Citation
Do Buyer and Seller Motivations Affect Transaction Prices in Commercial Real Estate Markets? Evidence from Tax-Deferred Exchanges and Other Conditions of Sale

Material Information

Title:
Do Buyer and Seller Motivations Affect Transaction Prices in Commercial Real Estate Markets? Evidence from Tax-Deferred Exchanges and Other Conditions of Sale
Creator:
PETROVA, MILENA T. ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Conditional sales ( jstor )
Market prices ( jstor )
Motivation ( jstor )
Prices ( jstor )
Property relinquishment ( jstor )
Property replacement ( jstor )
Sales representatives ( jstor )
Sales transactions ( jstor )
Taxes ( jstor )
Taxpaying ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Milena T. Petrova. Permission granted to University of Florida to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
8/31/2006
Resource Identifier:
485046703 ( OCLC )

Downloads

This item has the following downloads:

petrova_m ( .pdf )

petrova_m_Page_039.txt

petrova_m_Page_041.txt

petrova_m_Page_078.txt

petrova_m_Page_137.txt

petrova_m_Page_094.txt

petrova_m_Page_050.txt

petrova_m_Page_038.txt

petrova_m_Page_128.txt

petrova_m_Page_106.txt

petrova_m_Page_105.txt

petrova_m_Page_008.txt

petrova_m_Page_030.txt

petrova_m_Page_139.txt

petrova_m_Page_102.txt

petrova_m_Page_076.txt

petrova_m_Page_065.txt

petrova_m_Page_133.txt

petrova_m_Page_123.txt

petrova_m_Page_045.txt

petrova_m_Page_015.txt

petrova_m_Page_079.txt

petrova_m_Page_058.txt

petrova_m_Page_009.txt

petrova_m_Page_100.txt

petrova_m_Page_063.txt

petrova_m_Page_093.txt

petrova_m_Page_084.txt

petrova_m_Page_028.txt

petrova_m_Page_054.txt

petrova_m_Page_064.txt

petrova_m_Page_107.txt

petrova_m_Page_027.txt

petrova_m_Page_081.txt

petrova_m_Page_066.txt

petrova_m_Page_148.txt

petrova_m_Page_113.txt

petrova_m_Page_069.txt

petrova_m_Page_056.txt

petrova_m_Page_049.txt

petrova_m_Page_047.txt

petrova_m_Page_098.txt

petrova_m_Page_142.txt

petrova_m_Page_135.txt

petrova_m_Page_089.txt

petrova_m_Page_140.txt

petrova_m_Page_037.txt

petrova_m_Page_031.txt

petrova_m_Page_024.txt

petrova_m_Page_074.txt

petrova_m_Page_143.txt

petrova_m_Page_035.txt

petrova_m_Page_004.txt

petrova_m_Page_059.txt

petrova_m_Page_032.txt

petrova_m_Page_090.txt

petrova_m_Page_029.txt

petrova_m_Page_068.txt

petrova_m_Page_101.txt

petrova_m_Page_018.txt

petrova_m_Page_072.txt

petrova_m_Page_051.txt

petrova_m_Page_132.txt

petrova_m_Page_095.txt

petrova_m_Page_110.txt

petrova_m_Page_134.txt

petrova_m_Page_149.txt

petrova_m_Page_138.txt

petrova_m_Page_125.txt

petrova_m_Page_055.txt

petrova_m_Page_136.txt

petrova_m_Page_092.txt

petrova_m_Page_013.txt

petrova_m_pdf.txt

petrova_m_Page_073.txt

petrova_m_Page_083.txt

petrova_m_Page_014.txt

petrova_m_Page_046.txt

petrova_m_Page_120.txt

petrova_m_Page_150.txt

petrova_m_Page_070.txt

petrova_m_Page_001.txt

petrova_m_Page_103.txt

petrova_m_Page_085.txt

petrova_m_Page_034.txt

petrova_m_Page_007.txt

petrova_m_Page_127.txt

petrova_m_Page_087.txt

petrova_m_Page_005.txt

petrova_m_Page_077.txt

petrova_m_Page_033.txt

petrova_m_Page_086.txt

petrova_m_Page_111.txt

petrova_m_Page_151.txt

petrova_m_Page_152.txt

petrova_m_Page_057.txt

petrova_m_Page_003.txt

petrova_m_Page_067.txt

petrova_m_Page_099.txt

petrova_m_Page_010.txt

petrova_m_Page_096.txt

petrova_m_Page_023.txt

petrova_m_Page_011.txt

petrova_m_Page_122.txt

petrova_m_Page_141.txt

petrova_m_Page_082.txt

petrova_m_Page_118.txt

petrova_m_Page_115.txt

petrova_m_Page_036.txt

petrova_m_Page_071.txt

petrova_m_Page_091.txt

petrova_m_Page_117.txt

petrova_m_Page_147.txt

petrova_m_Page_052.txt

petrova_m_Page_017.txt

petrova_m_Page_144.txt

petrova_m_Page_088.txt

petrova_m_Page_060.txt

petrova_m_Page_108.txt

petrova_m_Page_025.txt

petrova_m_Page_124.txt

petrova_m_Page_016.txt

petrova_m_Page_153.txt

petrova_m_Page_126.txt

petrova_m_Page_097.txt

petrova_m_Page_026.txt

petrova_m_Page_131.txt

petrova_m_Page_053.txt

petrova_m_Page_075.txt

petrova_m_Page_104.txt

petrova_m_Page_130.txt

petrova_m_Page_114.txt

petrova_m_Page_022.txt

petrova_m_Page_040.txt

petrova_m_Page_112.txt

petrova_m_Page_042.txt

petrova_m_Page_109.txt

petrova_m_Page_020.txt

petrova_m_Page_145.txt

petrova_m_Page_006.txt

petrova_m_Page_021.txt

petrova_m_Page_012.txt

petrova_m_Page_121.txt

petrova_m_Page_062.txt

petrova_m_Page_146.txt

petrova_m_Page_044.txt

petrova_m_Page_129.txt

petrova_m_Page_119.txt

petrova_m_Page_048.txt

petrova_m_Page_019.txt

petrova_m_Page_116.txt

petrova_m_Page_002.txt

petrova_m_Page_061.txt

petrova_m_Page_080.txt

petrova_m_Page_043.txt


Full Text












DO BUYER AND SELLER MOTIVATIONS AFFECT TRANSACTION PRICES IN
COMMERCIAL REAL ESTATE MARKETS?
EVIDENCE FROM TAX-DEFERRED EXCHANGES AND OTHER CONDITIONS
OF SALE















By

MILENA T. PETROVA


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2006


































Copyright 2006

by

Milena Petrova
































This dissertation is dedicated to my parents.















ACKNOWLEDGMENTS

I would like to sincerely thank David Ling, my dissertation chair, for his guidance

and support and making sure I successfully complete this work. I would also like to thank

Wayne Archer, Andy Naranjo, Ronald Ward, Mahen Nimalendran and Thomas Barkley

for their helpful comments and suggestions. I would like to gratefully acknowledge the

valuable suggestions by participants at the AREUEA doctoral meeting in Boston and

seminar participants from the University of Florida, Syracuse University, Baruch College

and Florida International University. I thank CoStar Group, Inc. for providing the data

used in this dissertation. Finally, I thank my parents and my husband, for their patience

and support.

















TABLE OF CONTENTS



page

A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TABLES.. .. ............... .. .. .... ......................... vii

LIST OF FIGURES ........................ ........................... ix

ABSTRACT .....................x.............. ......................

1 INTRODUCTION .................. ................ ...... .... .......... .. .......... 1

2 TAX-DEFERRED EXCHANGES..........................................................................7

Tax-Deferred Exchanges Overview ................................ ...............7
Types of Tax-Deferred Exchanges.................... .............................................8
Basic Requirements of a Valid Section 1031 Exchange ..............................10
Advantages of Tax-Deferred Exchanges.......................................................12
Drawbacks of Tax-Deferred Exchanges....................................................12
Exchanges and Price Effects ........................................................ 13

3 OTHER ATYPICAL MOTIVATIONS ..........................................15

Purchases by Out-of-State Buyers ............................................................. .. ...15
Condominium Conversion..................................... ......... 16
Portfolio Sales................................... .........17
Sale-Leasebacks...................................... ............................... .........18

4 THEORETICAL MOTIVATION FOR PRICE EFFECTS AND CONCEPTUAL
FRAMEW ORK .................. ........................... 20

Theoretical M otivation for Price Effects .............................................. ......20
Price Pressure H ypothesis ............................................ ............... 21
Im perfect Substitute Hypothesis.............................. ................... 22
Tax Capitalization Hypothesis ................................ ............... 22
Other Factors ........................................ ........23
Conceptual Fram ew ork.................................... ............... 23

5 SIMULATING THE MAGNITUDE OF PRICE EFFECTS FOR EXCHANGES ...31



v










6 DATA AND M ETHODOLOGY ........................................................ 41

D ata ...................................... ..................................................41
Apartments ......................................................44
Office Properties................ ....... ............... 48
Retail Properties ................................................ ........ 51
R research M methodology ............................................................53

7 RESULTS FOR RESIDENTIAL REAL ESTATE..................................................61

8 RESULTS FOR OFFICE PROPERTIES.............. ............ ...............84

9 RESULTS FOR RETAIL PROPERTIES .........................................108

10 CONCLUSION................... .. ............ ..........133

BIOGRAPHICAL SKETCH ........................ ...... .........................142
















LIST OF TABLES


Table page

1 Incremental NPV of Apartment Exchange as a Percent of Replacement Property
V alue ............................................................34

2 Incremental NPV of Non-Residential Exchange as a Percent of Replacement
Property Value.................... .. ...... .................. 39

3 Description of Size of Exchange M market ............................................................. 43

4 Description of Apartment Property Sales by Markets ..........................................46

5 Description of Office Property Sales by Markets ..................................... 49

6 Description of Retail Property Sales by Markets............. ....................52

7 List of Regression Variables ......................... ......... ...........58

8 Summary Statistics of Apartment Data by M arkets............. ............................... 62

9 Differences in Mean Prices of Control Sample and Identified Groups of Interest
for Apartment Properties.......... .................................66

10 Regression Statistics for OLS Model with Structural Characteristics and
Submarket Dummies by Apartment Markets............... ..............69

11 Regression Statistics for OLS Model with Structural Characteristics, Submarket
Dummies and Longitude, Latitude Coordinates by Apartment Markets .............75

12 Marginal Effects for Significant Coefficients for Variables of Interest................ 81

13 Summary Statistics of Office Data by Markets................................................ 86

14 Differences in Mean Prices of Control Sample and Identified Groups of Interest
for Office Properties................ .... .............. 90

15 Regression Statistics for OLS Model with Structural Characteristics and
Submarket Dummies by Office Markets....................... ..................93

16 Regression Statistics for OLS Model with Structural Characteristics, Submarket
Dummies and Longitude, Latitude Coordinates by Office Markets ....................102









17 Marginal Effects for Significant Coefficients for Variables of Interest in Office
Regressions...................................... ......... 107

18 Summary Statistics of Retail Data by M markets .....................................................109

19 Differences in Mean Prices of Control Sample and Identified Groups of Interest
for R detail Properties .............. ...................... ....... ....... .. ...... ........115

20 Regression Statistics for OLS Model with Structural Characteristics and
Submarket Dummies by Retail Markets .................... ..... .............118

21 Regression Statistics for OLS Model with Structural Characteristics, Submarket
Dummies and Longitude, Latitude Coordinates by Retail Markets................. 125

22 Marginal Effects for Significant Coefficients for Variables of Interest in Retail
Regressions...................................... .........13 1
















LIST OF FIGURES

Figure page

1 D irect E change (Sw ap) ........................................ .................9

2 Delayed Exchange with Intermediary............................. ...............9

3 Impact of Atypical Motivation on Price ........................................................21
















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

DO BUYER AND SELLER MOTIVATIONS AFFECT TRANSACTION PRICES IN
COMMERCIAL REAL ESTATE MARKETS?
EVIDENCE FROM TAX-DEFERRED EXCHANGES
AND OTHER CONDITIONS OF SALE

By

Milena T. Petrova

August 2006


Chair: David C. Ling
Major Department: Finance, Insurance and Real Estate

Heterogeneous buyer and seller motivations are common in real estate transactions.

However, there are only a few studies limited to one property type and market that

examine the effect of atypical motivations on observed selling price. I investigate several

heterogeneous motivations that could have an effect on prices of apartment, office and

retail properties in fifteen major metropolitan markets. In particular, I focus on analyzing

transactions motivated by tax-deferred exchanges. I find a significant positive price effect

related to replacement property exchanges. This result is robust across geographic

markets and property types. For example, I observe a price premium for replacement tax-

deferred exchanges ranging from 5% in Seattle to 18% in Denver for apartment

transactions, from 19% in Seattle to 36% in Chicago for office sales, and from 12% in

Seattle to 45% in Chicago for retail sales. This dissertation represents the first work that









measures the size of the exchange market nationwide, and defines conceptually and

empirically the magnitude of the effect of exchanges on transaction prices in different

markets across the country and across property types. The results imply that participants

in tax-delayed exchanges, need to be careful because the price they pay in the form of

higher replacement property price may offset, in whole or in part, the gain from

deferment of taxes.

I also analyze properties purchased by out-of-state buyers, sales that are part of

sale-leasebacks, portfolio transactions, or condominium conversions. I find that these

special motivations are generally associated with positive price premiums. These results

have important implications since they demonstrate not only that various atypical

motivations of buyers and sellers have an impact on transaction prices in commercial real

estate, but also that this impact is sensitive to the market, as well as to the property type.















CHAPTER 1
INTRODUCTION

I study the role that heterogeneous buyer and seller motivations play in determining

sales price in commercial real estate. Generally, price models are built based on the

assumptions that market participants have homogeneous motivations. In appraisal

literature, homogeneous motivation is central to the concept of market value.1 However,

various conditions of sale which can be viewed as distinct motivations appear to be quite

common in commercial real estate transactions. The literature documents multiple

situations in which sellers have heterogeneous motivation (Geltner, Kluger and Miller

(1988), Quan and Quigley (1991), Sirmans, Turnbull and Dombrow, 1995, Glower,

Haurin, Hendershott (1998)). Common atypical motivations include tax-deferred

exchanges, bank-foreclosed property sales, liquidation sales, eminent domain sales,

purchases by real estate investment trusts, purchases by tenants, purchases by out-of-state

buyers, condo conversions, sales-and-leasebacks, and portfolio sales.

Analyzing the impact of investor motivations on sales price is important for several

reasons. First, the sales comparison approach, which is one of the most widely used

methods for valuation in real estate, usually considers similar properties based on

1 Market value is defined as the following: "The most probable price which a property should bring in a
competitive and open market under all conditions requisite to a fair sale, the buyer and seller, each acting
prudently, knowledgeably and assuming the price is not affected by undue stimulus. Implicit in this
definition is the consummation of a sale as of a specified date and the passing of title from seller to buyer
under conditions whereby: (1) buyer and seller are typically motivated: (2) both parties are well informed
advised, and each acting in what he considers his own best interest: (3) a reasonable time is allowed for
exposure in the open market: (4) payment is made in terms of cash in U. S. dollars or in terms of financial
arrangements comparable thereto; and (5) the price represents the normal consideration for the property
sold unaffected by special or creative financing or sales concessions by anyone associated with the sale."
National Residential Real Estate Appraisal Institute, 2006









structural characteristics and location. However, if conditions of sale have an effect on

price, then investors need to be aware of such impact and know how to adjust

correspondingly.

Second, some motivations appear to have different roles in different parts of the

country. For example, during the last six years tax-deferred exchanges represented

approximately 19 percent of total commercial real estate sales. However, in some areas of

the country, and in particular in the West Coast, exchanges represented close to 50

percent of all sales. In addition, exchanges tend to be more frequent for apartments than

for other types of commercial real estate, such as office and retail. This raises questions

as to whether the impact of sale conditions varies by markets, as well as property type.

Therefore, while analyzing the impact of buyer and seller motivation is important, it is

not sufficient to conduct the analysis based on one market and property type, which is a

common feature of the existent literature. By conducting a comprehensive analysis across

major metropolitan markets, as well as property types, this dissertation can be used as an

important reference by appraisers for the adjustment of prices when the sales comparison

approach is used.

Finally, although there is a large body of literature devoted to examining the

influence of structural characteristics and location on sales price, only a few papers

examine the impact of atypical motivations on price. Slade (2004) in a review article

provides a summary of possible conditions that influence motivation. He discusses four

major conditions that cause atypical motivations:

* Tax-Deferred Exchanges
* Bank-Foreclosed Properties
* Acquisitions by REITs
* Out-of-State Buyers









He also mentions two other conditions that may influence transactions price:

reduced marketing period (liquidation) and eminent domain.

In his review, Slade (2004) identifies one study on tax-deferred exchanges (Holmes

and Slade, 2001), seven papers on bank-foreclosed properties (Shilling et al., 1990,

Forgey et al., 1994, Hardin and Wolverton, 1996, Lambson, McQueen and Slade, 2004,

Downs and Slade, 1999, and Munneke and Slade, 2000, 2001), two published studies

involving sales by REITs (Hardin and Wolverton, 1999, and Lambson, McQueen and

Slade, 2004), and two published papers examining the influence out-of-state buyers

(Lambson, McQueen and Slade, 2004, and Turnbull and Sirmans, 1993).

Holmes and Slade's analysis (2001) shows that in the Phoenix market, tax-deferred

exchanges are associated with a 7.9 percent price premium. Bank-foreclosed apartments

are associated with a 22 23 percent price discount (Hardin and Wolverton, 1996;

Lambson, McQueen and Slade, 2004). In bank-foreclosed offices, Munneke and Slade

(2000, 2001) find discounts ranging from 11 percent to 31 percent. In apartment

acquisitions by REITs, different studies find premiums ranging from zero to 28 percent

(Hardin and Wolverton, 1996, Lambson, McQueen and Slade, 2004). In properties sold

to out-of-state buyers, Lambson, McQueen and Slade (2004) observe a 5 to 7 percent

premium for apartments.

A related stream of literature analyzes the impact of seller's motivation on selling

time. For example, in a home sellers survey, Glower, Haurin and Hendershott (1998) find

that a seller, who at the time of listing has a planned time to move, sells more quickly

than one that does not.









The majority of these studies are based on either homes or residential real estate.

Only Downs and Slade (1999) and Munneke and Slade (2000, 2001) study property types

that are different from residential in this case offices. All of the previous studies, with the

exception of one, focus on one market and one property type. The study by Hardin and

Wolverton (1999) analyses REIT purchases in three markets Atlanta, Phoenix and

Seattle.

I examine several conditions of sale which represent distinct motivations that are

frequently seen in comparable sales data and could influence sales price. In particular, I

focus attention on the use of tax-deferred exchanges nationwide and their effect on

observed transaction prices.

Tax-deferred exchanges are transactions in which a taxpayer is able to defer

payment of some, or all, of the federal income taxes associated with the disposition of

real property by acquiring another property of "like kind." Although Section 1031 of the

Internal Revenue Code (IRC) dates back to the 1920's, exchanges under the original

restrictions could only be completed as a simultaneous swap of properties among two or

more parties. The required simultaneous exchange of property severely limited the

usefulness of Section 1031 exchanges as a tax deferral tool because of the difficulty of

synchronizing the close of two or more complex transactions.

Only in 1984, in response to an earlier court decision related to the "Starker" case

(Starker vs. United States, 602 F. 2d 1341 (9th cir., 1979)), did Congress amend the

original regulations to allow taxpayers more time to complete the transaction. More

specifically, a taxpayer who initiates a Section 1031 tax-deferred exchange has up to 45

days after the disposition of the "relinquished" property to identify a replacement









property and 180 days (135 beyond the 45-day period) to complete the delayed exchange

by acquiring the replacement property (Internal Revenue Code Section, Title 26, Section

1031). Nevertheless, the Section 1031 exchange market did not fully evolve until 1991

when the Internal Revenue Service (IRS) issued final regulations for initiating and

completing delayed Section 1031 exchanges.

The use of tax-deferred exchanges has grown significantly during the last decade.

For example, in 2004, an estimated 80 percent of all commercial real estate transactions

on the West Coast involved the use of an exchange (McLinden, 2004).

Many real estate practitioners argue that taxpayers should use exchanges whenever

possible, despite higher average transaction/selling costs than "regular" fully taxable

sales. However, such advice may not fully account for the potential price discount

associated with the sale of the relinquished property in an exchange or the price premium

that may be required to obtain the replacement property. Separately or together, these

price effects may fully or partially offset the tax deferral benefits of an exchange.

Since most tax-deferred exchanges are negotiated and closed in private markets,

quantification of the size and scope of the real estate Section 1031 exchange market has

not been possible. In addition, very little is known about the effects of tax-deferred

exchanges on observed transaction prices and whether the effects have varied over time,

across geographic markets, or across property types. In fact, only a handful of papers

have investigated the pricing of Section 1031 exchanges (Holmes and Slade, 2001;

Lambson, McQueen and Slade, 2004). As discussed above, a major weakness of the prior

studies is that they examine just one market and one property type (the Phoenix

apartment market); therefore the results are difficult to generalize.









This dissertation represents the first work that defines conceptually and empirically

the magnitude of the effect of exchanges on transaction prices in different markets across

the country and across property types. I also examine possible price effects associated

with properties that are purchased by out-of-state buyers, as well as sales that are part of

condo conversions, portfolio transactions, or sale-leaseback transactions.

The remainder of the dissertation is organized as follows. Chapter 2 presents an

overview of tax-deferred exchanges and discusses their potential advantages and

disadvantages. Chapter 3 discusses out-of-state buyer motivations, condo conversions,

sale-leasebacks and portfolio sales. Chapter 4 reviews the theoretical background for

observing price effects associated with atypical motivations, develops a conceptual

framework for quantifying the benefits of Section 1031 exchanges, and simulates a range

of possible transaction price effects. Chapter 5 presents a simulation analysis of the

theoretical benefits from using an exchange. In Chapter 6, I present the empirical data set

and research methodology. Chapter 7 contains the results of the empirical analysis for

residential real estate. Chapters 8 and 9 present the results of the empirical analysis for

office and retail properties, correspondingly. Finally, in Chapter 10, I summarize the

results and offer some concluding comments.















CHAPTER 2
TAX-DEFERRED EXCHANGES

Tax-Deferred Exchanges Overview

Realized gains from the sale or exchange of real property must generally be

recognized for federal income purposes (Internal Revenue Code Section 1001(c)). In

general, the realized gain is equal to the net selling price of the property minus the

adjusted tax basis.1 However, under Section 1031 of the IRC, real estate owners who

dispose of their investment, rental, or vacation property and reinvest the net proceeds in

other "like kind" property are able to defer recognition of the capital gain realized on the

sale of the relinquished property.

It is important to note that a Section 1031 exchange is, strictly speaking, a tax

deferral technique. The taxpayer's basis in the replacement property is set equal to the

transaction price of the replacement property, minus the gain deferred on the disposition

of the relinquished property. Therefore, when (if) the replacement property is

subsequently disposed of in a fully taxable sale, the realized gain will equal the deferred

gain plus any additional taxable gain realized since the acquisition of the replacement

property. However, if the subsequent disposition of the replacement property is also

structured in the form of a Section 1031 exchange, the realized gain can again be

deferred.2


1 The adjusted tax basis is equal to the original cost basis (including the value of the land), plus the cost of
any capital improvements undertaken since acquisition of the property, minus cumulative depreciation.

2 Tax deferral turns into permanent tax savings upon the death of the taxpayer because the basis of the
property is "stepped-up" to its current fair market value. Thus, the taxpayer's heirs can dispose of the









In order for the exchanging taxpayer to avoid completely the recognition of the

accrued taxable gain, he or she must acquire a property of equal or greater value than the

relinquished property. In addition, the taxpayer must use all of the net cash proceeds

generated from the sale of the relinquished property to purchase the replacement

property. The transaction is taxable to the extent that (1) the value of the replacement

property is less than the value of the relinquished property and (2) there is cash left over

after the purchase of the replacement property.

Types of Tax-Deferred Exchanges

There are a number of ways in which a Section 1031 exchange can be structured

involving two or more of the following parties:

* Taxpayer: elects to relinquish his property via a Section 1031 exchange.
* Seller: owns the real estate that the taxpayer acquires as the replacement property.
* Buyer: purchaser of the taxpayer's relinquished property.
* Qualified Intermediary: independent agent who facilitates the exchange. The
qualified intermediary (QI) takes an assignment of rights in the sale of the
relinquished property and the purchase contract for the replacement property. In
short, the QI buys and then resells the properties for a fee.

Although rare, the two-party exchange is the purest form of exchange. The

transaction involves two parties who simultaneously exchange ("swap") properties. Title

to the relinquished property is conveyed by the taxpayer to the seller and title to the

replacement property is conveyed by the seller to the taxpayer. The two party exchange

in depicted in Figure 1.

Since the swapped properties are rarely of equal value, the party with the least

valuable position will have to pay cash (or its equivalent) to the other party in order to

balance the equity positions. Cash received as part of the transaction must be recognized


property in a fully taxable sale and not have to pay taxes on gains deferred through one or more Section
1031 exchanges.









as a taxable gain in the year of the exchange. Thus, taxpayers exchanging into less

valuable properties lose a portion of the tax deferral benefits associated with like-kind

exchanges.3


Relinquished
Property

Replacement
Property


Figure 1. Direct Exchange (Swap)

Most Section 1031 transactions are "delayed" exchanges that involve the use of a

qualified intermediary. In a delayed exchange, ownership of the relinquished property is

transferred to the buyer. However, the buyer of the relinquished property transfers the

agreed-upon cash amount to the QI, not the taxpayer. This first phase of the delayed

exchange, often referred to as the taxpayer's "down-leg," is depicted in the top portion of

Figure 2.


Cash FIRST PHASE
("DOWN-LEG")






Cash SECOND PHASE
("UP-LEG")


Figure 2. Delayed Exchange with Intermediary



3 To the extent an exchanging taxpayer must recognize a portion of the realized gain because of the receipt
of cash, the taxpayer's basis in the replacement property is higher and, therefore, any subsequent realized
gain from the sale of the replacement property will be lower.


TAXPAYER


I SELLER


TAXPAYER^H


TAXPAYER^H










The cash paid by the buyer is "parked" with the QI until the taxpayer is able to identify

and close on the replacement property.

Within 45 days of sale of the relinquished property, the taxpayer must "identify"

the replacement property. The identification must be specific, such as the address of the

property to be acquired. To allow for the possibility that the taxpayer may not be able to

come to terms with the owner of the potential replacement property, the taxpayer may

designate more than one replacement property.4

The taxpayer must acquire one or more of the identified replacement properties

within 180 days of the date of the closing of the relinquished property; that is, the 45 and

180 day periods run concurrently. There are no exceptions to these time limits and failure

to comply will convert the transaction to a fully taxable sale. At the closing of the

replacement property, the QI transfers cash to the seller of the replacement property and

the seller transfers ownership to the taxpayer. This second phase of the delayed exchange,

often referred to as the taxpayer's "up-leg," is depicted in the bottom portion of Figure

2.5

Basic Requirements of a Valid Section 1031 Exchange

In general, both real and personal property can qualify for tax-deferred treatment.

However, some types of property are specifically disqualified; for example, stocks,

bonds, notes, and ownership interests in a limited partnership or multi-member limited



4 More specifically, the taxpayer can (1) identify up to three properties of any value or (2) identify more
than three properties so long as their combined values do not exceed 200 percent of the value of the
relinquished property.

5 If the taxpayer closes on the replacement property prior to the sale of the relinquished property, the
transaction becomes a ic\ iii" exchange. This dataset does not include a large enough sample of reverse
exchanges to examine empirically how the pricing of such exchanges varies from delayed exchanges and
fully taxable sales. I therefore do not discuss reverse exchanges here.









liability company.6 Both the relinquished property and the replacement property must be

held for productive use in trade or business or held as a "long-term investment." Thus,

personal residences and property held for sale to consumers (i.e., "dealer" property)

cannot be part of a Section 1031 exchange.7 A holding period equal to or greater than one

year is commonly assumed to qualify the relinquished property as a long-term investment

for the purposes of implementing a tax-deferred exchange; however, the one year rule of

thumb has no basis in statutory or case law.

For a transaction to qualify as a Section 1031 exchange there must (1) be a

reciprocal exchange (rather than a sale for cash) and (2) the exchange must involve "like-

kind" property. An exchange is clearly created by the use of a QI and the required

exchange documentation. Like kind means "similar in nature or character." In fact,

virtually any real estate is like-kind to any other real estate. However, real property is not

like-kind to personal property. Therefore, for example, a warehouse cannot be exchanged

for jewelry. In addition, foreign property cannot be exchanged for U.S. property. With the

development of IRS regulations concerning Section 1031, tax-deferred exchanges are

also used to trade lines of business, such as such as television and radio stations,

newspapers, distributorships, and franchises, including, among others, sports teams, beer

distributorships, and professional service practices (McBurney, 2004). Because a line of

business includes multiple classes of assets real, personal and intangible property an

exchange for each class needs to be completed (McBurney and Boshkov, 2003;

McBurney, 2004).


6 Since 2003, a percentage ownership interest as a tenant-in-common (TIC) is qualified property for the
purposes of a Section 1031 exchange. The taxpayer, however, must be careful that the TIC has been
structured to avoid its re-characterization by the IRS as a partnership for federal income tax purposes.

7 Vacation homes will only qualify if they have been rented out the majority of the year.









Advantages of Tax-Deferred Exchanges

The tax literature and popular press point to several motivations for use of Section

1031 exchanges. First, exchanges serve as an effective shelter from taxes, thereby

preserving investment capital. In addition, exchanges can be used to upgrade portfolios

(Fickes, 2003). By deferring taxes, the taxpayer can also leverage appreciation and afford

to acquire a larger/higher priced replacement property. Section 1031 exchanges can also

be used to consolidate or diversify properties, exchange low-return properties for high-

return properties, or to substitute depreciable property for non-depreciable property

(Wayner, 2005a and 2005b).

Drawbacks of Tax-Deferred Exchanges

Despite the advantages of tax-deferral, Section 1031 exchanges have several

drawbacks. First, the taxpayer's basis in the replacement property is set equal to the

market value of the replacement property, minus the deferred gain. Thus, the larger the

amount of tax-deferral, the smaller is the depreciable basis in the replacement property

and, therefore, the smaller is the allowable deduction for depreciation. Moreover, the

larger the amount of tax-deferral, the larger will be the realized gain if and when the

replacement property is subsequently disposed of in a fully taxable sale.

A second disadvantage is that the transaction costs (both monetary and non-

monetary) associated with initiating and completing the exchange will likely exceed the

costs of a fully taxable sale. The additional costs may include settlement fees,

intermediary fees, and attorney preparation fees (Wayner, 2005b). These first two

disadvantages are explicitly considered in the conceptual model presented in Chapter 4.

An additional disadvantage is that Section 1031 exchanges do not allow for the

recognition of a loss for tax purposes. Thus, taxpayers will avoid using exchanges if they









have not realized a positive capital gain. Also, unlike the proceeds from a "cash out"

refinancing, tax-deferred exchanges do not provide a method for drawing tax-free cash

out of the relinquished property. This is because any cash received in the year of the

exchange is fully taxable.

Exchanges and Price Effects

If a taxpayer is successful in completing a simultaneous or delayed tax-deferred

exchange, the realized tax liability will be deferred until the replacement property is

subsequently disposed of in a fully taxable sale. A portion of the realized gain will be

recognized in the tax year in which the exchange occurs to the extent that the value of the

relinquished property exceeds the value of the replacement property. The present value of

the income tax deferral benefit is therefore a function of the magnitude of the deferred

capital gain, the expected holding period of the replacement property, and the applicable

discount rate.

A taxpayer entering into a tax-deferred exchange can afford to accept an offer for

the relinquished property that is lower than the investment value he or she places on the

property by an amount that is equal to, or less than, the present value of the income tax

deferral benefit. That is, depending on current market conditions, including liquidity, the

negotiating abilities of the taxpayer and potential buyers, and whether or not potential

buyers are aware the taxpayer is initiating a Section 1031 exchange, the selling taxpayer

may be willing or required to "share" a portion of the expected tax deferral benefits with

the buyer of the relinquished property. Since Section 1031 exchanges are a tax-deferral

technique, sellers will not enter in exchanges, unless their property has appreciated in

value and (or) has significantly depreciated. If avoiding capital gains tax is the main

motivation for participating in an exchange then all else equal relinquished properties will









have higher values than properties that are not part of an exchange. Therefore, the net

effect of discussed above factors is not clear. I expect to find that prices paid for

relinquished properties will be higher or less than transaction prices in fully taxable sales,

all else equal.

In contrast, taxpayers face significant compliance risk when seeking to complete

the second leg of a tax-deferred exchange by identifying and purchasing a replacement

property within the 45- and 180-day time limits. The strict time requirements imposed by

IRS regulations, in addition to the complicated nature of tax-deferred exchanges, can lead

to a temporary increase in demand in order for investors to find replacement properties

and close the exchanges in a timely manner. In perfect markets, a temporary increase in

demand by exchange motivated investors has no effect on market prices because supply

can instantaneously respond. However, commercial real estate markets are known to be

"thin" and less elastic. Therefore, I expect that buyers closing on a tax-deferred exchange

transaction may be willing or required to give up some of their benefits from deferring

taxes and pay a premium for the replacement property relative to its fair market value in

order to acquire their property within the time constraint. In a competitive market the

amount of the price premium will not exceed the expected present value of tax deferral,

with the actual magnitude of the premium again depending on market liquidity, the

negotiating abilities of the taxpayer and potential sellers, and whether or not potential

sellers are aware the taxpayer is attempting to complete a Section 1031 exchange.














CHAPTER 3
OTHER ATYPICAL MOTIVATIONS

Purchases by Out-of-State Buyers

Anecdotal evidence suggests that out-of-state buyers, especially from higher-cost

areas, pay more for real estate than in-state buyers, especially those residing in lower cost

areas. This observation is explained by buyers "anchoring" to the higher values in their

home area and therefore being willing to pay a premium for real estate in lower-cost

areas. An example of anchoring are sales in neighboring states to California. Investors

from California will be more willing to offer a premium for a property in Arizona or Las

Vegas, since they are anchoring on the high prices of real estate in their home state. The

"anchoring" phenomenon is explained by behavioral literature. Slovic and Lichtenstein

(1971) and Tversky and Kahneman (1974) were the first academics to discuss heuristics

and biases. In the real estate literature, Northcraft and Neale (1987) presented strong

evidence of anchoring in property pricing that was similar for both amateurs and real

estate professionals. Other real estate studies that find evidence for anchoring include

Black and Diaz (1996), Diaz and Hansz (1997), Diaz and Wolverton (1998), and Diaz,

Zhao and Black (1999). Additional evidence of anchoring in real estate is seen in

appraisal smoothing. A large body of literature discusses smoothing in appraisal based

indexes, due to appraisal values lagging true prices and being too reliant on historical

prices (Geltner (1989) and Webb (1994)).

Tumbull and Sirmans (1993) attribute observed out-state buyers' price premiums to

higher search costs. In their model, buyers with higher search costs will search less than









investors with lower search costs, and therefore will tend to overpay on average.

Lambson, McQueen and Slade (2004) identify three factors that contribute to observed

premiums: biased beliefs, high search costs and time pressure, namely the haste

associated with out-of-state buyers. Their paper examines 2,854 apartment sales in

Phoenix during 1990 mid 2002 and finds a 5 7 percent premium associated with

purchases by out-of-state buyers. One limitation of the Lambson, McQueen and Slade

study is that it is based on one market and one property type. Therefore, it is not clear to

what extent their findings carry over to other property types and markets. The objective

of this paper is to examine not only whether there is any price premium associated with

purchases by out-of-state buyers, but also to establish to what extent such premiums vary

across markets and by property types.

Condominium Conversion

The conversion of rental properties to condominium ownership have been very

popular in the recent past, induced by increasing house prices and lagging rent levels.

This trend is expected to slow or be reversed in 2006 given an oversupply of condos and

shortage of apartments in some markets. In competitive markets the motivation of the

condo converter should not have an impact on sales prices. However, as Lambson,

McQueen and Slade (2004) point out "apartment complexes trade infrequently with high

transaction costs... and real estate buyers have heterogeneous information" (p. 86).

Therefore, an investor that is buying a multifamily property with the objective of

converting its units to condos may be willing to pay a premium because of the expected

higher price he will net per condo sold.









Portfolio Sales

Previous studies on portfolio sales focus on the stock price effects from

announcements of portfolio sales. A portfolio sale is defined as a transaction in which

two or more unrelated properties are sold to the same buyer. Studies that are based on

property sales prior to 1992 found no abnormal returns associated with portfolio sales

announcements. Glascock, Davidson and Sirmans (1991) examined 51 real estate

portfolio purchases prior to 1986 and find that abnormal returns are insignificantly

different from zero. McIntosh, Ott and Liang (1995) reached a similar conclusion based

on a 54-transaction sample during 1968-1990, in which all of the acquirers are Real

Estate Investment Trusts (REITs). Booth, Glascock and Sarkar (1996) studied a sample

of 94 portfolio acquisitions and also failed to observe any significant abnormal returns.

These findings are consistent with expectations in competitive markets, since they signify

that no change in shareholder wealth results whether assets are acquired and sold

separately or together.

However, a more recent study based on the "modern" post-1992 REIT market by

Campbell, Petrova and Sirmans (2003) finds average abnormal returns for the acquirer of

approximately 0.5 percent in a study based on 209 portfolio acquisitions during 1995-

2001. As the authors point out, these results suggest a clear change in the regime of such

transactions, post 1992, when new regulations relaxed the restrictions on ownership

concentrations for Real Estate Investment Trusts and thereby led to an increase of capital

flows to REITs, especially by institutions. Campbell, Sirmans and Petrova (Ibid.) present

evidence that the observed returns are related to the positive effect of reconfirming

geographical focus, signaling by taking on private debt and private placement of stock

with institutions.









None of the existing studies addresses whether properties with similar

characteristics will sell at a premium (discount) if they are part of a portfolio transaction.

By bundling together several properties in one transaction, buyers may enjoy acquisition

economies of scale, decreased transaction costs, as well as decreased search costs. A

portfolio acquisition can provide the buyer with quick exposure or focus in a desired

geographical or property market. However, such transactions are also more complex to

negotiate and complete. All else equal, a portfolio buyer may be willing to give up some

of the expected benefits from acquisition and pay a premium to obtain a desired portfolio.

The premium paid will also depend on the market power of the real estate portfolio

buyer. Therefore, in cases involving large influential buyers (e.g. REITs, or institutions) a

discount associated with real estate portfolio purchase may be observed. Anecdotal

evidence from the real estate professional press suggests that portfolio acquisitions by

REITs are frequently the quickest and cheapest way for REITs to acquire properties.

They can consequently turn around and dispose of less attractive properties.

Sale-Leasebacks

In a sale-leaseback transaction, a firm sells an asset, such as real estate property or

equipment, to another firm and simultaneously leases it back. Academic research focuses

on tax motivation as the major source of value creation in corporate sale-leasebacks (see

Miller and Upton (1976), Lewellen, Long and McConnell (1976), Myers, Dill and

Bautista (1976), Brealey and Yong (1980)). Recent studies including Smith and

Wakeman (1984), Alvayay, Rutherford and Smith (1995), Moyer and Krishnan (1995)

and Lasfer and Levis (1998) confirm the importance of tax related motivations to lease or

buy, but also acknowledge other non-tax incentives. Leasing offers benefits to the lessor

by increased non-debt tax shields through depreciation. Sale-leasebacks create value









when there is a difference between applicable tax rates of the lessor and lessee, namely

the lessor is in a higher tax bracket, while the lessee is in a lower tax bracket. Studies on

the effect of sale-leaseback announcements on the lessee's and lessor's share prices

record a positive effect on the lessee's share price (see Slovin, Sushka and Poloncheck

(1990), Allen, Rutherford and Springer (1993), Ezzell and Vora (2001) and Fisher

(2004)). Slovin, Sushka and Poloncheck (1990) conclude that the observed positive

market reaction to announcements of a sale-leaseback is due to the perception of

reduction of present value of expected taxes and present evidence that gains from sale-

leasebacks accrue only for the lessees. In similar spirit, Lewis and Schallheim (1992)

assume that leasing offers the opportunity for transferring non-debt tax shields and posit

that if the lessee firm can locate a lessor firm who is more able to enjoy such tax shields

then "the buyer will pay more than they are worth to the lessee" (p.498, Ibid.). According

to the authors, this higher price takes the form of reduced lease payments.

In this study, sale-leasebacks are viewed as another example of atypical motivation.

Under the Tax Capitalization Hypothesis and consistent with the expectations by Lewis

and Schallheim (1992), I hypothesize that any expected tax benefits by the lessor may be

capitalized into a higher purchase price. Therefore, everything else equal, properties that

are part of sale-leaseback transactions will have higher sales price.















CHAPTER 4
THEORETICAL MOTIVATION FOR PRICE EFFECTS
AND CONCEPTUAL FRAMEWORK

Theoretical Motivation for Price Effects

Generally, price models assume that buyers (sellers) are homogeneous in their

motivation to buy (sell). Various conditions of sale, which can be viewed as distinct

motivations appear to be quite common in commercial real estate transactions. Such

conditions of sale can impact transaction prices. Motivated buyers can create a temporary

increase in demand in property markets. In perfect markets, a temporary increase in

demand by motivated investors has no effect on market prices because supply can

instantaneously respond. However, in commercial real estate markets the supply of

available properties is less elastic to shocks in demand (Wheaton and Torto, 1990; Eppli

and Shilling, 1995). Jones and Orr (1999) point to differences in elasticity across real

estate markets, with inelastic supply most severe in retail and office properties.

Therefore, equilibrium price may change in order to eliminate excess demand. In sales

motivated by a tax-deferred exchange, an out-of-state buyer, a condo-converter, a sale-

leaseback transaction or a portfolio sale, I expect a positive effect on selling price, due to

increased demand based on any of these atypical motivations. This effect is presented in

Figure 3.

In finance theory, both the Price Pressure Hypothesis and the Imperfect Substitute

Hypothesis allow for changes in equilibrium price in response to shocks in demand.










Price with
Ln (Sales Price) Atypical
Motivation




-- True Price






Square Feet of
Improvements


Figure 3. Impact of Atypical Motivation on Price

Price Pressure Hypothesis

The Price Pressure Hypothesis (PPH), developed first by Scholes (1972) posits that

if the number of shares outstanding is increased by a secondary offering or large block

sales, investors need to be offered a "sweetener" in the form of a reduced share price, so

that they are willing to hold more shares. This leads to a temporary decline in share

prices. Scholes (1972) tests this hypothesis by examining 1,200 secondary distributions

of large block sales during 1947-1965, but finds no significant price effect.

Kraus and Stoll (1972) further develop the PPH and posit that in an imperfect

market, with few investors, trading may produce significant price changes if the

expectations of the marginal seller of the security are different from the marginal buyer.

For example, it may be difficult for a large seller to distribute his shares at the same price.

As a result, two types of distribution effects may arise. The first type is due to liquidity

costs. This is a temporary effect in which the different costs of finding willing investors

can move the transaction price away from the equilibrium price. In support of the PPH,









Kraus and Stoll find evidence for some form of distribution effect. Mikkelson & Partch

(1985), Hess & Frost (1982), and Harris & Gurel (1986) also find empirical evidence

supporting the PPH.

Imperfect Substitute Hypothesis

The second hypothesis that supports price changes in response to excess demand is

the Imperfect Substitute Hypothesis (ISH). The ISH is developed as the second

distribution effect discussed by Kraus & Stoll (1972). This effect is due to different

investor preferences for a given security. This is a permanent price effect, and depends on

the number of investors and on the substitutability of one security for another. In public

securities markets, the Imperfect Substitute Hypothesis allows for equilibrium price

changes to eliminate the excess demand. Evidence of demand induced price shocks can

be found in the impact of large block trades on prices (Scholes (1972), Kraus and Stoll

(1972) and Mikkelson and Parch (1985)).

Section 1031 exchanges may create a similar tax-induced demand shock. In

particular, the compliance risk in delayed exchanges could create increased acquisition

demand, which may force the taxpayer to share some of the expected tax-deferral benefits

with the seller of the replacement property in the form of an increased purchase price. In

addition, if the seller of the replacement property is aware that the taxpayer is seeking to

complete a tax-deferred exchange, the taxpayer's bargaining position is clearly

compromised.

Tax Capitalization Hypothesis

In competitive markets, the value of commercial real estate fully reflects the current

and expected future tax treatment of depreciable assets (Hendershott and Ling, 1984, and

Ling and Whinihan, 1985). Expected increases in tax liabilities are capitalized in the









value of the assets, resulting in lower asset values. The use of a tax-free exchange

constitutes a "reverse tax-capitalization" (Holmes & Slade, 2001). The buyer can

distribute some of his expected tax benefits due to tax-deferral in order to outbid other

potential buyers.

With sale-leasebacks, leasing offers the opportunity for transferring non-debt tax

shields (Lewis and Schallheim, 1992). Under the Tax Capitalization Hypothesis and

consistent with the expectations by Lewis and Schallheim (1992), I hypothesize that any

expected tax benefits by the lessor can be expressed in the form of a higher purchase

price.

Other Factors

With out-of-state buyers a possible premium in price can be explained by

anchoring. Evidence for anchoring in real estate is given by Northcraft and Neale (1987),

Black and Diaz (1996), Diaz and Hansz (1997), Diaz and Wolverton (1998), and Diaz,

Zhao and Black (1999). A second explanation for a possible premium involves higher

search costs for out-of-state buyers (Tumbull and Sirmans, 1993). Time pressure to buy

is a third possible explanation for expected price premiums (Lambson, McQueen and

Slade, 2004).

Conceptual Framework

The following framework will focus on conceptualizing the net benefits from using

an exchange. Assume a taxpayer who owns an income producing property has decided

that the risk-return characteristics of her portfolio would be enhanced by disposing of the

asset and reinvesting his (her) equity into a replacement property located in a market with

more growth potential. Assume also that the replacement property has already been

identified. The first strategy available to the taxpayer is to dispose of the existing









property in a fully taxable sale and then use the net proceeds, along with additional equity

capital, to acquire the replacement property. The second option is to take advantage of

Section 1031 of the IRC and exchange out of the existing property and into the

replacement property. The second strategy would allow the taxpayer to defer recognition

of the taxable gain that has accrued on the existing property. The net present value of the

sale-purchase strategy, NPVSALEt, assuming all-equity financing, can be represented as

S(1- o)I, DEP,," p,2 SC -2 ,gCG 2, ZdREC4P,2,s
NPVSALE, =(ATSP, -pt) +- 7, (1)
z=1 (1+k)' (1+k)"
where:
ATSP' = the net after-tax proceeds from the sale of the existing property at
time t;
Pt = the acquisition price of the replacement property at time t;
ro = the taxpayer's marginal tax rate on ordinary income;
I, = the expected net cash flow of the replacement property in year i of
the expected n-year holding period;
DEP2," = allowable depreciation on the replacement property in year i,
conditional on a sale-purchase strategy;
k = the required after-tax rate of return on unlevered equity;
P 2 = the expected price of the replacement property in year t+n;
SC,, = expected selling costs on the disposition of the replacement
property in year t+n;
Tg = the tax rate on capital gain income;
CG,2 = expected capital gain income on the sale of the replacement
property in year t+n, conditional on a sale-purchase strategy;
Td = the tax rate on depreciation recapture income; and
RECAP,, = depreciation recapture income on the sale of the replacement
property in year t+n, conditional on a n-year sale-purchase
strategy.

The first term on the right-hand-side of equation (1) represents the additional equity

capital that must be invested at time t under the sale-purchase strategy, and is equal to the

after-tax proceeds from a fully taxable sale minus the acquisition price of the replacement

property at time t. As is detailed below, if the price of the replacement property is equal









to the price of the existing property, then ATSP,' P,2 is equal to total taxes due on the

sale of the existing property, plus total selling costs.

The second term on the right-hand-side of equation (1) represents the cumulative

present value of the replacement property's net cash flows from annual operations, plus

the present value of the annual tax savings generated by depreciation. Annual

depreciation, DEP2 ", is equal to


DEP2 = L (2)
RECPER

where P2 is the acquisition price of the replacement property, L is the percentage of P2

that represents non-depreciable land, and RECPER is the allowable cost recovery period

for the replacement property.1 Since the replacement property is purchased with the

proceeds from a fully taxable sale, the original tax basis of the replacement property is

"stepped up" to equal the total acquisition price, P2, thereby maximizing allowable

depreciation deductions over the expected n-year holding period.

The third and final term on the right-hand-side of equation (1) represents the

expected after-tax cash proceeds from the sale of the replacement property at the end of

the assumed n-year holding period. Deducted from the expected selling price of the

replacement property at time t+n are the following: expected selling costs (SC,2), the

expected capital gain tax liability (cg CG,2 ), and the expected depreciation recapture

tax (dr RECAP, ).



1 Congressional legislation has repeatedly altered the period of time over which rental real estate may be
depreciated. Currently, residential real property (e.g., apartments) may be depreciated over no less than 27
and 1/2 years. The cost recovery period for nonresidential real property (e.g., shopping centers, industrial
warehouses, and office buildings) is 39 years.










Under current federal income tax law, all taxable income from property sales must

be classified as either ordinary income, capital gain income, or depreciation recapture

income. The distinctions are important because capital gain income, under the tax rules in

place in 2006, is subject to a maximum 15 percent tax rate.2 In contrast, the maximum

statutory rate on ordinary income is 35 percent.

Assuming the taxpayer's asset is classified as trade or business property and has

been held for more than one year, the total taxable gain on the sale of the replacement

property in year t+n has the following two components

CG 2,s =(P -SC2) -UNDBASIS2,s (3)

and


RECAP,2, = DEp2, (4),


where UNDBASIS,2, is the un-depreciated cost basis of the replacement property at time

t+n. More specifically, UNDBASIS 2 is equal to acquisition price of the replacement

property (P,2), plus any capital expenditures over the n-year holding period.3 Note that

the magnitude of UNDBASIS2, is conditioned upon whether the replacement property

was acquired via an exchange or with a sale-purchase strategy. Also note that CG2, is

the amount by which the original acquisition price of the replacement property (plus any

subsequent capital improvements) is expected to increase in nominal value over the n-



2 From 1997 to May 6, 2003, the maximum capital gain tax rate was 20 percent.

3 According to the IRS, a capital expenditure increases the market value of the property. In contrast,
expenditures deemed by the IRS to be "operating" expenses maintain, but do not fundamentally alter, the
market value of the property. Capital expenditures are not depreciable in the year in which they are
incurred. Rather, they are added to the tax basis of the property and then systematically expensed through
annual depreciation deductions.









year holding period. Also, RECAP,2" captures the portion of the total taxable gain on sale

that results from depreciation. Total taxes due on the sale of the replacement property in

year t+n, conditional on a sale-purchase strategy, are therefore expected to be

TDS, = c CG2 + drRECAP2,. (5)

Next, the components of the after-tax proceeds from a fully taxable sale of the

relinquished property at time t (the first term on the right-hand-side of equation (1)) are

examined. Note that

ATSP' = (P' SC) -, gCG] rdRECAP1, (6)

where P' is equal to the price of the relinquished property and SCd is equal to total

selling costs at time t. Similar to the subsequent sale of the replacement property at time

t+n, the capital gain and depreciation recapture portions of the total taxable gain from the

sale of the relinquished property at time t are

CG\ = (P1 SC ) UNDBASIS} (7)

and

h
RECAP,' = DEf (8)


where UNDBASIS} is the undepreciated cost basis of the relinquished property at time t

andDEP1 is equal to


DEP1 _(1 L0 )^h (9)
RECPER

Pth represents the acquisition price of the existing property when purchased h

years ago, L,_ is the percentage of the original acquisition price that was non-









depreciable, and RECPER is the allowable cost recovery period for the existing property.

Total taxes due on the sale of the relinquished property at time t are therefore expected to

be

TDS = TcgCG + TdrRECAlP (10)

The second acquisition option available to the taxpayer is to take advantage of

Section 1031 and exchange into the replacement property. The net present value of the

exchange strategy, assuming all-equity financing, can be represented as

NPVEX 1 EC 2 (- oDEP2,e pt2 n-SC2 cgCG,2- p p2,e
NPVEXt = t -(ECI -k)( +k)
=1 (1+k)j (1+k)"

(11)

where:

EC = the total cost of exchanging out of the relinquished property and
into the replacement property at time t;
DEP"2, = depreciation on the replacement property in year i, conditional on
an exchange strategy;
CG2 = the expected capital gain income on the sale of the replacement
property in year t+n, conditional on an exchange strategy; and
RECAP2,e = depreciation recapture income on the sale of the replacement
property in n years assuming an exchange at time t.

All other variables are as previously defined. The capital gain and recapture

components of the total taxable gain on the sale of the replacement property at time t+n,

conditional on an exchange strategy are

CG 2,e(P2 -SC2 )-UNDBASIS2,e (12)

and


RECAP =iDEP2 (13)









where UNDBASIS2, is equal to the acquisition price of the replacement property (p,2)

minus the taxable gain that was deferred at time t (DEFGAINt) by executing an exchange

strategy. DEFGAIN, = CG} +RECAP,', and DEP2 e is equal to


12 -LP2 DEFGAIN )
DEP2,e _(1 L D t (14)
RECPER

where L^ is the percentage of the replacement property's acquisition price that is non-

depreciable and RECPER is the allowable cost recovery period for the existing property.

Note that reducing the tax basis of the replacement property by the amount of the

deferred gain insures thatDEP2,s > DEP2 ". Total taxes due on the sale of the

replacement property in year t+n, conditional on an exchange strategy at time t, are

expected to be

TDS, = cgCG, +rRECAP2. (15)

The taxpayer should exchange into the replacement property if the net present value

of the exchange strategy [equation (11)] exceeds the net present value of the sale-

purchase strategy [equation (1)]. To determine the net present value of the sale-purchase

strategy, equations (7) and (8) are first substituted into equation (6). Equations (2), (3),

(4), and (6) are then substituted into equation (1). To determine the net present value of

the exchange strategy, equations (12), (13), and (14) are substituted into equation (11).

Finally, subtraction of equation (1) from equation (11) produces the following expression

for the incremental NPV of the exchange strategy









NCNPV SC EC (DEI -DEP2,e) dr (RECAP RECAP)2,e)
INCVP V, = [SCr ECt + TDSt ] 7 +
1 (1+k)' (1+k)"

S(CG,- _CG, )
+nn(16)
(l+k)"

The first term in equation (16), [SC1 EC, + TDS} ], captures the immediate net

benefit of tax deferral. Note that if the time t selling costs associated with the sale-

purchase strategy and exchange strategy are equal, the advantage of the exchange is equal

to TDS,, the deferred tax liability. To the extent exchanges are more expensive to

execute than sales, SC} EC, will be negative and this incremental outflow will be

netted against the positive deferral benefits.

As noted above, the tax basis of the replacement property is reduced by the amount

of the taxable gain deferred by the exchange, which insures that DEP2 > DEp2 ". The

second term in equation (16) captures the cumulative present value of the foregone

depreciation deductions over the n-year holding period. However, to the extent annual

depreciation deductions are reduced by an exchange, the amount of depreciation

recaptured when the replacement property is sold in year t+n is reduced by an exchange.

The present value of the reduced depreciation recapture taxes is reflected in the third term

in equation (16).

Finally, because the tax deferral associated with an exchange reduces the tax basis

in the replacement property, the taxable capital gain due on the sale of the replacement

property will be larger with an exchange. The negative effects of the increased capital

gain tax liability on the incremental NPV of an exchange are captured by the fourth term

in equation (16).














CHAPTER 5
SIMULATING THE MAGNITUDE OF PRICE EFFECTS FOR EXCHANGES

As previously discussed, taxpayers face significant compliance risk when seeking

to complete a tax-deferred exchange. Moreover, the exchanging taxpayer may have

compromised his or her bargaining position with potential sellers of replacement

properties. As a result, the taxpayer may be forced to pay a premium for the replacement

property, assuming the marginal (price determining) buyers and sellers in the market are

not motivated by Section 1031 tax deferral benefits and compliance issues. In a

competitive market, the magnitude of the price discounts accepted by sellers of

relinquished properties and the price premiums paid by acquirers of replacement

properties should not exceed the incremental NPV of the exchange strategy, with the

actual magnitude of the premium depending on market liquidity, the negotiating abilities

of the taxpayer and other potential buyers and sellers, and whether or not potential buyers

and sellers are aware the taxpayer is attempting to complete a Section 1031 exchange.

Before turning to the empirical estimates of the price discounts offered by sellers

of relinquished properties and the price premiums paid by purchasers of replacement

properties, equation (16) is used to simulate the magnitude of INCNPVt under a number

of plausible assumptions. Simulated values of INCNPV, are then divided by the price of

the replacement property at time t to determine the percentage price effect. These

simulations are intended to quantify the maximum percentage price effects that are likely

to be found in the subsequent empirical work.









To solve equation (16) numerically, the following base-case assumptions for the

parameter values are made:

* Price of relinquished and replacement property: Pt = Pt
* Cost recovery period (RECPER): 27.5 years for residential and 39 years for non-
residential commercial properties
* Selling costs (SCI and SC,'): 3 percent of sale price
* Exchange costs (ECt): equal to SCI
* Ordinary income tax rate (r,): 35 percent
* Capital gain tax rate (cg): 15 percent
* Depreciation recapture tax rate (rd,): 25 percent
* After-tax discount rate (k): 8 percent
* Non-depreciable portion of original tax basis (L h_ and Lz ): 20 percent

The price of the replacement property is assumed to be equal to the price of the

relinquished property to abstract for any effects unequal equity positions would have on

time t inflows and outflows as well as future depreciation deductions. Note that the

assumed magnitude of p1 = p2 does not affect the numerical simulation results because

INCNPVt is divided by the price of the replacement property to produce a percentage

price effect. Other key variables in the calculation of INCNPVt include the number of

years since acquisition of the relinquished property, HOLD1, the annualized rate of price

appreciation since acquisition of the relinquished property, 7r1, and the expected holding

period of the replacement property, HOLD2.1

Table 1 represents the simulation results for residential commercial real estate. The

top panel in Table 1 contains the base case simulation results. One pattern is noteworthy:

the incremental value of an exchange is unambiguously positively related to HOLD1. For

example, assuming HOLD2 = 8, HOLD1 = 5, and 7nc = 6 percent, INCNPVt is equal to



1 It is straightforward to show that the value of INCNPV, from equation (16) is not affected by the rate at
which the replacement property is expected to appreciate in nominal value.









2.48 percent of P,2. This implies the taxpayer could afford to pay up to a 2.48 percent

premium for the replacement property, assuming they did not agree to a price discount on

the sale of the relinquished property.

As HOLD1 increases to 10, the maximum price impact rises from 2.48 percent to

4.03 percent. Assuming HOLD1 = 20, the maximum price impact increases further to

5.33 percent. In short, the relative attractiveness of the exchange strategy is

unambiguously positively related to the magnitude of the accumulated gain, all else

equal.

The relation between INCNPVt and 7c1, however, for a given HOLD1 is less clear.

For example, assuming HOLD1 = 5, increased price appreciation produces slight

increases in INCNPVt. However, with HOLD1 = 20, higher values of 7 produce lower

values of INCNPVt. With HOLD1 = 10, the relation between x7t and INCNPVt is sensitive

to the assumed value of HOLD2.

All else equal, the value of tax deferral increases with its duration. However, the

top panel of Table 1 indicates that INCNPV, increases with HOLD2, but at a decreasing

rate. For expected holding periods longer than eight to ten years, INCNPVt is largely

unaffected by increases in HOLD2 and, in fact, for holding periods in excess of 16 years

the value of INCNPVt begins to decrease.

The premium price effect in Panel A ranges from 1.85 percent to 9.00 percent of

value. These results clearly indicate that any price discounts or premiums observed in the

data are likely to vary depending on the magnitude of the taxpayer's accumulated gain on

the relinquished property. However, the size of the taxpayer's accumulated taxable gain

is not observable in the data set. The maximum price premiums displayed in Panel B of











Table 1. Incremental NPV of Apartment Exchange as a Percent of Replacement Property
Value
Holding Period of Replacement Property Holding Period of Replacement Property
(HOLD') (n') 4 8 12 16 20 24 28 4 8 12 16 20 24 28


Panel A:
Tg =15%,k=8%,
EC, =SC t
5 2%
5 6%
5 10%
5 20%
10 2%
10 6%
10 10%
10 20%
20 2%
20 6%
20 10%
28 2%
28 6%
28 10%
Panel C:
T = 15%,k=8%,
EC, =1.2* SC
5 2%
5 6%
5 10%
5 20%
10 2%
10 6%
10 10%
10 20%
20 2%
20 6%
20 10%
28 2%
28 6%
28 10%
Panel E:
cg =20%,
k=8%,EC, SC1
5 2%
5 6%
5 10%
5 20%
10 2%
10 6%
10 10%
10 20%
20 2%
20 6%
20 10%
28 2%
28 6%
28 10%


Panel B:
T' =15%,k=10%,
EC, SC 1


1.85 2.14 2.27 2.30 2.28 2.23 2.17
2.00 2.48 2.69 2.75 2.71 2.63 2.53
2.11 2.74 3.03 3.10 3.06 2.95 2.81
2.30 3.21 3.61 3.72 3.66 3.51 3.31
3.45 4.01 4.26 4.33 4.29 4.20 4.07
3.20 4.03 4.40 4.50 4.44 4.30 4.12
3.03 4.04 4.49 4.61 4.54 4.37 4.15
2.81 4.05 4.61 4.75 4.67 4.46 4.19
5.80 6.80 7.25 7.36 7.29 7.13 6.91
4.12 5.33 5.87 6.02 5.93 5.73 5.46
3.35 4.67 5.26 5.41 5.32 5.10 4.81
7.04 8.29 8.85 9.00 8.91 8.71 8.43
4.15 5.50 6.11 6.27 6.17 5.95 5.65
3.19 4.57 5.19 5.36 5.26 5.03 4.72


1.33 1.60 1.73 1.76 1.74 1.69 1.63
1.47 1.94 2.15 2.21 2.17 2.10 1.99
1.58 2.21 2.49 2.56 2.52 2.42 2.28
1.78 2.67 3.08 3.18 3.12 2.97 2.77
2.92 3.48 3.73 3.79 3.75 3.66 3.54
2.67 3.49 3.86 3.96 3.90 3.77 3.58
2.50 3.50 3.95 4.07 4.00 3.83 3.61
2.29 3.52 4.07 4.21 4.13 3.92 3.65
5.28 6.26 6.71 6.82 6.75 6.59 6.37
3.59 4.80 5.34 5.48 5.39 5.19 4.93
2.83 4.13 4.72 4.87 4.78 4.56 4.27
6.52 7.76 8.32 8.46 8.38 8.17 7.89
3.63 4.97 5.57 5.73 5.63 5.41 5.11
2.66 4.04 4.65 4.82 4.72 4.49 4.19


1.46 1.93 2.20 2.33 2.39 2.40 2.38
1.89 2.69 3.15 3.38 3.47 3.49 3.45
2.24 3.30 3.90 4.21 4.33 4.35 4.30
2.84 4.36 5.21 5.65 5.83 5.86 5.78
2.77 3.71 4.24 4.51 4.62 4.64 4.59
3.15 4.54 5.32 5.72 5.88 5.91 5.84
3.39 5.08 6.03 6.51 6.71 6.75 6.66
3.71 5.78 6.95 7.55 7.79 7.83 7.73
4.76 6.42 7.36 7.84 8.04 8.07 7.99
4.32 6.35 7.50 8.08 8.32 8.36 8.26
4.12 6.32 7.56 8.19 8.45 8.49 8.38
5.85 7.94 9.12 9.72 9.97 10.01 9.90
4.59 6.85 8.13 8.78 9.04 9.09 8.97
4.17 6.49 7.80 8.46 8.74 8.78 8.66


2.01 2.35
2.25 2.84
2.45 3.23
2.79 3.90
3.75 4.44
3.64 4.66
3.57 4.80
2.79 3.90
6.34 7.55
4.76 6.25
4.06 5.66
7.71 9.24
4.87 6.53
3.93 5.62


2.51 2.55 2.55 2.51
3.10 3.18 3.16 3.11
3.57 3.67 3.66 3.58
4.38 4.53 4.51 4.41
4.74 4.83 4.82 4.76
5.11 5.24 5.22 5.13
5.35 5.51 5.49 5.37
4.38 4.53 4.51 4.41
8.09 8.25 8.23 8.12
6.91 7.11 7.07 6.94
6.37 6.59 6.55 6.40
9.91 10.12 10.09 9.95
7.26 7.48 7.44 7.29
6.37 6.60 6.56 6.40


Panel D:
T =15%,k=10%,
E&, 1.2* SC ,
1.48 1.81 1.96
1.72 2.30 2.55
1.92 2.68 3.02
2.26 3.35 3.84
3.22 3.90 4.19
3.11 4.12 4.56
3.04 4.26 4.80
2.95 4.45 5.11
5.80 7.01 7.54
4.23 5.71 6.36
3.52 5.12 5.82
7.18 8.70 9.37
4.34 5.99 6.71
3.40 5.08 5.82
Panel F:
cg =20%,
k=10%,EC,=SCt
1.66 2.22 2.52
2.23 3.18 3.68
2.69 3.95 4.61
3.49 5.27 6.23
3.17 4.28 4.87
3.74 5.38 6.25
4.11 6.10 7.16
4.60 7.04 8.34
5.47 7.43 8.48
5.19 7.58 8.86
5.06 7.65 9.03
6.74 9.21 10.52
5.56 8.22 9.64
5.16 7.89 9.35


2.00 2.00 1.97 1.92
2.63 2.62 2.56 2.49
3.12 3.11 3.04 2.94
3.99 3.96 3.86 3.72
4.29 4.27 4.21 4.12
4.69 4.67 4.58 4.46
4.96 4.94 4.83 4.67
5.31 5.28 5.14 4.96
7.71 7.68 7.57 7.42
6.56 6.53 6.39 6.21
6.04 6.00 5.86 5.66
9.57 9.54 9.40 9.21
6.93 6.89 6.74 6.54
6.05 6.01 5.86 5.65




2.66 2.72 2.74 2.72
3.93 4.03 4.05 4.03
4.94 5.07 5.11 5.08
6.69 6.88 6.93 6.89
5.16 5.28 5.31 5.28
6.68 6.85 6.89 6.86
7.68 7.89 7.94 7.90
8.97 9.24 9.30 9.24
8.99 9.20 9.25 9.21
9.48 9.74 9.80 9.75
9.70 9.98 10.05 9.99
11.16 11.42 11.49 11.43
10.33 10.62 10.69 10.63
10.06 10.35 10.42 10.36









Table 1 are based on an increase in the assumed after-tax equity discount rate (k) to 10

percent from 8 percent. All other variables remain at their base case levels. Comparison

of Panel A and Panel B demonstrates that a higher discount rate unambiguously increases

the incremental value of the exchange option. This is because the value of tax deferral

produced by the exchange is immediate. In contrast, the foregone depreciation deductions

and the increased capital gain tax liability at sale that results from the decreased tax basis

both occur in subsequent years. Thus, the present value of these future cash outflows is

reduced by a higher discount rate. The net percentage price benefit from exchange in this

panel ranges from 2 percent to slightly above 10 percent.

The calculated price impacts reported in Panel C of Table 1 assume the discount

rate has been reset to 8 percent, but that the dollar costs of executing an exchange (ECt)

are 20 percent higher than the costs of a fully taxable sale (SCd). As expected, higher up-

front exchange costs reduce the maximum benefit of an exchange (relative to the base

case in Panel A). However, decreases in the maximum price impacts are modest,

averaging approximately one-half of a percentage point across varying assumptions for

HOLD1, HOLD2, and 70. The price benefits in this panel range from 1.33 percent to 8.46

percent.

In Panel D, the dollar costs of executing an exchange (ECt) are assumed to be 20

percent higher than the costs of a fully taxable sale (SCI) and the discount rate is set to

10 percent. Hence, Panel D has the same assumptions as Panel B, except for an increased

cost associated with exchange. This simulation represents a combination of higher after

tax discount rate, which has immediate positive effect to the incremental NPV and

increased cost of using an exchange which has a slight negative impact on the net benefit









of using an exchange. Therefore, the values in this panel are lower than in Panel B, but

higher than in Panel C. The maximum price impact is 9.57 percent for HOLD1 = 28,

HOLD2 = 16 and 7nc = 2 percent.

Panel E of Table 1 reports the results assuming a tax rate on capital gain income of

20 percent (the maximum statutory cg from 1999 to 2003). Clearly, the immediate value

of tax deferral is larger the higher is -C. However, the simulated price effects in Panel E

are not uniformly higher than those reported in Panel A. This is because a higher capital

gain tax rate will also increase the tax liability that results from the eventual sale of the

replacement property. The longer the expected holding period of the replacement

property, the more likely it is that the immediate tax deferral benefits associated with a

higher rC will exceed the present value of the increased taxes due on the subsequent sale

of the replacement property. This anticipated result is confirmed in panel E. That is,

increasing -rc from 15 percent to 20 percent produces larger values of INCNPV,, except

in some cases where the magnitude of the deferred gain is small (i.e., when HOLD1 and

7n1 are small) or when the expected holding period of the replacement property is

relatively short.

Panel F of Table 1 reports the results assuming a tax rate on capital gain income of

20 percent and after tax discount rate of 10 percent. Values tend to be higher than in

Panel B, where Tcg is set to 15 percent. However, similar to Panel E, values are only

higher for longer holding periods and appreciation rates. The difference between Panel E

and Panel F is that in Panel F values increase faster, which expresses the added positive

effect of the increased discount rate. The maximum price impact in this panel is 11.49

percent for HOLD1 = 28, HOLD2 = 24 and 7nt = 2 percent.









Overall, the maximum effect a Section 1031 exchange is likely to have on

observed transaction prices in a competitive residential market (when the marginal buyer

and seller are not exchange motivated) is estimated to range from about 1 percent to

approximately 11.5 percent.

In Table 2 I repeat the simulation analysis for non-residential real estate. The only

difference in this case is that non-residential real estate, such as office, industrial and

retail properties, has a 39 year cost recovery period (RECPER). All other assumptions in

the simulation analysis remain the same.

Table 2 reveals results similar to Table 1. All else equal, the value of tax deferral tends to

increase with its duration. Therefore, with the longer depreciation recovery period of 39

years, the maximum net benefit from using an exchange will also be higher. However,

with longer recovery period, benefits from depreciation each year are smaller compared

to if faster depreciation schedule is used. Hence, values in Table 2 will tend to be smaller

than the corresponding values in Table 1 for small appreciation rates. For appreciation

rate 7t1 = 10 percent and higher, the faster depreciation effect is offset by higher

appreciation and values in Table 2 become higher than the corresponding values in Table

1.

I record a maximum benefit based on all simulations of 13.25 percent in Panel F for

HOLD' = 39, HOLD2 = 39 and 7c1 = 2 percent. There is a U-shaped relationship between

INCNPVt and HOLD2. At first INCNPVt increases with HOLD2, it peaks at year 25 for

Panels A though D and then it decreases at a slow rate. With Panels E and F, where the

tax rate on capital gain income is set to 20 percent there is a strictly positive relationship









between INCNPVt and HOLD2; however, after year 20 INCNPVt increases at a

decreasing rate.

The premium price effects in Panel A vary from 1.51 percent to 9.85 percent of

value. In Panel B of Table 2 the assumed after-tax equity discount rate (k) is increased to

10 percent. Comparison of Panel A and Panel B demonstrates the incremental value of

the exchange is increased by as little as 0.14 percent, for holding periods of 5 years and

appreciation rate of 25 percent, to more than 1.3 percent, for holding periods of 39 years

for both the relinquished and the replacement property. The percentage price net benefit

from exchange in this panel ranges from 1.65 percent to close to 11 percent.

Panels C and D repeat the simulations of Panels A and B, with costs of executing

an exchange (ECt) set to be 20 percent higher than the costs of a fully taxable sale (SC,).

Higher up-front exchange costs again reduce the maximum benefit of an exchange.

Maximum price differences in both panels are about 55 percentage points lower than the

corresponding values in Panels A and B (9.30 percent in Panel C vs. 9.85 percent in

Panel A, and 10.40 percent in Panel D vs. 10.96 percent in Panel C).

Panels E and F of Table 2 report the results based on the same assumptions as

Panels A and B, but assuming a tax rate on capital gain income of 20 percent (the


maximum statutory rcg from 1999 to 2003). As with the residential simulations, the

simulated price effects in Panels E and F are not uniformly higher than those reported in

Panels A and B. A higher capital gain tax rate is also associated with increased tax

liability when the taxpayer eventually sells the replacement property in an ordinary sale.

With a higher capital gains tax the added benefit of using an exchange is larger for

higher appreciation rates and longer holding periods. The maximum price impact in












Table 2. Incremental NPV of Non-Residential Exchange as a Percent of Replacement
Property Value
Holding Period of Replacement Property Holding Period of Replacement Property
(HOLD ) (n&) 5 10 20 25 30 39 5 10 20 25 30 39


Panel A:
- 15%0,k=8%,oEC,=SC l
5 2% 1.51 1.81 2.02 2.03 2.02 1.99
5 6% 1.87 2.45 2.84 2.87 2.85 2.78
5 10% 2.15 2.95 3.50 3.53 3.51 3.41
5 20% 2.65 3.83 4.64 4.69 4.65 4.51
10 2% 2.85 3.47 3.89 3.91 3.89 3.82
10 6% 3.08 4.10 4.79 4.84 4.81 4.68
10 10% 3.23 4.51 5.39 5.45 5.41 5.25
10 20% 3.42 5.05 6.17 6.24 6.19 5.99
20 2% 4.86 5.98 6.73 6.78 6.75 6.61
20 6% 4.16 5.69 6.73 6.80 6.75 6.56
20 10% 3.85 5.56 6.73 6.80 6.75 6.54
30 2% 6.18 7.65 8.65 8.71 8.67 8.49
30 6% 4.39 6.14 7.33 7.41 7.35 7.14
30 10% 3.83 5.67 6.93 7.01 6.95 6.72
39 2% 6.92 8.62 9.78 9.85 9.80 9.59
39 6% 4.31 6.15 7.41 7.49 7.43 7.20
39 10% 3.74 5.61 6.89 6.97 6.91 6.68
Panel C:
T 150%,k=8%,EC, 1.2*sc }
5 2% 0.97 1.27 1.47 1.48 1.47 1.43
5 6% 1.33 1.90 2.29 2.32 2.30 2.23
5 10% 1.62 2.41 2.95 2.98 2.96 2.86
5 20% 2.12 3.29 4.09 4.14 4.10 3.96
10 2% 2.32 2.92 3.34 3.36 3.34 3.27
10 6% 2.54 3.55 4.24 4.29 4.25 4.13
10 10% 2.69 3.97 4.84 4.90 4.85 4.70
10 20% 2.89 4.51 5.61 5.69 5.63 5.44
20 2% 4.33 5.43 6.18 6.23 6.20 6.06
20 6% 3.63 5.15 6.18 6.24 6.20 6.01
20 10% 3.32 5.02 6.18 6.25 6.20 5.99
30 2% 5.65 7.10 8.10 8.16 8.12 7.94
30 6% 3.85 5.59 6.78 6.86 6.80 6.59
30 10% 3.30 5.13 6.37 6.45 6.39 6.17
39 2% 6.38 8.08 9.23 9.30 9.25 9.04
39 6% 3.78 5.61 6.86 6.93 6.88 6.65
39 10% 3.20 5.07 6.33 6.42 6.36 6.13
Panel E:
Trg 20%, k=8%,EC, =sc
5 20o0 1.29 1.77 2.17 2.24 2.27 2.27
5 6% 1.96 2.87 3.63 3.76 3.81 3.82
5 10% 2.49 3.74 4.80 4.98 5.05 5.05
5 20% 3.42 5.26 6.82 7.08 7.19 7.20
10 2% 2.52 3.48 4.29 4.43 4.48 4.49
10 6% 3.34 4.94 6.29 6.51 6.60 6.61
10 10% 3.89 5.90 7.60 7.89 8.00 8.01
10 20% 4.60 7.15 9.31 9.67 9.81 9.82
20 2% 4.40 6.13 7.60 7.85 7.95 7.96
20 6% 4.78 7.16 9.18 9.51 9.65 9.66
20 10% 4.95 7.62 9.89 10.26 10.42 10.43
30 2% 5.69 7.98 9.93 10.25 10.38 10.39
30 6% 5.29 8.03 10.34 10.73 10.88 10.90
30 10% 5.17 8.04 10.47 10.88 11.04 11.05
39 2% 6.47 9.13 11.38 11.75 11.90 11.92
39 6% 5.41 8.29 10.72 11.13 11.29 11.30
39 10% 5.18 8.10 10.58 10.99 11.16 11.17


Panel B:
r 15%,k=100%,EC,-SC t
1.65 1.99 2.21 2.23 2.23 2.20
2.13 2.80 3.22 3.25 3.24 3.20
2.52 3.44 4.01 4.06 4.04 3.99
3.19 4.55 5.40 5.46 5.44 5.36
3.13 3.84 4.28 4.32 4.31 4.26
3.54 4.71 5.45 5.50 5.49 5.42
3.81 5.29 6.22 6.29 6.27 6.18
4.17 6.04 7.22 7.30 7.28 7.16
5.37 6.65 7.45 7.51 7.49 7.41
4.86 6.61 7.71 7.79 7.77 7.66
4.63 6.59 7.83 7.92 7.89 7.77
6.85 8.54 9.60 9.67 9.65 9.55
5.19 7.20 8.46 8.55 8.53 8.40
4.67 6.78 8.11 8.20 8.18 8.05
7.69 9.65 10.88 10.96 10.94 10.82
5.15 7.26 8.59 8.69 8.66 8.53
4.59 6.74 8.09 8.19 8.16 8.03
Panel D:
r 15%, k=10%, EC, 1.2C'c I
1.11 1.44 1.66 1.67 1.67 1.65
1.59 2.25 2.66 2.69 2.68 2.64
1.98 2.89 3.46 3.50 3.48 3.43
2.65 3.99 4.84 4.90 4.88 4.80
2.59 3.29 3.73 3.76 3.75 3.70
3.01 4.16 4.89 4.94 4.93 4.86
3.28 4.74 5.66 5.73 5.71 5.62
3.63 5.49 6.66 6.74 6.72 6.60
4.84 6.10 6.89 6.95 6.93 6.85
4.32 6.06 7.15 7.23 7.21 7.10
4.09 6.04 7.27 7.36 7.33 7.21
6.31 7.99 9.04 9.11 9.09 8.99
4.65 6.65 7.90 7.99 7.97 7.84
4.13 6.23 7.55 7.64 7.62 7.49
7.16 9.10 10.32 10.40 10.38 10.26
4.61 6.71 8.03 8.13 8.10 7.97
4.05 6.19 7.53 7.63 7.60 7.47
Panel F:
r 200%,k=100%,EC, sc I
1.48 2.01 2.42 2.48 2.50 2.51
2.31 3.33 4.11 4.22 4.27 4.27
2.98 4.39 5.45 5.61 5.67 5.68
4.14 6.21 7.79 8.02 8.11 8.12
2.89 3.97 4.80 4.92 4.96 4.97
3.97 5.76 7.12 7.32 7.40 7.41
4.68 6.93 8.65 8.90 9.00 9.02
5.60 8.46 10.63 10.96 11.08 11.10
5.08 7.03 8.51 8.73 8.81 8.83
5.71 8.39 10.42 10.72 10.83 10.86
6.00 9.00 11.28 11.62 11.75 11.77
6.59 9.17 11.12 11.41 11.52 11.54
6.36 9.43 11.77 12.11 12.25 12.27
6.29 9.52 11.97 12.33 12.47 12.50
7.51 10.50 12.76 13.10 13.23 13.25
6.54 9.77 12.22 12.58 12.72 12.75
6.32 9.61 12.10 12.47 12.61 12.64









Panel E is 11.92 percent for HOLD' = 39, HOLD2 = 39 and 1 = 2 percent, while the

maximum benefit from exchange in Panel F is 13.25 percent for HOLD1 = 39, HOLD2

39 and 71 = 2 percent.

To summarize the results in Table 2, an increase in the after tax-cost of equity by

two percent, all else the same, is associated with a maximum increase in the incremental

NPV from completing an exchange of slightly over 1.3 percent. Increase, in the dollar

costs associated with an exchange, relative to the cost of completing a taxable sale, is

associated with approximately a 0.5 percent decrease in NPV, all else the same, in the

worst case. A capital gain tax rate of 20 percent, all else the same is associated with, at

most, approximately a 4.5 percent higher allowed price premium. Maximum price

benefits vary between 8.5 and 11.5 percent for apartments and between 9.3 and 13.3

percent for non-residential real estate.

The next chapter describes the data and empirical methodology used to measure the

size of exchange premiums and discounts actually observed in commercial real estate

markets over the 1999 to 2005 period.















CHAPTER 6
DATA AND METHODOLOGY

Data

Property sales data is obtained from CoStar Group, Inc. The CoStar Comps

Professional database includes historical information on over 1.2 million confirmed

commercial real estate transactions from 1999 through the first half of 2005. The CoStar

database includes all sales in excess of $250,000 in more than 40 major U.S. markets.1

Land, mobile homes, and special use properties are excluded from this analysis. The

initial sample contains 270,415 confirmed sales in five property markets: office,

industrial, apartment, retail and hotel/motel. CoStar Comps Professional has a separate

attribute field that identifies whether the sale represents an exchange. Therefore, any

missing exchange identification is due to a lack of information, rather than CoStar's

failure to report the type of transaction. An observation is eliminated from the sample if it

could not be determined whether the sale was a part of a Section 1031 exchange. This

further reduces the sample to 158,196 observations.

CoStar Comps Professional also contains descriptive information on the type of

exchange (e.g. taxpayer's sale of relinquished property, simultaneous exchange, reverse

exchange, etc.) in detailed notes. Based on the manual inspection of these notes, each

exchange property sale is placed into one of the following categories:

1. Seller's relinquished property in delayed (Starker) exchange
2. Buyer's replacement property in delayed (Starker) exchange


1 The CoStar product used in this analysis is CoStar Comps Professional
(www.costar.com/products/comps/).









3. Seller's relinquished and buyer's replacement property in two separate transactions
4. Direct exchange (swap)
5. Seller's relinquished property in reverse exchange
6. Buyer's replacement property in reverse exchange
7. Reverse exchange (type not confirmed)
8. Exchange into Tenants-in-Common
9. Other tax-deferred exchange exchange which cannot be categorized in any of the
above types or where CoStar was unable to confirm its type.

To assure reliability of the data, CoStar requires agents to physically inspect the

site and record a variety of property characteristics and transaction details. I therefore

exclude sales not confirmed by CoStar. In addition, I exclude all transactions with

recorded sales price below $250,000. CoStar covers comprehensively only transactions

that are above this threshold, although in some cases brokers do report smaller

transactions. For the sake of consistency such smaller transactions are eliminated from

the sample. The final sample has 124,830 transactions which facilitates a comprehensive

empirical investigation of the Section 1031 exchange market, as well as other atypical

motivations, not possible with previous datasets. Of the 124,830 usable sales transactions,

23,989 (or 19 percent) involved the use of a Section 1031 exchange.

Table 3 summarizes the number of transactions by year and property type. The year

1999 contains fewer transactions than year 2000. Also in 2005 there is only data for half

of the year through June 2005.

A substantial increase in the use of exchanges in commercial real estate markets

has been discussed in the popular press (e.g., McLinden, 2004). However, the data

necessary to support a comprehensive analysis has not been available. Table 3 breaks

down the sample by exchange and non-exchange transactions. The table reveals that

exchanges as percentage of all sales are very stable over the 1999-2005 sample period.

For example, between 1999 and 2000, the total number of commercial real estate











Table 3. Description of Size of Exchange Market
Property 1999 2000 2001 2002 2003 2004 Jun-05 Total
Type


Exchange
Non-
Apartment n
exchange
All

Exchange
Non-
Industrial
exchange
All


Office





Retail


Exchange
Non-
exchange
All

Exchange
Non-
exchange
All


Exchange
Non-
Hotel/ Motel None
exchange
All


Total


Exchange
Non-
exchange
All


587 2011 1774 2089 2096 2016

1291 4754 4505 4683 4728 4996

1878 6765 6279 6772 6824 7012

222 757 640 630 671 749

1172 4248 4210 4437 4469 4711

1394 5005 4850 5067 5140 5460

180 675 587 575 631 671

1028 3512 3430 3388 3519 3993

1208 4187 4017 3963 4150 4664

252 904 755 897 1023 958

1372 4836 4839 5300 5389 5186

1624 5740 5594 6197 6412 6144


16 47 41 37

88 338 328 301

104 385 369 338


1257 4394 3797 4228 4464 4437

4951 17688 17312 18109 18439 19245

6208 22082 21109 22337 22903 23682


636

1381

2017

231

1308

1539

241

1074

1315

294

1249

1543

10

85

95

1412

5097

6509


11209

26338

37547

3900

24555

28455

3560

19944

23504

5083

28171

33254

237

1833

2070

23989

100841

124830









exchanges grew from 1,257 to 4,394; however the exchanges' share of all verified CoStar

transactions remained in the 18-20 percent range. Notably, the year of 2003, which is the

year in which the maximum statutory capital gain rate was decreased from 20 percent to

15 percent, was not associated with a decrease in the number of exchanges. In fact, the

number of exchanges increased from 4,228 in the previous year to 4,464. A decrease in

the maximum capital gain rate as the simulation analysis suggests can possibly make an

exchange a less attractive alternative if the tax deferral is the main motivation to

participate in such a transaction.

Inspection of Table 3 reveals that exchanges are more frequently used in apartment

markets. Of the 37,547 verified apartment transactions in the sample, 11,209

(approximately 30 percent) involved the use of an exchange. Moreover, this percentage

has remained remarkably stable during 1999 2005. Among other property types,

exchanges generally account for 10-18 percent of all transactions.

Tables 4 through 6 present breakdowns of property sales for all 46 markets covered

by CoStar for apartment, office and retail properties, respectively. Markets are sorted

alphabetically to make it easier for the reader to locate his or her market of interest. The

number of transactions for each market and the percentage of total transactions by

property type are presented. The tables show that there is a substantial variability in terms

of transactions observed in different markets.

Apartments

Table 4 reveals that the largest apartment market is Los Angeles with 25 percent of

all sales. New York City is the second largest market with 12 percent of all transactions.

The smallest market (Charlotte) contains only 12 usable sales observations. The table

reveals that relatively few markets account for a major share of all apartment sales. For









example, more than half of all transactions are concentrated in less than 10 markets and

the top 20 markets account for 86 percent of all sales.

Table 4 also contains the percentage of exchanges observed in each market, as well

as the breakdown of the type of exchanges observed: relinquished, replacement,

relinquished and replacement, direct swaps, and other types. Table 4 clearly shows that

the use of exchanges varies substantially across the major markets. Importantly, in eight

markets (Marin-North SF Bay Area, Portland, Reno, Sacramento, San Diego, San

Francisco, San Jose and Seattle) exchanges represent the dominant form of property

transaction. Interestingly, all of these markets are located in the Western U.S., and most

are in California. This is consistent with anecdotal evidence that exchanges are more

common on the West Coast.

One factor cited for the wider use of Section 1031 exchanges in the Western US is

that exchanges are related to the "real estate booms in the West in the 1960s, 1970s and

1980s, which made investors more entrepreneurial" (McLinden, 2004). Another possible

explanation is the rapid appreciation of real estate in major metropolitan areas in

California in the last few years. Anecdotal evidence suggests that homes have appreciated

at an annual rate of 20 percent in Southern California over the last five years. In the other

major markets tracked by CoStar, exchanges have been much less prominent. For

example, over the 1999-2005 time period, only 1.6 percent of apartment sales in New

York involved an exchange.

The distribution of exchange types of shows that the number of relinquished and

replacement exchanges by markets is quite similar. This is expected, since for each

exchange involving a relinquished property, there should be at least one replacement







46


Table 4. Description of Apartment Property Sales by Markets


Name


Apts. % of Apt. % Reliq. Repl. .
All Total Exch. Both Direct Other
All Total Exch. Exch. Exch. Exch.


Los Angeles
New York City
San Diego
Chicago
Seattle
Phoenix
Oakland
Miami
San Francisco
Denver
Portland
Ft. Lauderdale
Riverside/San
Bernardino
Tucson
Washington, DC
San Jose
Dallas/Fort Worth
Colorado Springs
Boston
Tampa
Sacramento
Cincinnati/Dayton
Atlanta
Detroit/Toledo
Las Vegas
Fresno
Cleveland/Akron
North SF Bay Area
Houston
West Palm Beach
Stockton/Modesto
New Jersey
Orlando
Austin
Philadelphia
Minneapolis
Baltimore
Columbus
Jacksonville
Ventura
St Louis
Kansas City
Salt Lake City
San Antonio
Reno
Charlotte


9450 25.17% 3574 37.82% 1139
4586 12.21% 73 1.59% 20
2448 6.52% 1419 57.97% 477
1991 5.30% 377 18.94% 162
1686 4.49% 900 53.38% 207
1442 3.84% 143 9.92% 31
1398 3.72% 645 46.14% 256
1102 2.93% 81 7.35% 30
1029 2.74% 521 50.63% 275
1002 2.67% 473 47.21% 141
748 1.99% 516 68.98% 87
733 1.95% 85 11.60% 22

712 1.90% 283 39.75% 77
615 1.64% 212 34.47% 34
579 1.54% 31 5.35% 10
569 1.52% 287 50.44% 122
546 1.45% 129 23.63% 22
527 1.40% 38 7.21% 22
506 1.35% 64 12.65% 20
497 1.32% 37 7.44% 10
421 1.12% 261 62.00% 71
408 1.09% 53 12.99% 21
398 1.06% 24 6.03% 6
350 0.93% 44 12.57% 24
336 0.89% 125 37.20% 22
316 0.84% 107 33.86% 23
297 0.79% 20 6.73% 14
296 0.79% 183 61.82% 67
287 0.76% 55 19.16% 12
277 0.74% 27 9.75% 7
274 0.73% 121 44.16% 44
249 0.66% 8 3.21% 3
236 0.63% 16 6.78% 2
227 0.60% 32 14.10% 10
201 0.54% 19 9.45% 14
170 0.45% 65 38.24% 21
132 0.35% 13 9.85% 2
126 0.34% 55 43.65% 18
125 0.33% 7 5.60% 0
106 0.28% 44 41.51% 18
46 0.12% 10 21.74% 5
40 0.11% 10 25.00% 5
20 0.05% 7 35.00% 3
16 0.04% 3 18.75% 0
15 0.04% 8 53.33% 1
12 0.03% 4 33.33% 0


1348 514
27 5
490 360
114 41
349 99
68 17
200 152
42 4
137 94
189 106
163 102
46 5

108 66
120 51
10 1
83 63
43 8
8 5
38 1
20 2
104 73
24 4
8 0
12 0
82 12
54 13
5 1
55 50
34 2
13 4
35 29
4 0
9 1
7 4
1 0
26 10
10 1
24 11
3 1
11 11
2 1
5 0
2 2
1 0
6 1
3 1









exchange. However, if the taxpayer fails to identify replacement properties in a timely

fashion, or close the exchange according to the guidelines issued by the IRS, no

replacement exchange takes place. Also, it is common for the replacement property to be

of a different type than the original (relinquished property). Finally, replacement

properties may also be located in different markets than the original properties.

Table 4 shows that direct exchanges are quite rare. Out of 9,450 sales in Los

Angeles only 24 transactions represented direct swaps. This is also the largest number of

direct exchanges recorded for any market. The small number of direct swaps is not

surprising given how difficult it is to match the requirements of both taxpayers and

complete a direct swap of properties.

For the purposes of the regression analysis that follows in a later chapter, I exclude

all direct exchanges to eliminate the possible bias that their inclusion may induce in the

model. "Other" exchanges, which mostly include exchanges whose type has not been

confirmed, can introduce similar problems as with direct exchanges; hence, these

transactions are also eliminated from the sample. I also exclude sales that are associated

with any other "special condition" that is not of interest for this study. Examples of such

special conditions are sales that are part of an auction, bankruptcy or sales that involve

building contamination, natural disaster damage, or the threat of contamination. In total,

CoStar delineates more than 30 such unusual conditions, which potentially have an effect

on observed transaction prices. I therefore eliminate all observations that contain a sale

condition which is not analyzed by this study. The only special conditions that I allow are

related to condominium conversions and portfolio sales. Since, as noted previously, the

number of sales in the top 20 apartment markets accounts for 86 percent of all apartment









sales, I focus in the empirical analysis on property sales in these markets only. In

addition, I exclude markets for which there is not a sufficient number of exchanges to

generate meaningful statistical tests. These excluded markets are Miami, Ft. Lauderdale,

Washington, DC, Dallas, and Colorado Springs. This leaves us with 15 markets, which

represent 75 percent of all apartment sales. The final sample has 23,640 apartment

transactions over the 1999-2005 time period.

Office Properties

Table 5 presents information on office property sales by markets and shows that in

this sample the largest office market is Los Angeles with 8.5 percent of all sales. Phoenix

is the second largest market with 7.2 percent of all transactions, while Washington, DC is

the third largest market with 6.8 percent of all office transactions. The smallest market

(San Antonio) contains only 22 sales observations.

The table reveals a different picture than with apartments. First, there is less

concentration of transactions in the largest markets. Consequently, the top 20 markets

account for 75 percent of all sales. Second, the distribution of exchanges across markets

is also quite different. Although the use of exchanges varies substantially across major

metropolitan areas, in none of the markets do exchanges represent the dominant form of

property transaction.2 The observation that markets located in the Western U.S. have the

highest number of exchanges remains unchanged. The distribution of relinquished and

replacement types of exchanges also varies across the sample. In approximately one half

of the markets, replacement exchanges are either equal to or as much as twice the number

of relinquished exchanges. In some markets, sharp contrasts are observed. For example,


2 Salt Lake City is the only market that has a share of exchanges that is above 50 percent, but this market
can be ignored since it only has 26 transactions. Therefore, percentages may be biased due to the small
sample.











Table 5. Description of Office Property Sales by Markets
Name Office % of Office % Reliq. Repl. Relq. & Direct Other
All Total Exch. Exch. Exch. Exch. Repl. Exch. Exch.
Exch.


Los Angeles
Phoenix
Washington D.C.
Chicago
Atlanta
Seattle
Denver
San Diego
New York City
Tampa
New Jersey
Detroit/Toledo
Miami
Philadelphia
Ft. Lauderdale
Orlando
Dallas/Fort Worth
Boston
Oakland
Baltimore
Colorado Springs
West Palm Beach
Portland
Sacramento
Las Vegas
Tucson
Riverside/San
Bernardino
Houston
Cleveland/Akron
San Jose
Cincinnati/Dayton
Austin
North SF Bay Area
Jacksonville
San Francisco
Fresno
Columbus
Stockton/Modesto
Minneapolis
Ventura
St Louis
Kansas City
Charlotte
Reno
Salt Lake City
San Antonio


2001 8.51% 446 22.29%
1683 7.16% 126 7.49%
1600 6.81% 120 7.50%
1163 4.95% 116 9.97%
1073 4.57% 36 3.36%
1039 4.42% 340 32.72%
902 3.84% 241 26.72%
858 3.65% 282 32.87%
835 3.55% 42 5.03%
714 3.04% 49 6.86%
698 2.97% 34 4.87%
639 2.72% 46 7.20%
628 2.67% 28 4.46%
625 2.66% 30 4.80%
580 2.47% 41 7.07%
571 2.43% 26 4.55%
540 2.30% 68 12.59%
506 2.15% 19 3.75%
495 2.11% 167 33.74%
430 1.83% 25 5.81%
416 1.77% 38 9.13%
402 1.71% 33 8.21%
395 1.68% 159 40.25%
394 1.68% 159 40.36%
392 1.67% 117 29.85%
379 1.61% 60 15.83%
373 1.59% 142 38.07%

344 1.46% 39 11.34%
299 1.27% 20 6.69%
277 1.18% 65 23.47%
275 1.17% 9 3.27%
266 1.13% 19 7.14%
261 1.11% 102 39.08%
249 1.06% 7 2.81%
212 0.90% 69 32.55%
200 0.85% 59 29.50%
200 0.85% 63 31.50%
136 0.58% 31 22.79%
104 0.440o 16 15.38%
94 0.40% 18 19.15%
65 0.28% 7 10.77%
61 0.26% 11 18.03%
50 0.21% 8 16.00%
32 0.14% 9 28.13%
26 0.11% 14 53.85%
22 0.09% 4 18.18%


33
9
6
4
0
28
34
43
2
3
2
0
5
1
2
1
4
0
34
0
4
0
12
38
15
8
22









replacement exchanges in Phoenix outnumber relinquished exchanges by a factor of 7. In

Las Vegas, replacement exchanges are 5 times more frequent than relinquished property

sales and in Sacramento this ratio is equal to 3.4. A possible explanation for such

differences is the combination of several properties used to complete a replacement

exchange, as well as replacement property sales being completed outside of the particular

market or with a different property type. Table 5 reveals that direct exchanges are also

rare for office properties. The largest number of swaps is observed in Chicago, where

only 11 transactions represent direct exchanges.

As with apartments, I exclude all direct and "other" types of exchanges from the

sample in the regression analysis. I also exclude transactions associated with special

conditions that are not sale-leasebacks or portfolio sales. As noted previously, the office

sales in the top 20 office markets represent 75 percent of all office transactions. However,

after elimination from the sample of sales with the characteristics described above, eight

markets out of the largest 20 markets do not have a sufficient number of exchanges to

generate statistically meaningful tests. These markets are: Atlanta, New York City,

Northern New Jersey, Detroit, Miami, Philadelphia, Orlando, Boston and Baltimore. I

include in the sample the next largest markets that have a sufficient number of exchanges:

Sacramento, Las Vegas, Tucson, and Riverside/San Bernardino. Therefore, in the

empirical analysis I focus on the office property sales in the largest 15 markets that also

have a sufficient number of replacement and relinquished exchanges to generate any

results that are statistically meaningful. The sales in these 15 markets represent 56

percent of the office real estate market. The final sample has 8,871 office transactions

during the 1999-2005 time period.









Retail Properties

Table 6 presents a breakdown of retail property sales by markets. New York City

with 3,669 retail sales (11 percent of all sales) is the largest market, followed by Los

Angeles with 3,235 retail transactions (9.7 percent of all sales), and Chicago with 2,796

observations (8.4 percent of all sales). The smallest retail market is Reno, which only has

9 sales recorded. Similarly to the office sales, there is less concentration of transactions in

the largest markets. The top 20 markets account for 77 percent of all sales.

The use of exchanges in retail transactions also varies substantially across markets.

Similarly to the office transactions, tax deferred exchanges are not as popular as with

apartment sales. In none of the 46 CoStar markets do exchanges outnumber non-

exchanges. The largest percentage of exchanges is observed in Portland, where this type

of transaction represents more than 40 percent of all sales.3 Once again exchanges are

more frequently observed in the Western United States.

There is substantial variation in the distribution of relinquished and replacement

types of exchanges across the sample. In 10 of the retail markets relinquished exchanges

outnumber the replacement exchanges. In more than 50 percent of the sample the

replacement exchanges are less than twice as frequent as relinquished exchanges. Finally,

the replacement exchanges in 10 markets are more than 3 times higher than the

relinquished exchanges4. Notable markets, in which replacement exchanges significantly

outnumber relinquished exchanges include Tucson, with a ratio of replacement to

relinquished property sales equal to 5.6. Las Vegas has a ratio of 5.3, Riverside 4.9,

Dallas /Forth Worth 4.2, and Houston 3.8. Table 6 illustrates also the rareness of


3 In Reno 44 percent of all sales are exchanges; however this percentage is not reliable since it is based on
only 9 observations.
4 Reno and San Antonio are not included in this analysis, based on their small number of observations.












Table 6. Description of Retail Property Sales by Markets
Name Retail % of Retail % Reliq. Repl. Relq. & Direct Other
All Total Exch. Exch. Exch. Exch. Repl. Exch. Exch.
Exch.


New York City
Los Angeles
Chicago
Seattle
Phoenix
Atlanta
Detroit/Toledo
Denver
Tampa
Dallas/Fort Worth
San Diego
Boston
Washington, DC
Miami
Orlando
Oakland
Ft. Lauderdale
Houston
Tucson
Riverside/San
Bernardino
Las Vegas
West Palm Beach
Portland
Philadelphia
San Francisco
Cleveland/Akron
New Jersey
Cincinnati/Dayton
Colorado Springs
Baltimore
North SF Bay Area
Sacramento
Fresno
Jacksonville
San Jose
Columbus
Austin
Stockton/Modesto
Minneapolis
Ventura
St Louis
Kansas City
Charlotte
Salt Lake City
San Antonio
Reno


3669 11.03%
3235 9.73%
2796 8.41%
1546 4.65%
1443 4.34%
1323 3.98%
1068 3.21%
1035 3.11%
1017 3.06%
966 2.90%
918 2.76%
853 2.57%
841 2.53%
819 2.46%
799 2.40%
739 2.22%
733 2.20%
661 1.99%
584 1.76%
578 1.74%

561 1.69%
551 1.66%
535 1.61%
520 1.56%
492 1.48%
474 1.43%
461 1.39%
454 1.37%
444 1.34%
413 1.24%
356 1.07%
308 0.93%
291 0.88%
285 0.86%
274 0.82%
256 0.77%
246 0.74%
180 0.54%
146 0.44%
95 0.29%
95 0.29%
64 0.19%
52 0.16%
49 0.15%
20 0.06%
9 0.03%


60 1.64% 8
775 23.96% 247
381 13.63% 138
470 30.40% 127
145 10.05% 22
90 6.80% 11
61 5.71% 24
299 28.89% 82
43 4.23% 4
165 17.08% 19
317 34.53% 82
50 5.86% 14
43 5.11% 16
41 5.01% 10
35 4.38% 6
192 25.98% 63
47 6.41% 14
121 18.31% 21
106 18.15% 14
207 35.81% 27

199 35.47% 27
43 7.80% 12
220 41.12% 67
34 6.54% 4
175 35.57% 81
27 5.70% 6
14 3.04% 3
34 7.49% 4
39 8.78% 13
22 5.33% 11
118 33.15% 38
117 37.99% 26
40 13.75% 10
14 4.91% 3
80 29.20% 30
91 35.55% 23
28 11.38% 12
35 19.44% 10
22 15.07% 9
33 34.74% 8
11 11.58% 3
6 9.38% 1
10 19.23% 7
12 24.49% 6
7 35.00% 0
4 44.44% 0









direct exchanges in retail sales. Once again, the largest number of direct swaps is

observed in Chicago, where a total of 29 such transactions occurred.

As with apartments and office properties, I exclude all direct exchanges and "other"

types of exchanges from the regression sample. I also eliminate all transactions

associated with other special conditions that are not condominium conversions, sale-

leasebacks, or portfolio sales. The number of sales in the top 20 retail markets accounts

for 77 percent of all retail sales. However, after elimination of the above observations,

seven of the largest 20 markets do not have a sufficient number of exchanges to generate

statistical meaningful tests. These markets include New York City, Atlanta, Tampa,

Boston, Washington, DC, Miami and Orlando. Therefore, in order to have a sample of 15

markets I include the next largest markets that have also a sufficient number of

exchanges. These are Las Vegas and San Francisco. Hence, in the empirical analysis I

focus on the property sales in the largest 15 markets that also have a sufficient number of

replacement and relinquished exchanges to generate any results that are statistically

meaningful. The sales in these 15 markets represent 52 percent of the retail real estate

market. This yields a final sample of 12,015 office transactions over the 1999-2005 time

period.

Research Methodology

I use standard hedonic regression techniques to assess the influence of tax-deferred

exchanges, as well as other sale conditions on observed transaction prices. Standard

hedonic models include the log of the transaction price or rent as the dependent variable

and a set of independent variables that capture the site, structural, and location

characteristics of the property. For example, Frew and Jud (2003) regress the observed

sale price on a number of independent variables, including the square footage of rental









space, land area, age, and number of units. In an apartment rent prediction model based

on 4,500 apartment complexes in eight markets, Valente et al. (2005) use the log of

asking rent as the dependent variable. For regressors they use square footage per unit,

number of floors, property age, submarket dummies, and year of sale dummies.

Building age, age squared, the square footage of improvements, building foot

print, lot size, number of units, and number of floors, are some of the most common

structural characteristics used in commercial property price or rent equations (see, for

example, Colwell, Munneke and Trefzger, 1998, and Saderion, Smith and Smith, 1993).

The choice of functional form is also very important in order to ensure that the model is

correctly specified. Weirick and Ingram (1990) provide an excellent analysis of various

approaches to functional forms in hedonic regressions, when the dependent variable is

selling price. In particular, the authors compare three standard functional forms:

* A linear model
* A semi-log model which uses the logarithmic transform of the dependent variable
(selling price)
* A log-linear model, which uses logarithmic transforms of both the dependent
variable as well as independent variables

As Weirick and Ingram (Ibid.) point out, the linear form has "serious deficiencies

from a market theory standpoint." Such models force the value of an extra square foot of

improvement for a 2,000 sq. ft. property to be the same as the value of an extra square

foot for a 10,000 sq. ft. property. The semi-log and log-linear models take into account

nonlinearities in the data. In addition, by using quadratic transformations of explanatory

variables (such as square footage and lot size) I can capture property value relationships

that are concave or convex in certain characteristics (Ibid.).









Recently, the focus of the residential hedonic pricing literature has shifted to the

proper control of location. One of the most important papers in this area is Clapp (2003),

who presents a semi-parametric method for valuing residential location and includes

latitude and longitude as explanatory variables. Case, Clapp, Dubin and Rodriguez (2004)

use a second order longitude-latitude expansion to control for location, as well as a

number of demographic characteristics, such as percent Black, Hispanic, etc.

Finally, by using geographic coordinates, Fik, Ling and Mulligan (2003) fully

account not only for the absolute location of the home, but also for relative location in a

metropolitan market. The authors use a complete variable interactive approach to model

the log of sale price as a function of structural characteristics, discrete location dummies,

location dummies and { x, y coordinates interacted with structural characteristics,

interaction terms between structural characteristics (e.g. age*sq. ft.), interacted structural

characteristics interacted with location dummies and triple interaction terms of location

dummies, geographic coordinates and structural characteristics. This fully interactive

specification allows Fik et al. (2003) to effectively estimate separate price surfaces for

identified sub-markets, rather than constrain the estimated coefficients on structural

characteristics to be constant across submarkets with the price surface shifted up or down

by location dummies only. In addition, interacting sub-market dummies with absolute

location allows Fik et al. (2003) to capture discontinuities or structural shifts that occur as

the price surface crosses submarket boundaries.

Table 7 provides definitions for all variables used in the hedonic regressions. The

dependent variable is LNPRICE the natural logarithm of the sale price. An advantage of

using the log of price is that less weight is given to extreme values than when using










untransformed prices. I also divide square footage of improvement and land square

footage by 1000 to keep size ranges consistent.

Following Weirick and Ingram (1990), I use a semi-log model with quadratic

transforms for square footage in thousands (SQFT), and land square footage in thousands

(LANDSQFT). With this semi-log form, unit price per unit change in the characteristic is

given by simply multiplying the estimated coefficient by the observed selling price.

To quantify the effect of Section 1031 exchanges, condominium conversions, sale-

leasebacks, portfolio sales, and out-of-state buyers on sale prices, I use a stepwise

estimation technique and estimate the following model separately for each of the

identified 15 markets in the apartment sales sample


LNPRIC, = a, +a1EXREPL+ a2EXRELQ+ aRELQ_ REPL+ aA GE+ aAGE2 + aSQFT
+ a7SQFT2 + a8LANDSQFT+ a9LANDSQF72 + PARKING+ a, FLOORS+ a, UNITS

+ f, CONDITION, + a13BUYEROUT+ a F .0JR+ a5 SUBSIDIZED+ a1l6CONDO
2005 P<43
+aC17CONDOCONV+a ,PORTSALE+ z,zYRn,+ 7, \\i)(U \ (17)
n=2000 s=2

I use a similar model to estimate the effect of tax-deferred exchanges,

condominium conversions, sale-leasebacks, portfolio sales, and out-of-state buyers on

property sale prices in offices and retail transactions

LNPRICE = ao +aEXREPL+ aEXRELQ+a3RELQ_ REPL+a4AGE+a AGE2 + 6SQFT

+ aSQFT2 + a LANDSQFT+ a LANDSQF72 + a,0PARKING+ a FLOORS+ E P, CONDITION
2005 P<43
+al2BUYEROUT+al3SALELEASEBICK+a14PORTSALE+ XnYR, + YS \\l)( u (18)
n=2000 s=2

In the estimation of the hedonic pricing equation, I include a dummy variable

(EXRELQ) that quantifies the extent to which transaction prices are lower (higher), all

else equal, if the seller of the property is a taxpayer initiating the "downleg" portion of a









delayed exchange (i.e., if the property is a "relinquished" property). I also include a

dummy variable (EXREPL) that quantifies the extent to which transaction prices are

higher, all else equal, if the buyer of the property is a taxpayer completing the "upleg"

portion of a delayed exchange (i.e., if the sample property is a "replacement" property).

Finally, I include a dummy variable (RELQ REPL) that quantifies the extent to which

prices are higher, all else equal, if the same property is used as both the relinquished

property for the seller and the replacement property for the buyer. The first exchange

involves the front end of a 1031 transaction for the seller, who now needs to find a

replacement property in order to complete the exchange. The second exchange involves

the back end of an exchange in which the taxpayer has already sold his property and uses

this sale to acquire a replacement property thereby completing his own exchange.

Lambson, McQueen and Slade (2004) find evidence that out-of state buyers pay

price premiums for apartment complexes in Phoenix, which the authors associate with

possible higher search costs and anchoring. I therefore use a dummy variable

BUYEROUTto control for any effects related to whether the buyer's principal residence

is out-of-state.

Two dummy variables (CONDOCONV and PORTSALE) are also used in the

apartment regressions to quantify the extent to which prices are higher, all else equal,

when the apartments are purchased with the intention to convert them to condos, or when

the transaction is part of a portfolio sale.

In the office and retail regressions, dummy variables (SALELEASEBACK and

PORTSALE) are used to quantify the price premium associated with sale-leasebacks or

portfolio sales.









Table 7. List of Regression Variables
Dependent Variable:
LNPRICE Natural logarithm of the sale price
Exchange Variables:
EXRELQ Binary variable set equal to one if transaction represents sale of a
relinquished property
EXREPL Binary variable set equal to one if transaction represents purchase of a
replacement property
RELQ REPL Binary variable set equal to one if transaction represents both sale of a
relinquished property and purchase of a replacement property

Building Characteristics:
AGE Age of the buildings) in years
SQFT Total improvements square footage in thousands
FLOORS Number of floors
UNITS Number of units
CONDITION, Physical condition of the property based on inspection. The categories
include below average, average, and above average. The omitted category
is average.
PARKING Number of parking spaces
SUBSIDIZED Binary variable set equal to one if property use is subsidized multi-family
SENIOR Binary variable set equal to one if property use is senior multi-family
CONDO Binary variable set equal to one if property use is multi-family
condominium


Year Dummies:
YR,


Yearly time periods from 1999 through 2005. Each year is included as a
binary variable except 1999, which is suppressed.


Site and Location Characteristics:
LANDSQFT Log of square footage of land in thousands
X Latitude of property
Y Longitude of property
SDUM, Binary variable signifying the submarket in which the property is located,
as defined by CoStar
{ l, Y} X, Y polynomial, where k,l >0, k+l<= 3

Deal Characteristics:
BUYEROUT Binary variable set equal to one if buyer lives out of state
CONDOCONV Binary variable set equal to one if the transaction was motivated by condo-
conversion
SALELEASEBACK Binary variable set equal to one if transaction was part of sale-leaseback
PORTSALE Binary variable set equal to one if transaction was part of portfolio sale









I use several variables to account for the relationship between selling price and

property structural characteristics. First, I expect a negative relation between age (AGE)

and price and a positive coefficient on age squared (AGE2). This expectation reflects the

frequently observed quadratic relation between price and age. A "vintage" effect is

sometimes observed, which is related to high prices for very old properties. I expect the

coefficients on SQFT, LANDSQFT, PARKING, FLOORS and UNITS to be positive.

The variable CONDITION, controls for building condition. I specify average

condition as the control group. With residential real estate, 79 percent of apartments are

reported to be in average condition, 14 percent of the apartment properties are

categorized by CoStar as being in above average condition, and 7 percent are labeled as

below average. In the office sample, 66 percent of the properties are classified to be in

average condition, 32 percent are above average, and only 3 percent are in below average

condition. Finally, with retail properties 70 percent are in average condition, 22 percent

are above average, and 8 percent are in below average condition.

I control for the effects of time by including dummies for each year in the sample

with 1999 as the base year for comparison. In the apartment regressions, I also determine

whether the use of the apartments is primarily as senior housing, subsidized housing, or

multifamily condominiums. The comparison group, which represents 98 percent of the

sample, is all other multifamily apartments.

Finally, I include dummy variables to control for differences across submarkets

within each major market. These submarkets are defined by CoStar. There are 405 unique

submarkets in the apartment sample. In the largest market, Los Angeles, 42 submarkets

are identified by CoStar. In the second largest market, New York, 30 different









submarkets are defined. In the smallest of the 15 apartment markets, Sacramento, there

are 21 submarkets.

In the office sample there are 488 distinct submarkets and that these submarkets

largely overlap with the submarkets defined in the apartment sample. There are 51

submarkets present in the Los Angeles office sample, 35 submarkets distinguished in the

Phoenix area and only 6 submarkets identified in Tucson, the smallest of the 15 studied

markets.

The retail sample contains 491 different submarkets. Los Angeles is once again the

largest market in this sample with 60 submarkets, Chicago, the second largest market in

the sample, has 37 submarkets, and San Francisco, which is the smallest market in the

sample with only 339 observations, has 33 submarkets defined.

The estimated models differ for residential and non-residential (office and retail)

properties. For each property type I run the specified model by market. In order to avoid

the effect of outliers in the data, I winsorize all continuous dependent variables in the

regressions at the top and bottom one percent of the distribution. The winsorising

procedure takes the non-missing values of a continuous variable sorted in ascending order

and replaces its one percent highest and lowest values by the next value counting inwards

from the extremes. The only exceptions are FLOORS, PARKING and UNITS, where I

winsorize at the top and bottom 0.5 percent of the distribution, to account for the narrow

distribution. Longitude and latitude coordinates are not winsorized.

The next chapter presents summary statistics of apartment transaction data by

market and the regression results from the models specified.














CHAPTER 7
RESULTS FOR RESIDENTIAL REAL ESTATE

This chapter focuses on the results from the empirical analysis on the apartment

market. An additional benefit of the analysis on apartment markets is the relative

simplicity and homogeneity of apartment leases, which simplifies modeling relative to

office and retail properties.

Summary statistics for the variables of interest are presented in Table 8. The first

two columns present a summary of the data at the aggregate level, while the remaining

columns present statistics for each of the 15 markets studied. The average apartment

complex in the sample is 49 years old, contains 23,034 square feet of improvements, is

built on 40,162 square feet of land area, has 27 units, 2.5 floors, 29 parking spaces and

sold for $2,194,040.

With an average age of 79 years properties in New York and Boston tend to be

much older than in other markets across the country. Phoenix has the newest apartments

with an average age of 27 years. The apartment buildings in the sample tend to be the

largest in Phoenix with an average size of 76,688 sq. ft. Oakland is the market with the

smallest average size of apartment buildings, both in terms of square footage (12,267)

and number of units (15).

Approximately 13 percent of the transactions involve the purchase of a replacement

property to finalize an exchange; 12 percent involve the sale of the relinquished property

in an exchange; and eight percent involve the sale of relinquished property that is also the

replacement property of the buyer in a separate exchange.


























-






o00










OC t-



S2
gg I
oc


0'o t-c


00C]3 t-
M r~ i^"


o
d
0















C -




0


0i
0



Cn
Co

0n


0


C] C] 0 ] C],,
C] itd-


S C,
'tm m o 0'!






oc
'I ^ 2

00 oo^
c^^-g ^S


ICCI ^^


d


[C

i C


0000
0000
(Ar "rtrs
z cz cz c


C


0000 o rnOOOOOOOOo tno













OrnOMOO 0000c
0 O CA -t00 0







S- -00000000000000000000



































000 C 00000000C0>0 C
m m C I-- oc C o -
00000000000000000000
ooo~ooooo~ooooot~im^^o
ddddddddddddddddddd


C>\ n C> '

C] oO


S o
M


IIu


crjcrj





















0000


CC





Eo











oc











-soc


'Clg~l~Cl
CC C. -^
- C CC t-





CCi
0000o
CC 9
z z z





\C D
CC Cl 3 -


Zi O CC
0 z









Cli
^0- CC CC -









^- CC
"100^


3^


rS

IC1\\D ^
m ( ^
"r ..^
rCn o (^ <
^r3 n


s s





S<1 at
CO


CZ CZ CZ CZ CZ CZ
000000
c1 1 1 1 1 c c c


CC
(^030
m zo
C Cl
- C- Z


\C C C l C Zt m C 0Z Z -3 C C C C \
Cl\CZOCZ~0ZZZZZ3l3lZ
z z z z z zz zz zz z z z












t- l ~ l ~ C C Z Z Cl C ~ C \
---CCZO C-Z0ZZZZ--Cl~l-
z z z z z zz zz zz z z z












Cl~~OCCC~ 3 ~ ZCCm
z z z z z zz zz zz z z z






0 C -OC\C C 0\C -CC- C 0 \C\C C l






ZZZZZOC-~ZZZZZ----Cl-
z z z z z zz zz zz z z z






0 C \COC CC- 0Z Z Z Z Z Z Z C 0 l






--- ~ o~ z ~ zzz z----------------
z z z z z zz zz zz z z z


*a

0









0


0
0




0
a,
o

*a
O


0-
1
1

0



o
o~


~a ss
























h 00






~C\D*~O*l








Cfi C'i Ct C






000





C0 Cl 0
Cl0~*Ci
cm *~ thC


COO
NOO ^ oo
\00200 00





. O .










0 t- 0 C 0
0


\D\D tC-\ i l




t-~- 0
-\D t-oc 0







C< 0" r

r Cl CD C1


C


0


00
u
1














0

0
so
(S

r^
u,
'S
O


VI
tI
a
0,


o
1^



a
^


SQ q
N N3


Y

00000
000000
SCZ, CZ, CZ, CZ, Z
(" I ("I
? SS S2S


1,00






Cl o
v0 0


Cl
8 -0


OOmN ltO- t-O-lOOOOO N N -
-00c mCl e0 HN-0 0 0- Cl
00000000000000000000














ddd-0dd0dddd- dd000 0 -- (N
~~~m00 C lC3~ ~ ~-3
Cl-~0~~300C-000003-Cl














Cl 00000000000000000000
-hN- O OOOOOOOOm'OO--













00000000000000000000














-i OOOm O OO -0
~30030000000003300000






N'- 00000 OOOOO - 0000 000 0
d000000000000 00--00


0



0





2
IA m
6D




v,

Q
`r

VIB


oCl


O~S









Approximately seven percent of all buyers reside out-of-state. There are 1,749

transactions in which the buyer was out of state. Phoenix is the market with the highest

percentage of out-of-state buyers. They represent 59 percent of the 1,291 observations.

Other markets of potential interest when quantifying the effect of having an out-of-state

buyer include: Tucson with 43 percent out-of-state buyers, Portland with 23 percent,

Denver with 13 percent, Seattle 10 percent, and New York 4 percent.

Approximately two percent of all sales were part of a portfolio sale; there are 488

portfolio sales in the sample. Denver is the market with the highest percentage of

portfolio sales; they represent 6 percent of all sales. The other markets with a large

enough number of portfolio sales to generate statistical significance are: New York City,

4 percent of all sales; San Francisco, 4 percent of sales; Phoenix, 3 percent of all

observations; and Chicago, 2 percent of sales. Portfolio sales and out-of-state buyers in

Los Angeles are only 1 percent each, but there are 7,761 observations in this market.

Slightly over 200 properties, or one percent of all sales were were purchased by

condo converters. There are only two markets in which I could look for any price effects

from condo-conversion motivation: San Diego with 77 condo conversions, and Boston

with 34 such transactions.

Table 9 performs t-tests for differences between the mean price of apartments in a

control group that contains all properties not associated with any conditions and the

groups of properties that are the subject of study. These are properties that represent a

replacement exchange only; properties that are associated with a relinquished exchange

only; properties that served as both the replacement and relinquished property; purchases

by out-of-state buyers only; purchases by condo converters; and portfolio sales not









associated with any conditions. For example, the replacement exchanges group contains

all properties in which EXREPL is equal to 1, there are no conditions of sale

(CONDOCONV and PORTSALE are both equal to zero) and the buyer is not from out-of-

state.

Table 9. Differences in Mean Prices of Control Sample and Identified Groups of Interest
for Apartment Properties

S. Mean Value Standard
Apartments Observations Mean Value Standard T-test Value
of Sales Price Error

Control Group 14,383 1,703,610 30,440

EXREPL 2,786 1,878,481 59,843 -2.36*

EXRELQ 2,609 1,749,449 57,314 -0.61

RELQ REPL 1,687 1,707,154 48,233 -0.04

BUYEROUT 1,288 7,004,433 295,063 -39.36*

CONDOCONV 133 7,545,770 989,659 -17.68*

PORTSALE 310 4,650,359 352,059 -13.79*


The results show that replacement apartment exchanges are associated, on average, with a

10.3% price premium and the price difference is statistically significant. Relinquished

exchanges and transactions associated with both replacement and relinquished exchanges

have average prices that are not statistically significantly different from the average price

in the control group.

Sales that are completed by out-of-state buyers are associated with a price which

is four times higher than the average price of properties in the comparison group. The

results are similar with respect to sales to condo converters, which are associated with an

even higher premium, which is also statistically significant. Finally, I also observe that

portfolio sales are, on average, 173% more expensive than transactions in the comparison









sample. Therefore, I reject the null hypothesis that the difference between the mean price

in the comparison group and the mean price observed with replacement exchanges, out-

of-state buyers, condo conversions and portfolio sales correspondingly are equal to zero. I

was not able to reject the null hypothesis for sales involving relinquished exchanges and

two different exchanges. This result is consistent with our expectations.

Next, I present the results from estimating equation (17) for each of the 15 markets in

Table 10.

I perform each of the regressions using a stepwise estimation procedure to decide

which of the submarket dummies to leave in the final model. All other dependent

variables are not subject to the procedure. I use a robust estimation method to account for

potential heteroskedasticity; therefore all reported p-values are adjusted values.

The reported results in Table 10 show that the estimated coefficients on the

structural attributes are of the predicted sign and statistically significant in most of the

market regressions. PARKING tends to be positive and insignificant. However, in two

markets, San Diego and New York City, it is negative and significant. It also has a

negative sign in the Chicago and Los Angeles regressions, with p-values of 0.14 and 0.16

correspondingly.

This finding is not surprising and is specific to higher density areas in which more

parking is usually associated with apartment communities that are distant from the

centers of the city and hence tend to be cheaper. The coefficient on UNITS is positive in

all regressions but one, and is significant in 6 of the 15 models. In Phoenix, UNITS

reverses sign and becomes negative and significant at less than one percent level. Phoenix

is the market in which the largest average transactions are observed. The average









apartment property in the sample has 86 units and 76,688 sq. ft. of improvements. In

contrast, the average apartment property sold in the 15 markets has 27 units and 23,034

sq. ft. of improvements.

The estimated coefficient on the variable of interest, EXREPL, is positive and

significant in 12 of the 15 regressions. This provides evidence that buyers of replacement

properties are paying statistically significant price premiums in the majority of the larger

markets contained in the regression sample.

The coefficient estimates ofEXRELQ are generally positive and significant in 7 out

of the 15 regressions. However, in only three of the regressions are the coefficient

estimates on relinquished exchanges higher than the coefficient estimates of replacement

exchanges. Casual observation shows that relinquished exchanges have significant and

positive coefficients in markets that have seen more appreciation in residential real estate

than others. In addition, it is important to remember that if the main motivation to enter

into an exchange is to postpone capital gain taxes, properties that are part of relinquished

exchanges will tend to be ones that have seen more capital appreciation than other

markets. Hence, all else the same, they will tend to be more expensive properties.

Therefore, the observed positive coefficients provide evidence that sellers enter in tax

delayed exchanges when their properties have appreciated in value. This positive effect

on price may offset any other possible negative effects related with relinquished

exchanges and thus I can not make any conclusions about the magnitude of the price

impact related to a sale being part of a relinquished exchange.

With replacement exchanges, there are no such issues; and therefore all else equal

the coefficient on EXREPL can be directly associated with a price premium paid when











Table 10. Regression Statistics for OLS Model with Structural Characteristics and Submarket
Dummies by Apartment Markets


Market


Oakland Phoenix Portland


Observations
EXREPL


EYRELQ


RELQ_ REPL


AGE


AGE2


SQFT


SQFT2


LANDSQFT


LANDSQFT2


PARKING


FLOORS


UNITS


CONDITION BA


CONDITION AA


BUYEROUT


SENIOR


SUBSIDIZED


New
Los
Boston Chicago Denver Angeles York
Angeles C

400 1585 782 7761 4067
0.139 0.107 0.160 0.076 0.094
0.10 0.01 0.00 0.00 0.43
-0.145 0.137 0.140 0.043 0.081
0.23 0.00 0.00 0.00 0.15
0.756 0.260 0.192 0.093 -0.005
0.00 0.00 0.00 0.00 0.98
-0.008 -0.005 -0.007 -0.009 -0.003
0.12 0.05 0.04 0.00 0.10
0.000 0.000 0.000 0.000 0.000
0.71 0.10 0.21 0.00 0.03
0.021 0.030 0.029 0.039 0.028
0.00 0.00 0.00 0.00 0.00
0.000 0.000 0.000 0.000 0.000
0.00 0.00 0.00 0.00 0.00
0.003 -0.001 0.000 0.000 0.010
0.03 0.62 0.63 0.87 0.00
0.000 0.000 0.000 0.000 0.000
0.02 0.18 0.31 0.20 0.00
0.001 -0.001 0.000 -0.001 -0.002
0.48 0.14 0.22 0.16 0.05
-0.036 0.024 0.041 0.044 0.081
0.26 0.23 0.08 0.02 0.00
0.008 0.002 0.002 0.006 0.003
0.01 0.17 0.02 0.00 0.07
-0.101 -0.202 -0.194 -0.089 -0.008
0.22 0.00 0.00 0.00 0.69
-0.057 0.167 -0.018 0.116 0.094
0.51 0.00 0.73 0.00 0.00
0.051 0.065 0.038 0.016 0.040
0.66 0.65 0.47 0.77 0.30
-0.003 0.024 0.607 -0.419 -0.634
0.99 0.93 0.00 0.13 0.00
-0.536 0.174 0.060 -0.069 -0.364
0.20 0.25 0.60 0.67 0.08


1176
0.071
0.03
0.050
0.07
0.113
0.00
0.000
0.96
0.000
0.14
0.043
0.00
0.000
0.00
0.000
0.89
0.000
0.35
0.000
0.80
0.118
0.00
0.001
0.79
-0.094
0.01
0.099
0.20
-0.206
0.22
0.257
0.05
-1.110
0.25


1291
-0.020
0.67
0.000
1.00
-0.108
0.18
-0.010
0.00
0.000
0.72
0.023
0.00
0.000
0.00
0.000
0.37
0.000
0.95
0.000
0.82
0.035
0.14
-0.001
0.00
-0.232
0.00
0.064
0.20
0.109
0.00
0.394
0.17
-0.015
0.84


490
0.086
0.06
0.087
0.13
0.113
0.01
-0.008
0.02
0.000
0.03
0.027
0.00
0.000
0.00
0.002
0.12
0.000
0.12
0.000
0.99
0.085
0.06
0.002
0.33
-0.035
0.58
0.174
0.01
-0.047
0.32
0.310
0.36
0.277
0.00











Table 10. Continued


New
Boston Chicago Denver York Oakland Phoeni
Angeles City

0.154 0.186 0.307 0.171 -0.484 -0.007 0.103


0.09


0.00


0.00


-0.038 0.098 -0.038 0.044


0.21


0.79


0.509 0.158 -0.104 -0.012
0.01 0.02 0.16 0.82


0.90


0.093 0.516 0.046
0.03 0.04 0.47


0.072


-0.042
0.48


-0.089 0.145 0.067 0.032 -0.038 0.134 0.136


0.58 0.01


0.34


0.00


-0.005 0.270 0.338 0.142 0.040 0.457 0.179


0.00


0.00


0.60


0.00


0.112 0.364 0.261 0.295 0.139 0.627 0.262


0.00


0.00


0.07


0.00


0.332 0.541 0.279 0.523 0.373 0.657 0.279


0.00


0.00


0.00


0.00


0.394 0.622 0.220 0.753 0.507 0.802 0.443


0.00


0.00


0.00


0.00


0.559 0.691 0.381 0.911 0.641 0.858 0.539


PORTSALE


YR2000


YR2001


YR2002


YR2003


YR2004


YR2005

SD UMi
(not
reported)
CONST


R-squared


0.00


0.00


0.00


x Portland


Market


CONDO


14.276 13.295 12.631 13.132 12.079 12.517 13.372 12.858


0.00 0.00
0.88 0.80


0.00


0.83


0.00
0.83


0.00


0.93


All standard errors are adjusted for potential heteroskedasticity. P-values are reported under the coefficient estimates and are in
bold and italics. Dependent Variable is LNPRICE Log of selling price; EXREPL Binary variable set equal to one if
transaction represents sale of a replacement property. EXRELQ Binary variable set equal to one if transaction represents sale of
a relinquished property; RELQ_REPL Binary variable set equal to one if transaction represents both sale of relinquished
property and purchase of a replacement property. AGE Age of the buildings) in years; AGE2 Age squared; SQFT Square
footage of total improvements; SQFT2 Square footage of total improvements squared; LANDSQFT Square footage of land;
LANDSQFT2 Square footage of land squared; PARKING Parking, defined as number of parking spaces; FLOORS Number
of floors; UNITS Number of units; CONDITION, Physical condition of the property based on inspection. The categories
include below average, average and above average. The omitted category is average; BUYEROUT Binary variable set equal to
one if buyer lives out of state; SENIOR Binary variable set equal to one if property use is senior multi-family; SUBSIDIZED
Binary variable set equal to one if property use is subsidized multi-family; CONDO Binary variable set equal to one if property
use is multi-family condominium; CONDOCONV- Binary variable set equal to one if transaction was part of condo conversion;
PORTSALE Binary variable set equal to one if transaction was part of portfolio sale; YR, Yearly time periods from 1999
through 2005. Each year is included as a binary variable except 1999, which is suppressed; SDL \I Binary variable signifying
the submarket in which the property is located, as defined by CoStar brokers.


0.030
0.84
0.006
0.99
-0.601
0.00
-0.045
0.58
-0.091
0.24
-0.028
0.73
0.095
0.27
0.146
0.08
0.343
0.00


CONDOCONV


0.00 0.00











Table 10. Continued


San San
Diego Francisco


San Jose Seattle Tucson


Observations
EXREPL


EXRELQ


RELO REPL


AGE


AGE2


SQFT


SQFT2


LANDSQFT


LANDSQFT2


PARKING


FLOORS


UNITS


CONDITION BA


CONDITION AA


BUYEROUT


SENIOR


SUBSIDIZED


CONDO


593
0.098
0.03
0.120
0.02
0.234
0.00
-0.012
0.01
0.000
0.33
0.028
0.00
0.000
0.00
0.001
0.21
0.000
0.08
0.000
0.47
-0.001
0.98
0.000
0.81
-0.061
0.53
0.047
0.38
-0.013
0.89
0.172
0.18
0.358
0.00
-0.040
0.61


370
0.158
0.00
0.184
0.00
0.182
0.00
-0.009
0.04
0.000
0.34
0.028
0.00
0.000
0.00
0.002
0.43
0.000
0.29
-0.001
0.18
0.143
0.04
0.001
0.43
-0.153
0.11
0.048
0.47
0.057
0.74



0.002
0.98


2031
0.075
0.00
0.030
0.19
0.087
0.00
-0.002
0.25
0.000
0.81
0.030
0.00
0.000
0.00
0.001
0.25
0.000
0.36
-0.001
0.03
0.112
0.01
0.007
0.00
-0.101
0.02
0.175
0.00
0.019
0.78
0.151
0.04
-0.226
0.17
0.174
0.01


796
0.078
0.01
0.068
0.01
0.115
0.00
0.000
0.90
0.000
0.51
0.094
0.00
-0.001
0.00
-0.001
0.90
0.000
0.09
0.001
0.79
0.072
0.00
0.001
0.80
-0.186
0.00
0.180
0.00
-0.017
0.83



-0.181
0.20
0.341
0.00


471
0.101
0.00
0.004
0.87
0.082
0.05
-0.002
0.75
0.000
0.97
0.053
0.00
0.000
0.00
0.005
0.14
0.000
0.17
-0.002
0.33
0.137
0.00
0.020
0.02
-0.113
0.06
0.140
0.01
-0.019
0.79
-0.888
0.00
-1.378
0.00
4.651
0.00


1281 546
0.055 0.044
0.03 0.28
0.016 0.031
0.59 0.61
0.094 0.124
0.02 0.02
-0.002 -0.005
0.36 0.13
0.000 0.000
0.69 0.60
0.025 0.021
0.00 0.00
0.000 0.000
0.00 0.00
0.000 0.002
0.67 0.05
0.000 0.000
0.64 0.54
0.001 0.001
0.18 0.00
0.158 -0.003
0.00 0.60
0.002 0.001
0.14 0.47
-0.140 -0.163
0.00 0.00
0.107 0.055
0.00 0.45
-0.015 0.190
0.75 0.00
0.061 0.248
0.80 0.00
0.102 0.400
0.14 0.00
0.325 0.424
0.00 0.06


Market


Riverside
/San
Bernardino


Sacramento











Table 10. Continued


Market


CONDOCONV


PORTSALE


YR2000


YR2001


YR2002


YR2003


YR2004


YR2005


SDUMi (not
reported)
CONST


Riverside
/San
Bernardino
-0.338
0.12
0.037
0.70
0.106
0.19
0.258
0.00
0.343
0.00
0.472
0.00
0.726
0.00
0.998
0.00


12.528
0.00


Sacramento


-0.061
0.82
-0.798
0.00
0.147
0.07
0.221
0.01
0.474
0.00
0.620
0.00
0.855
0.00
0.877
0.00



12.704
0.00


San San
Diego Francisco

0.160
0.02
-0.248 -0.156
0.16 0.00


0.146
0.19
-0.550 0.111


0.154
0.16


0.052 0.208 0.108 -0.007 -0.128


0.27


0.00


0.211 0.321


0.00


0.00


0.445 0.376


0.00


0.00


0.00


0.89


0.347 0.022 -0.024


0.00


0.68


0.363 0.077


0.00


0.14


0.80
0.035
0.71


0.664 0.386 0.375 0.148 0.074


0.00


0.00


0.907 0.454


0.00


0.00


0.00 0.01
0.397 0.163


0.00


0.00


0.127
0.15


0.842 0.510 0.370 0.299 0.309


0.00


0.00


0.00


0.00


0.00


13.363 12.940 13.050 12.804 12.869


0.00


0.00


0.00


0.00


0.00


R-squared 0.91 0.91 0.85 0.84 0.92 0.89 0.92
All standard errors are adjusted for potential heteroskedasticity. P-values are reported under the coefficient estimates and are in
bold and italics. Dependent Variable is LNPRICE Log of selling price; EXREPL Binary variable set equal to one if
transaction represents sale of a replacement property. EXRELQ Binary variable set equal to one if transaction represents sale of
a relinquished property; RELQ_REPL Binary variable set equal to one if transaction represents both sale of relinquished
property and purchase of a replacement property. AGE Age of the buildings) in years; AGE2 Age squared; SQFT Square
footage of total improvements; SQFT2 Square footage of total improvements squared; LANDSQFT Square footage of land;
LANDSQFT2 Square footage of land squared; PARKING Parking, defined as number of parking spaces; FLOORS Number
of floors; UNITS Number of units; CONDITION, Physical condition of the property based on inspection. The categories
include below average, average and above average. The omitted category is average; BUYEROUT Binary variable set equal to
one if buyer lives out of state; SENIOR Binary variable set equal to one if property use is senior multi-family; SUBSIDIZED
Binary variable set equal to one if property use is subsidized multi-family; CONDO Binary variable set equal to one if property
use is multi-family condominium; CONDOCONV- Binary variable set equal to one if transaction was part of condo conversion;
PORTSALE Binary variable set equal to one if transaction was part of portfolio sale; YR, Yearly time periods from 1999
through 2005. Each year is included as a binary variable except 1999, which is suppressed; SDL I Binary variable signifying
the submarket in which the property is located, as defined by CoStar brokers.


San Jose Seattle Tucson









the sale is part of a replacement exchange. The coefficient on RELQ REPL also is

positive and significant in 13 out of the 15 markets. As discussed, this coefficient

represents the combined price effect of relinquished and replacement exchange

motivation. The magnitude of this coefficient tends to be larger than the coefficient on

replacement exchanges.

The coefficient on the variable indicating that the buyer is out-of-state,

BUYEROUT, is statistically and economically significant in just two markets Phoenix

and Tucson. These are also markets of specific interest when analyzing the price

premiums associated with out-of-state buyers. In Phoenix 59 percent of sales were to out-

of-state buyers, while in Tucson their share was 44 percent.

With condominium conversions there are only two markets of potential interest -

San Diego and Boston. I find that the coefficient on the variable indicating a purchase by

a condo converter, CONDOCONV, is economically and statistically significant in San

Diego. The coefficient is also significant in New York City and Oakland.

The coefficient on the variable indicating a portfolio sale, PORTSALE, is

statistically significant in 3 out of the 5 markets of interest Chicago, New York City

and San Francisco. However, contrary to our intuition the coefficient on PORTSALE in

San Francisco is negative, rather than positive.

The estimated year dummies, with 1999 as the omitted year, are generally positive

and significant. Moreover, the magnitude and significance of the time dummy

coefficients reveal substantial price appreciation over the seven year study period.

Finally, although submarket dummy variables are not reported in the regression tables,

the model fit is improved significantly by the use of the submarket location controls.










In OLS Model II, I add controls for absolute location by using a third order

expansion of the property's latitude and longitude coordinates. This expansion yields nine

additional explanatory variables.1 The resulting model specification has the following

form:


LNPRICEm = ao +a EXREPL+a2EXRELQ+a RELQ _REPL+ A 4GE+a 5AGE2 + 6SQFT
+ aiSQFT2 + acLANDSQFT+ c9LANDSQF72 + PARKING+ 1 FLOORS+ a,2UNITS

+ ZfCONDJITIOY+a 3BUYEROUT+ a4SENIOR+ axSUBSIDIZE+ al6CONDO
2005 P_43 r r
+cO7CONDOCONV+a1,PORTSALE+ X,,YR, + '6 -'/i( 'I + 2 YZ,1(Xk ,Y) (19)
n=2000 s=2 k=0l=0

Table 11 presents the regression results for Model II. Controlling for absolute

location, in addition to relative location, improves slightly (by about 0.01) the fit of the

models. Adjusted R-squared for the market regressions range from 0.81 to 0.92. The

regression results are virtually unchanged from the results reported for Model 1.

Therefore, I will only discuss results related to the key variables of interest.


The coefficient on the variable indicating a replacement exchange, EXREPL, is

positive and significant in 12 of the 15 regressions. This reconfirms the previous result

that buyers of replacement properties are paying statistically significant price premiums

in the majority of the sample markets. The coefficient on the variable indicating a

relinquished exchange, EXRELQ, remains generally positive and significant in 7 out of

the 15 regressions. The size of coefficient estimates remains largely unchanged. The

coefficient on RELQ REPL is again positive and significant in 13 out of the 15 markets

and its magnitude remains larger than the coefficient on replacement exchanges.

1 I also experiment with a model where I control for differences in the effects of key structural
characteristics across submarkets by interacting age and the square footage of improvements with the sub
market dummies. The Third Model yields a slightly better fit R-squares are increased by 1-5 percent
across markets. However, it sacrifices even more degrees of freedom. In addition, coefficients on age and
square footage, due to the presence of interaction variables, are mostly insignificant and hard to interpret.
Therefore, I do not report the results from the third model in the dissertation.











Table 11. Regression Statistics for OLS Model with Structural Characteristics, Submarket
Dummies and Longitude, Latitude Coordinates by Apartment Markets


Boston Chicago Denver


Los New York
Los New York Oakland Phoenix Portland
Angeles City


400 1585 782 7761
0.145 0.099 0.162 0.076
0.09 0.02 0.00 0.00
-0.131 0.128 0.137 0.046
0.28 0.00 0.00 0.00
0.754 0.234 0.191 0.092


0.00 0.00 0.00 0.00
AGE -0.008 -0.005 -0.006 -0.010
0.13 0.08 0.05 0.00
AGE2 0.000 0.000 0.000 0.000
0.71 0.16 0.22 0.00
SQFT 0.021 0.029 0.029 0.039
0.00 0.00 0.00 0.00
SQFT2 0.000 0.000 0.000 0.000
0.00 0.00 0.00 0.00
LANDSQFT 0.002 0.000 0.000 0.001
0.04 0.95 0.68 0.60
LANDSQFT2 0.000 0.000 0.000 0.000
0.02 0.34 0.33 0.30
PARKING 0.001 -0.001 0.000 -0.001
0.43 0.19 0.21 0.17
FLOORS -0.039 0.024 0.042 0.043
0.23 0.19 0.07 0.02
UNITS 0.009 0.002 0.002 0.006
0.01 0.19 0.02 0.00
CONDITION BA -0.093 -0.191 -0.199 -0.086
0.25 0.00 0.00 0.00
CONDITION AA -0.063 0.160 -0.015 0.114
0.47 0.00 0.78 0.00
BUYEROUT 0.056 0.072 0.035 0.013
0.63 0.61 0.50 0.80
SENIOR -0.017 -0.002 0.548 -0.418
0.91 0.99 0.01 0.12
SUBSIDIZED -0.531 0.223 0.049 -0.074
0.20 0.13 0.66 0.66


Observations
EXREPL


EYRELQ


RELQ_ REPL


4067
0.108
0.35
0.046
0.45
-0.010
0.97
-0.002
0.21
0.000
0.23
0.029
0.00
0.000
0.00
0.009
0.00
0.000
0.00
-0.002
0.11
0.072
0.00
0.003
0.08
-0.019
0.34
0.071
0.00
0.038
0.31
-0.677
0.00
-0.328
0.13


Market


1176 1291 490
0.070 -0.020 0.086
0.03 0.66 0.06
0.054 0.000 0.088
0.05 1.00 0.13
0.115 -0.106 0.111
0.00 0.19 0.02
-0.001 -0.010 -0.008
0.60 0.00 0.02
0.000 0.000 0.000
0.29 0.72 0.03
0.043 0.023 0.027
0.00 0.00 0.00
0.000 0.000 0.000
0.00 0.00 0.00
0.000 0.000 0.002
0.78 0.36 0.13
0.000 0.000 0.000
0.40 0.97 0.13
0.000 0.000 0.000
0.90 0.81 0.99
0.114 0.035 0.086
0.00 0.14 0.06
0.001 -0.001 0.002
0.85 0.00 0.33
-0.091 -0.232 -0.040
0.01 0.00 0.54
0.089 0.064 0.172
0.25 0.20 0.01
-0.222 0.110 -0.046
0.18 0.00 0.33
0.241 0.398 0.270
0.08 0.17 0.41
-1.060 -0.018 0.245
0.27 0.81 0.03











Table 11. Continued


Market

CONDO


Los New York
Boston Chicago Denver s C Y
Angeles City

0.140 0.171 0.298 0.154 -0.441


0.11


CONDOCONV


0.00 0.07


0.01


-0.035 0.028 -0.045 0.041


Oakland Phoenix Portland

0.018 0.103 0.028
0.74 0.12 0.85
1.415 0.049 0.006


0.74 0.74


0.75 0.87


0.514 0.149 -0.098 -0.005 0.086


0.01 0.02 0.18 0.92
-0.087 0.142 0.074 0.033


0.60 0.01


0.30 0.10


0.04


0.145 -0.041 -0.605
0.33 0.50 0.00


-0.050 0.135 0.138 -0.049


0.53


-0.003 0.272 0.351 0.146 0.034


0.98 0.00 0.00 0.00


0.65


0.119 0.366 0.271 0.298 0.129


0.00


0.00 0.00


0.09


0.356 0.535 0.287 0.529 0.365


0.04 0.00


0.00 0.00


0.00 0.03 0.55
0.468 0.180 -0.095
0.00 0.01 0.23
0.636 0.263 -0.032
0.00 0.00 0.70
0.665 0.279 0.094


0.00


0.401 0.642 0.233 0.759 0.506


0.01 0.00 0.00 0.00


0.00


0.552 0.694 0.393 0.916 0.645


0.00 0.00


0.00 0.00


0.00


0.000 0.000 0.000 0.000 0.000


0.54 0.00 0.18 0.00
0.000 0.000 0.000


0.00


0.00


0.71 0.24


0.000
0.08


0.00


0.28


0.812 0.443 0.143
0.00 0.00 0.09
0.864 0.540 0.337
0.00 0.00 0.00
0.000 0.000 0.000
0.00 0.96 0.97
0.000 0.000 0.000
0.00 0.42 0.73


0.000
0.60


SDUMi (not reported)
CONST


7.226 13.126 33.875 -33.261 -62.993 9.494 22.172 27.363


0.34 0.09
0.88 0.81


0.03 0.00
0.90 0.83


0.00
0.83


0.16
0.89


0.85


All standard errors are adjusted for potential heteroskedasticity. P-values are reported under the coefficient estimates and are in
bold and italics. Dependent Variable is LNPRICE Log of selling price; EXREPL Binary variable set equal to one if
transaction represents sale of a replacement property. EXRELQ Binary variable set equal to one if transaction represents sale of
a relinquished property; RELQ_REPL Binary variable set equal to one if transaction represents both sale of relinquished
property and purchase of a replacement property. AGE Age of the buildings) in years; AGE2 Age squared; SQFT Square
footage of total improvements; SQFT2 Square footage of total improvements squared; LANDSQFT Square footage of land;
LANDSQFT2 Square footage of land squared; PARKING Parking, defined as number of parking spaces; FLOORS Number
of floors; UNITS Number of units; CONDITION, Physical condition of the property based on inspection. The categories
include below average, average and above average. The omitted category is average; BUYEROUT Binary variable set equal to
one if buyer lives out of state; SENIOR Binary variable set equal to one if property use is senior multi-family; SUBSIDIZED
Binary variable set equal to one if property use is subsidized multi-family; CONDO Binary variable set equal to one if property
use is multi-family condominium; CONDOCONV- Binary variable set equal to one if transaction was part of condo conversion;
PORTSALE Binary variable set equal to one if transaction was part of portfolio sale; YR, Yearly time periods from 1999
through 2005. Each year is included as a binary variable except 1999, which is suppressed; SDL \I Binary variable signifying
the submarket in which the property is located, as defined by CoStar brokers; X latitude of property; Y -longitude of property.


PORTSALE


0.44


0.99


YR2000


YR2001


YR2002


YR2003


YR2004


YR2005


R-squared











Table 11. Continued


Riverside Sacrame
Sacrame
/San
Bernard nto
Bemnardmin


San San
Diego Francisco


San Jose Seattle Tucson


Observations
EXREPL


EXRELQ


RELO REPL


AGE


AGE2


SQFT


SQFT2


LANDSQFT


LANDSQFT2


PARKING


FLOORS


UNITS


CONDITION BA


CONDITION AA


BUYEROUT


SENIOR


SUBSIDIZED


CONDO


593
0.096
0.03
0.122
0.02
0.225
0.00
-0.012
0.01
0.000
0.36
0.028
0.00
0.000
0.00
0.001
0.17
0.000
0.08
0.000
0.52
-0.001
0.97
0.000
0.77
-0.056
0.58
0.046
0.39
-0.021
0.83
0.175
0.18
0.375
0.00
-0.057
0.48


370
0.158
0.00
0.182
0.00
0.182
0.00
-0.009
0.04
0.000
0.33
0.028
0.00
0.000
0.00
0.002
0.46
0.000
0.32
-0.001
0.18
0.147
0.04
0.001
0.45
-0.159
0.10
0.047
0.49
0.047
0.77



0.021
0.86


2031
0.072
0.00
0.028
0.23
0.087
0.00
-0.002
0.21
0.000
0.78
0.030
0.00
0.000
0.00
0.001
0.24
0.000
0.35
-0.001
0.04
0.109
0.01
0.007
0.00
-0.100
0.02
0.173
0.00
0.016
0.81
0.156
0.03
-0.191
0.28
0.170
0.02


796
0.076
0.02
0.062
0.01
0.109
0.00
-0.001
0.75
0.000
0.66
0.093
0.00
-0.001
0.00
0.001
0.84
0.000
0.33
0.001
0.84
0.070
0.00
0.001
0.72
-0.182
0.00
0.150
0.00
-0.008
0.92



-0.235
0.10
0.365
0.00


471
0.099
0.00
0.002
0.96
0.084
0.04
-0.002
0.74
0.000
0.99
0.053
0.00
0.000
0.00
0.005
0.15
0.000
0.18
-0.002
0.33
0.137
0.00
0.020
0.02
-0.112
0.06
0.139
0.01
-0.017
0.81
-0.891
0.00
-1.388
0.00
4.640
0.00


1281 546
0.053 0.045
0.04 0.28
0.013 0.034
0.67 0.59
0.090 0.130
0.02 0.02
-0.002 -0.006
0.38 0.07
0.000 0.000
0.72 0.39
0.026 0.021
0.00 0.00
0.000 0.000
0.00 0.00
0.000 0.002
0.63 0.05
0.000 0.000
0.65 0.59
0.001 0.001
0.20 0.00
0.158 -0.003
0.00 0.58
0.002 0.001
0.15 0.33
-0.139 -0.145
0.00 0.00
0.107 0.054
0.00 0.46
-0.012 0.197
0.81 0.00
0.061 0.248
0.80 0.00
0.085 0.446
0.24 0.00
0.348 0.401
0.00 0.07


Market











Table 11. Continued


San San
Diego Francisco


San Jose Seattle Tucson


CONDOCONV


PORTSALE


YR2000


YR2001


YR2002


YR2003


YR2004


YR2005


SDUMi (not
reported)
CONST


R-squared


-0.345
0.07
0.026
0.79
0.111
0.17
0.261
0.00
0.345
0.00
0.474
0.00
0.724
0.00
1.002
0.00
0.000
0.64
-0.001
0.67
-0.001


3.215


-0.066 0.152


0.81


0.143


0.02


-0.807 -0.252


0.00
0.144
0.07
0.217
0.01
0.474
0.00
0.620
0.00
0.852
0.00
0.876
0.00
0.000
0.89
0.000
0.73


0.15
0.046
0.33
0.209
0.00
0.441
0.00
0.664
0.00
0.907
0.00
0.848
0.00
0.000
0.00
0.000
0.02


-0.160
0.00
0.208
0.00
0.317
0.00
0.366
0.00
0.379
0.00
0.463
0.00
0.506
0.00
0.000
0.00
0.000
0.00


-0.555 0.093 0.124


0.15
0.107
0.00


-0.008 -0.118
0.88 0.19


0.351 0.024 -0.028


0.00


0.77


0.368 0.078 0.037


0.00


0.69


0.381 0.149 0.071
0.00 0.01 0.43
0.399 0.167 0.129
0.00 0.00 0.14
0.374 0.306 0.319


0.00


0.00


0.000 0.001 0.000
0.10 0.00 0.03
0.000 0.007 0.000
0.28 0.00 0.07


0.009
0.00



28.361 -9.089 -18.833 -8.118 20.801 28.200


0.36


0.60
0.85


0.69


0.85


All standard errors are adjusted for potential heteroskedasticity. P-values are reported under the coefficient estimates and are in
bold and italics. Dependent Variable is LNPRICE Log of selling price; EXREPL Binary variable set equal to one if
transaction represents sale of a replacement property. EXRELQ Binary variable set equal to one if transaction represents sale of
a relinquished property; RELQ_REPL Binary variable set equal to one if transaction represents both sale of relinquished
property and purchase of a replacement property. AGE Age of the buildings) in years; AGE2 Age squared; SQFT Square
footage of total improvements; SQFT2 Square footage of total improvements squared; LANDSQFT Square footage of land;
LANDSQFT2 Square footage of land squared; PARKING Parking, defined as number of parking spaces; FLOORS Number
of floors; UNITS Number of units; CONDITION, Physical condition of the property based on inspection. The categories
include below average, average and above average. The omitted category is average; BUYEROUT Binary variable set equal to
one if buyer lives out of state; SENIOR Binary variable set equal to one if property use is senior multi-family; SUBSIDIZED
Binary variable set equal to one if property use is subsidized multi-family; CONDO Binary variable set equal to one if property
use is multi-family condominium; CONDOCONV- Binary variable set equal to one if transaction was part of condo conversion;
PORTSALE Binary variable set equal to one if transaction was part of portfolio sale; YR, Yearly time periods from 1999
through 2005. Each year is included as a binary variable except 1999, which is suppressed; SD \ I Binary variable signifying
the submarket in which the property is located, as defined by CoStar brokers; X- latitude of property; Y longitude of property.


Market


Riverside
/San
Bernardino


Sacrame
nto









The coefficient on BUYEROUT remains positive and significant in Phoenix and

Tucson. The coefficient on CONDOCONV remains positive and significant in San

Diego. The coefficient on PORTSALE is economically and statistically significant in

Chicago, New York City, Portland and San Francisco. PORTSALE is also positive and

significant in Boston, but negative and significant in Portland and Sacramento. However

these results will not be discussed since they are based on too few observations.

Based on OLS Model II, the percentage sales price changes corresponding to the

estimated (statistically significant) coefficients for EXREPL, EXRELQ, RELQ REPL,

CONDOCONV, BUYEROUT, and PORTSALE for each market are presented in Table 12.

Percentage price effects are calculated using 100*g = 100*{exp(x) 1) consistent

with Halvorsen and Palmquist (1980). In this equation g is the relative effect on sale price

(PRICE) of the presence of a condition (e.g. replacement exchange (EXREPL),

relinquishment exchange (EXRELQ), etc); x is EXREPL or EXRELQ, or some other

condition.

The percentage price effect of the sale being part of a replacement property

exchange ranges from 5.43 percent (in Seattle) to 17.58 percent in Denver, while the

effect of the transaction being part of relinquished property exchange on sales price is

between 4.73 percent in Los Angeles to 20.02 percent in Sacramento. The price premium

associated with a sale being part of both a replacement and relinquished exchange ranges

from 8.81 percent in San Jose to 26.26 percent in Chicago. Although the coefficient on

relinquished property is harder to interpret, because of the issues discussed, it is clear that

replacement exchanges have a positive effect on price. Moreover, in the majority of the

markets (12 out of 15) this price effect is statistically and economically significant.









In comparison, the simulation analysis discussed in Chapter 5 predicts that when

the holding period, both for the relinquished and replacement property, does not exceed 5

years, the observed incremental NPV from a 1031 exchange is 1 4 percent. With a 10

year holding period, the incremental NPV increases to 4 8 percent, depending on the

assumptions for the other key variables. Finally, with a 20 year holding period

incremental value from using an exchange ranges from 4.8 to 10 percent. Therefore,

participants in tax-delayed exchanges that have a short-term investment horizon need to

be very careful, since the wealth that they lose in the form of a higher replacement

property price may offset, in whole or in part, the gain from the deferment of taxes.

Out-of-state buyers are associated with a price premium of 11.61 percent in

Phoenix and 21.77 percent in Tucson. Condominium conversions are associated with a

price premium of 16.37 percent of sales price in San Diego. Finally, portfolio sales result

in 8.99 percent higher prices in NYC, a 16.10 percent increase in price in Chicago and a

14.79 percent positive impact on price in San Francisco.

Robustness of Results

Omitted Variables Issues

As noted previously there is a concern that coefficient estimates may be biased due

to omitted variables. The magnitude of the coefficients of the variable representing a

relinquished exchange (EXRELQ) clearly shows that properties that become part of

relinquished exchanges may have some extraordinary characteristics which are not

captured by the variables in the model. I follow the Haurin (1988) and Glower, Haurin

and Hendershott (1998) approach to form a variable that controls for unusual












Table 12. Marginal Effects for Significant Coefficients for Variables of Interest

Market Obs EXREPL EXRELQ RELQ REPL BUYEROUT CONDOCONV PORTSALE

Boston 400 15.55%

Chicago 1585 10.41% 13.65% 26.36% 16.10%

Denver 782 17.58% 14.65% 21.00%

Los Angeles 7761 7.90% 4.72% 9.63%

New York City 4067 8.99%

Oakland 1176 7.28% 5.53% 12.24%

Phoenix 1291 11.61%

Portland 490 8.96% 11.69%

RiversideSan 593 10.03% 13.01% 25.23%
Bernardino

Sacramento 370 17.12% 20.02% 19.96%

San Diego 2031 7.44% 9.09% 16.37%

San Francisco 796 7.86% 6.39% 11.55% -14.79%

San Jose 471 10.41% 8.81%

Seattle 1281 5.43% 9.46%

Tucson 546 13.91% 21.77%









characteristics of a property. Each property's atypicality is computed based on the

hedonic equation's coefficients, and the formula provided in Glower, Haurin and

Hendershott (1998)


Y exp(a + b,h,) exp(a + b,hb)
ATYP = '


*100%


(20)


The property atypicality measure, A TYP is presented as percentage of selling price

(SP); h, are the physical characteristics, h1, are the mean values of these characteristics,

and a and b, are the intercept and the slope estimates from the hedonic regression (Ibid.).

The hedonic equation used has the following form

LNPRICkE = a + ajAGE+ a2AGE2 + a3SQFT+ a4SQFT2
+ a5LANDSQFT+ a6LANDSQF72 + a7PARKING
3
+ a8FLOORS+ a9UNITS+ Z /,CONDITION + a SENIOR
i=2
+ a SUBSIDIZED+ al2CONDO (21)

Location characteristics are excluded. The fundamental value of a property is presented

by the value predicted from a hedonic price equation based on a sample of apartment

properties sold across all markets that are not associated with any conditions, completed

with out-of-state buyers, or part of an exchange.

Next, I include the atypicality measure in the model specified by the equation

LNPRICF = ao + xaEXREPL+ a2EXRELQ+ o3RELQ_REPL+ 3BUYEROUT+ a9A TYP+
P<43 r r k
+ 1,SMDUM,+ y 1k,(Xk Y) (22)
s=2 k=01=0
The estimated coefficients from this model are not different from the ones

estimated with OLS Model II. This leads us to conclude that OLS Model II is well

specified.









Endogeneity Issues

There is a concern that the variable representing relinquished exchanges, EXRELQ,

is not exogenous with respect to selling price. For example, an omitted factor such as

extreme price appreciation, which is correlated with both the probability for a sale to be

part of a relinquished exchange as well as selling price, could cause a significant

relationship between the variable representing that the sale is part of a relinquished

exchange and the selling price. To address these concerns, I perform a Durbin-Wu-

Hausman (DWH) test for endogeneity. I estimate the base OLS Model II using two-stage

least squares regressions, where EXRELQ is an endogenous variable. The DWH test first

estimates the endogenous variable as a function of all exogenous variables. In the second

stage it regresses the dependent variable (in this case, natural log of price) on all variables

and includes the residuals of the endogenous variable. The second stage coefficient

estimates, when performing this procedure, are not different from the coefficients

estimated in the original model. Also the DWH test shows that there is no severe bias in

the OLS model estimates. Therefore, the results from the OLS model appear to be robust

to endogeneity concerns.














CHAPTER 8
RESULTS FOR OFFICE PROPERTIES

This chapter focuses attention on results from the empirical analysis on the

commercial real estate office market. There are 8,871 observations in the office property

sample used for regression analysis. Summary statistics for the variables of interest are

presented in Table 13. Statistics at the aggregate data level are presented in the first two

columns. With office properties, the average property price is about 2.5 times higher than

apartments. The mean sales price is $5,353,894 and the standard deviation of price is

$9,734,759. Office transactions are, on average, much larger than our apartment sales.

The average office property sold in the sample has 41,718 square feet of improvements

and is built on 78,952 square feet of land. Also, the office properties in our sample are

newer than apartments, averaging 29 years of age. A large portion of the office properties

are classified as being in excellent condition (32 percent), 66 percent are in good

condition, and only 3 percent are in below average condition. The average office property

has 2.43 floors and 100 parking spaces.

There are fifteen markets represented in this sample including: Chicago,

Dallas/Forth Worth, Denver, Fort Lauderdale, Las Vegas, Los Angeles, Oakland,

Phoenix, Riverside/San Bernardino, Sacramento, San Diego, Seattle, Tampa, Tucson and

Washington, DC. Los Angeles is the largest market represented with 1,491 sales,

followed by Phoenix with 995 observations. The smallest market is Tucson with 264

office transactions.









Table 13 also presents summary statistics of regression variables by markets. I

notice that the largest average transactions are observed in the Washington, DC area with

an average sales price of $14,800,000, which is a significantly higher than the average

price of $5,353,894 for the entire office sample. Washington, DC also has the largest

average size office property in the sample with a mean square footage of 93,519, which is

also significantly higher than the average square footage size of properties in the sample,

41,718. In contrast, the smallest average transactions based on both price and square

footage of improvements, are observed in Tucson. The average office property in Tucson

had a selling price of only $1,713,384 and is just 15,370 sq. ft. in size. Office properties

sold in Oakland and Los Angeles tend to be the oldest with average ages of 34.82 and

37.96 years, respectively. The market with the newest office properties is Las Vegas. The

mean office building age in Las Vegas is only 14.54 years with a standard deviation of

15.89.

Approximately 12 percent of the transactions in the office sample involve the

purchase of a replacement property to complete an exchange; 6 percent involve the sale

of a relinquished property; and 3 percent represent both a sale of relinquished property

and purchase of a replacement property in a separate exchange. This is in contrast to the

exchange distribution in apartment markets, where replacement and relinquished property

exchanges were approximately equal in number. The retail data, which will be analyzed

in the next chapter, also contain a much higher number of replacement exchanges than

relinquished exchanges. This leads us to the conclusion that with offices and retail

properties it is often the case that more than one property is involved in the replacement

exchange.





























C> c I--
mt t- OC t-
ID IDCII









S OC OC
C] -


m C ob1




In O -
^ri^ ^ oo i




ro oo o ^ o



-N C> 00
'i t-d
^M^ ro
mD c^cio


C o: d d
m





SC> OC ol
oooo m
C>C]oCC>



C] -N C] '



oor
C]In
s


0- 0C 00 CC C> C>
- 0 I
m m o n o^- o


biDOb
IlbCCI
C\lOr~lD
i
i


CA C O m


m
C]
SC]
~c~r


(fl CA t C> OC m o m mt
oN m V- c V) V x














OOOOmOOOO N iiO














- m n IO Dm -














000oo O(- D m m -D I-- D D D













D D C] C M D D -- C ID









-cn
^o o

II


I


II


o
























0 \D 00 -












6O6
^ t- \D c








000 Cl 00
0

o~


t(- 00 0 cm
O <0^ (-
t-O- c


> 0

3Cl
(~ 0
o
a
9


oo 00
00 ^
0'. 0'.3


IA0
~OC 0
31~-


003



t--\D ~0


r-\ o 0 C
~IOC\


2

CI "


0


0\00 d ^r|r-i^-i
\CZ CZ CZ CZ CZ
-\ CZ Cl CZ CZ CZ CZ
^^ \i r r r^r^r
-%ur %uruur


r-- N m C oli mN V) o 'tN r-- N x
NO- t 0 DmOOOO N (N 00
00000000000000000














SD NOOmoD 0 0 C -- NN-l -
00000000000000000















00000000000000000













CD 0 OC D CD CD C- 0 CD 0 C ( Cl n








00000000000000000








666666666 c66 c666 c


o
0
in

'a





o







^a





9 3

Pa
3

si
0
-:


~~~Fu

o

























































ClDh

\D Cl


0 \OD
O1OO
chi6
Cl


OClC


lf0 0 C -
&0do0rt<
\D cm --



-r- 3 0


01 00lf



00 cd











Oh \D Cl
c\D 00


t- \D

0 COC~


I \D Cl1
Cl 00




CD~l


0 l


C
o





00


2


0

VI


0
ac

















O








0
0S
Ki
i<
O



i<
o^
VI





^a


IA
O








VI


00
11C
00
C



as

B a"


0\00 d ^r|r-i^-i
o\CZC CZ, CZ, Cs C
o\CZ CZ CZ CZ CZ,
^^ \i r r r^r^r
Q^ Q^ Q^ ^ Q^ Q^ Q


S 0 OO D 0 0 \ m m 00-
00000000000000C0O









00000000000000000















00000000000000000







66666666666666666






m r Cl Clo t t^- O^ ^t- Cl \D- cn 00 o 00 \D-
O OOtmC 000----CNO
0000000000000000














NO~mobN-OOOm ---0
00Om-0000^Dm 00000
00000000000000000















666OOm66666666606


C"I~31Cl \D O
o3~-t300-








0 -~\D -~t-00
ClODho~hCl-
CIC"I- 3Chi
-CO 0


~
B FY











Approximately 16 percent of office property buyers reside out-of-state; out of

8,871 office sales 1,418 involved an out-of-state buyer. This is in sharp contrast to

apartment sales where only 7 percent of sales were completed by out-of-state buyers. The

markets with the highest share of out-of-state buyers are: Washington, DC, 44 percent;

Las Vegas, 38 percent; Phoenix, 30 percent; and Dallas/Forth Worth, 25 percent. Other

markets of potential interest when quantifying the price effect of out-of-state buyers

include Denver with 16 percent out-of-state buyers; Chicago, 11 percent; Tampa, 14

percent; Tucson, 13 percent; San Diego, 10 percent; Seattle, 9 percent; Fort Lauderdale, 8

percent; and Los Angeles, 4 percent.

Approximately 4 percent of all sales are part of sale-leaseback transactions; there

are 342 sale-leaseback transactions in the office sample. The market with the highest

percentage of sale-leasebacks is Seattle, where this type of transaction represented 7

percent of all sales. Other markets of potential interest when quantifying the effect sale-

leasebacks have on sales price include Los Angeles (56 observations), Washington, DC

(32 observations) and San Diego (30 observations).

On average, portfolio sales comprise only 2 percent of all sales in the office

sample; therefore there are only 222 portfolio sales in the sample. Washington, DC is the

market with the highest percentage of portfolio sales, where such transactions are

observed in 6 percent of all sales. The only other market where I observe a sufficient

number of portfolio sales for the purposes of quantifying potential price premiums

associated with the motivation to bundle several transactions together is Chicago, which

has 30 portfolio sales.




Full Text

PAGE 1

DO BUYER AND SELLER MOTIVATIONS A FFECT TRANSACTION PRICES IN COMMERCIAL REAL ESTATE MARKETS? EVIDENCE FROM TAX-DEFERRED EX CHANGES AND OTHER CONDITIONS OF SALE By MILENA T. PETROVA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by Milena Petrova

PAGE 3

This dissertation is dedicated to my parents.

PAGE 4

ACKNOWLEDGMENTS I would like to sincerely th ank David Ling, my dissertation chair, for his guidance and support and making sure I successfully comple te this work. I would also like to thank Wayne Archer, Andy Naranjo, Ronald Ward, Mahen Nimalendran and Thomas Barkley for their helpful comments and suggestions. I would like to gratef ully acknowledge the valuable suggestions by participants at the AREUEA doctoral meeting in Boston and seminar participants from the University of Florida, Syracuse University, Baruch College and Florida International Univ ersity. I thank CoStar Group, Inc. for providing the data used in this dissertation. Fi nally, I thank my parents and my husband, for their patience and support. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................iv LIST OF TABLES ............................................................................................................vii LIST OF FIGURES ...........................................................................................................ix ABSTRACT .........................................................................................................................x 1 INTRODUCTION ........................................................................................................1 2 TAX-DEFERRED EXCHANGES ...............................................................................7 Tax-Deferred Exchanges Overview.............................................................................7 Types of Tax-Deferred Exchanges ........................................................................8 Basic Requirements of a Valid Section 1031 Exchange .....................................10 Advantages of Tax-Deferred Exchanges .............................................................12 Drawbacks of Tax-Deferred Exchanges ..............................................................12 Exchanges and Price Effects ...............................................................................13 3 OTHER ATYPICAL MOTIVATIONS .....................................................................15 Purchases by Out-of-State Buyers ..............................................................................15 Condominium Conversion ..........................................................................................16 Portfolio Sales .............................................................................................................17 Sale-Leasebacks ..........................................................................................................18 4 THEORETICAL MOTIVATION FOR PRICE EFFECTS AND CONCEPTUAL FRAMEWORK ..........................................................................................................20 Theoretical Motivation for Price Effects ....................................................................20 Price Pressure Hypothesis ...................................................................................21 Imperfect Substitute Hypothesis ..........................................................................22 Tax Capitalization Hypothesis ............................................................................22 Other Factors .......................................................................................................23 Conceptual Framework ...............................................................................................23 5 SIMULATING THE MAGNITUDE OF PRICE EFFECTS FOR EXCHANGES ...31 v

PAGE 6

6 DATA AND METHODOLOGY ...............................................................................41 Data .............................................................................................................................41 Apartments ..........................................................................................................44 Office Properties ..................................................................................................48 Retail Properties ..................................................................................................51 Research Methodology ...............................................................................................53 7 RESULTS FOR RESIDENTIAL REAL ESTATE ....................................................61 8 RESULTS FOR OFFICE PROPERTIES ...................................................................84 9 RESULTS FOR RETAIL PROPERTIES ................................................................108 10 CONCLUSION .........................................................................................................133 BIOGRAPHICAL SKETCH ...........................................................................................142 vi

PAGE 7

LIST OF TABLES Table page 1 Incremental NPV of Apartment Exchange as a Percent of Replacement Property Value ........................................................................................................................34 2 Incremental NPV of Non-Residential Exchange as a Percent of Replacement Property Value ..........................................................................................................39 3 Description of Size of Exchange Market .................................................................43 4 Description of Apartmen t Property Sales by Markets .............................................46 5 Description of Office Property Sales by Markets ....................................................49 6 Description of Retail Pr operty Sales by Markets .....................................................52 7 List of Regression Variables....................................................................................58 8 Summary Statistics of Apartment Data by Markets .................................................62 9 Differences in Mean Prices of Control Sample and Identified Groups of Interest for Apartment Properties ..........................................................................................66 10 Regression Statistics for OLS Model with Structural Characteristics and Submarket Dummies by Apartment Markets ...........................................................69 11 Regression Statistics for OLS Model with Structural Characteristics, Submarket Dummies and Longitude, Latitude Coordinates by Apartment Markets .................75 12 Marginal Effects for Significant Coefficients for Variables of Interest ...................81 13 Summary Statistics of Office Data by Markets ........................................................86 14 Differences in Mean Prices of Control Sample and Identified Groups of Interest for Office Properties.................................................................................................90 15 Regression Statistics for OLS Model with Structural Characteristics and Submarket Dummies by Office Markets ..................................................................93 16 Regression Statistics for OLS Model with Structural Characteristics, Submarket Dummies and Longitude, Latitude Coordinates by Office Markets ......................102 vii

PAGE 8

17 Marginal Effects for Significant Coefficien ts for Variables of Interest in Office Regressions .............................................................................................................107 18 Summary Statistics of Retail Data by Markets ......................................................109 19 Differences in Mean Prices of Control Sample and Identified Groups of Interest for Retail Properties ..............................................................................................115 20 Regression Statistics for OLS Model with Structural Characteristics and Submarket Dummies by Retail Markets ................................................................118 21 Regression Statistics for OLS Model with Structural Characteristics, Submarket Dummies and Longitude, Latitude Coordinates by Retail Markets .......................125 22 Marginal Effects for Significant Coefficients for Variables of Interest in Retail Regressions .............................................................................................................131 viii

PAGE 9

LIST OF FIGURES Figure page 1 Direct Exchange (Swap) .........................................................................................9 2 Delayed Exchange with Intermediary .....................................................................9 3 Impact of Atypical Motivation on Price ...............................................................21 ix

PAGE 10

Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DO BUYER AND SELLER MOTIVATIONS A FFECT TRANSACTION PRICES IN COMMERCIAL REAL ESTATE MARKETS? EVIDENCE FROM TAX-DEFERRED EXCHANGES AND OTHER CONDITIONS OF SALE By Milena T. Petrova August 2006 Chair: David C. Ling Major Department: Finance, Insurance and Real Estate Heterogeneous buyer and seller motivations are common in real estate transactions. However, there are only a few studies limite d to one property type and market that examine the effect of atypical motivations on observed selling price. I investigate several heterogeneous motivations that could have an effect on prices of apartment, office and retail properties in fifteen major metropolitan markets. In particular, I focus on analyzing transactions motivated by tax-deferred exchanges. I find a significant positive price effect related to replacement property exchanges. This result is robust across geographic markets and property types. For example, I observe a price premium for replacement taxdeferred exchanges ranging from 5% in S eattle to 18% in Denver for apartment transactions, from 19% in Seattle to 36% in Chicago for office sales, and from 12% in Seattle to 45% in Chicago for retail sales. This dissertation represents the first work that x

PAGE 11

measures the size of the exchange market nationwide, and defines conceptually and empirically the magnitude of the effect of exchanges on transaction prices in different markets across the country and across property t ypes. The results imply that participants in tax-delayed exchanges, need to be careful because the price they pay in the form of higher replacement property price may offset in whole or in part, the gain from deferment of taxes. I also analyze properties purchased by out-of-state buyers sales that are part of sale-leasebacks, portfolio tr ansactions, or condominium co nversions. I find that these special motivations are generally associated with positive price premiums. These results have important implications since they de monstrate not only that various atypical motivations of buyers and sellers have an imp act on transaction prices in commercial real estate, but also that this impact is sensitive to the market, as well as to the property type. xi

PAGE 12

CHAPTER 1 INTRODUCTION I study the role that heterogeneous buyer a nd seller motivations play in determining sales price in commercial real estate. Ge nerally, price models are built based on the assumptions that market participants have homogeneous motivations. In appraisal literature, homogeneous motivation is cen tral to the concept of market value. 1 However, various conditions of sale which can be viewed as distinct motivations appear to be quite common in commercial real estate trans actions. The literature documents multiple situations in which sellers have heterogeneous motivation (Geltner, Kluger and Miller (1988), Quan and Quigley (1991), Sirmans, Turnbull and Dombrow, 1995, Glower, Haurin, Hendershott (1998)). Common atypical motivations include tax-deferred exchanges, bank-foreclosed property sales, liquidation sales, eminent domain sales, purchases by real estate inve stment trusts, purchases by tena nts, purchases by out-of-state buyers, condo conversions, sales-and-l easebacks, and portfolio sales. Analyzing the impact of investor motivati ons on sales price is important for several reasons. First, the sales comparison approac h, which is one of the most widely used methods for valuation in real estate, usually consider s similar properties based on 1 Market value is defined as the following: The most probable price which a property should bring in a competitive and open market under all conditions requisite to a fair sale, the buyer and seller, each acting prudently, knowledgeably and assuming the price is not affected by undue stimulus. Implicit in this definition is the consummation of a sale as of a specified date and the passing of title from seller to buyer under conditions whereby: (1) buyer and seller are typically motivated: (2) both parties are well informed advised, and each acting in what he considers his ow n best interest: (3) a reasonable time is allowed for exposure in the open market: (4) payment is made in terms of cash in U. S. dollars or in terms of financial arrangements comparable thereto; an d (5) the price represents the norm al consideration for the property sold unaffected by special or creativ e financing or sales concessions by anyone associated with the sale. National Residential Real Estate Appraisal Institute, 2006 1

PAGE 13

2 structural characteristics and location. Howeve r, if conditions of sale have an effect on price, then investors need to be aware of such impact and know how to adjust correspondingly. Second, some motivations appear to have di fferent roles in different parts of the country. For example, during the last six years tax-deferred exchanges represented approximately 19 percent of total commercial real estate sales. However, in some areas of the country, and in particular in the West Coast, excha nges represented close to 50 percent of all sales. In add ition, exchanges tend to be more frequent for apartments than for other types of commercial real estate, such as office and retail. This raises questions as to whether the impact of sale conditions varies by markets, as well as property type. Therefore, while analyzing the impact of buye r and seller motivation is important, it is not sufficient to conduct the analysis based on one market and property type, which is a common feature of the existent literature. By conducting a comprehe nsive analysis across major metropolitan markets, as well as property types, this dissertation can be used as an important reference by appraisers for the adju stment of prices when the sales comparison approach is used. Finally, although there is a large body of literature devoted to examining the influence of structural characteristics a nd location on sales price, only a few papers examine the impact of atypical motivations on price. Slade (2004) in a review article provides a summary of possible conditions that influence mo tivation. He discusses four major conditions that cause atypical motivations: Tax-Deferred Exchanges Bank-Foreclosed Properties Acquisitions by REITs Out-of-State Buyers

PAGE 14

3 He also mentions two other conditions th at may influence transactions price: reduced marketing period (liquidation) and eminent domain. In his review, Slade (2004) identifies one study on tax-deferred exchanges (Holmes and Slade, 2001), seven papers on bank-foreclosed properties (Shilling et al., 1990, Forgey et al., 1994, Hardin and Wolverton, 1996, Lambson, McQueen and Slade, 2004, Downs and Slade, 1999, and Munneke and Slade, 2000, 2001), two published studies involving sales by REITs (Hardin and Wolverton, 1999, and Lambson, McQueen and Slade, 2004), and two published papers examining the influence out-of-state buyers (Lambson, McQueen and Slade, 2004, a nd Turnbull and Sirmans, 1993). Holmes and Slades analysis (2001) shows that in the Phoenix market, tax-deferred exchanges are associated with a 7.9 percent price premium. Bank-foreclosed apartments are associated with a 22 23 percent pr ice discount (Hardin and Wolverton, 1996; Lambson, McQueen and Slade, 2004). In bank-foreclosed offices, Munneke and Slade (2000, 2001) find discounts ranging from 11 percent to 31 percent. In apartment acquisitions by REITs, different studies find premiums ranging from zero to 28 percent (Hardin and Wolverton, 1996, Lambson, McQueen and Slade, 2004). In properties sold to out-of-state buyers, Lambson, McQueen a nd Slade (2004) observe a 5 to 7 percent premium for apartments. A related stream of literature analyzes th e impact of sellers motivation on selling time. For example, in a home sellers surv ey, Glower, Haurin a nd Hendershott (1998) find that a seller, who at the time of listing has a planned time to move, sells more quickly than one that does not.

PAGE 15

4 The majority of these studies are based on either homes or residential real estate. Only Downs and Slade (1999) and Munneke and Slade (2000, 2001) study property types that are different from residential in this cas e offices. All of the prev ious studies, with the exception of one, focus on one market and one property type. The study by Hardin and Wolverton (1999) analyses REIT purchases in three markets Atlanta, Phoenix and Seattle. I examine several conditions of sale which represent distinct motivations that are frequently seen in comparable sales data and could influence sales price. In particular, I focus attention on the use of tax-deferred exchanges nationwide and their effect on observed transaction prices. Tax-deferred exchanges are transactions in which a taxpayer is able to defer payment of some, or all, of the federal inco me taxes associated w ith the disposition of real property by acquiring another property of like kind. Although Section 1031 of the Internal Revenue Code (IRC) dates back to the 1920s, exchange s under the original restrictions could only be completed as a simultaneous swap of properties among two or more parties. The required simultaneous ex change of property severely limited the usefulness of Section 1031 exchanges as a tax deferral tool because of the difficulty of synchronizing the close of two or more complex transactions. Only in 1984, in response to an earlier cour t decision related to the Starker case (Starker vs. United States, 602 F. 2d 1341 (9 th cir., 1979)), did Congress amend the original regulations to allow taxpayers mo re time to complete the transaction. More specifically, a taxpayer who initiates a Secti on 1031 tax-deferred exchange has up to 45 days after the disposition of the relinquished propert y to identify a replacement

PAGE 16

5 property and 180 days (135 beyond the 45-day peri od) to complete the delayed exchange by acquiring the replacement property (Internal Revenue Code Section, Title 26, Section 1031). Nevertheless, the Section 1031 exchange market did not fully evolve until 1991 when the Internal Revenue Service (IRS) i ssued final regulations for initiating and completing delayed Section 1031 exchanges. The use of tax-deferred exchanges has gr own significantly during the last decade. For example, in 2004, an estimated 80 percent of all commercial real estate transactions on the West Coast involved the use of an exchange (McLinden, 2004). Many real estate practitioners argue that taxpayers should use exchanges whenever possible, despite higher average transaction/selling costs than regular fully taxable sales. However, such advice may not fully account for the potential price discount associated with the sale of the relinquished property in an exchange or the price premium that may be required to obtain the replacem ent property. Separately or together, these price effects may fully or partially offset the tax deferral benefits of an exchange. Since most tax-deferred exchanges are ne gotiated and closed in private markets, quantification of the si ze and scope of the real estate Section 1031 exchange market has not been possible. In addition, very little is known about the effects of tax-deferred exchanges on observed transaction prices and whether the effects have varied over time, across geographic markets, or across property ty pes. In fact, only a handful of papers have investigated the pricing of Section 1031 exchanges (Holmes and Slade, 2001; Lambson, McQueen and Slade, 2004). As discus sed above, a major weakness of the prior studies is that they examine just one ma rket and one property type (the Phoenix apartment market); therefore the results are difficult to generalize.

PAGE 17

6 This dissertation represents the first work that defines conceptu ally and empirically the magnitude of the effect of exchanges on tr ansaction prices in different markets across the country and across property types. I also examine possible price effects associated with properties that are purcha sed by out-of-state buyers, as well as sales that are part of condo conversions, portfolio transactions or sale-leaseback transactions. The remainder of the dissertation is orga nized as follows. Chapter 2 presents an overview of tax-deferred exchanges and di scusses their potential advantages and disadvantages. Chapter 3 discusses out-ofstate buyer motivations, condo conversions, sale-leasebacks and portfolio sales. Chap ter 4 reviews the theo retical background for observing price effects associated with atyp ical motivations, develops a conceptual framework for quantifying the benefits of Section 1031 exchanges, and simulates a range of possible transaction price effects. Chapte r 5 presents a simulation analysis of the theoretical benefits from using an exchange. In Chapter 6, I present the empirical data set and research methodology. Chapter 7 contains the results of the empirical analysis for residential real estate. Chapters 8 and 9 pres ent the results of the empirical analysis for office and retail properties, correspondingl y. Finally, in Chapter 10, I summarize the results and offer some concluding comments.

PAGE 18

CHAPTER 2 TAX-DEFERRED EXCHANGES Tax-Deferred Exchanges Overview Realized gains from the sale or excha nge of real property must generally be recognized for federal income purposes (Internal Revenue Code Section 1001(c)). In general, the realized gain is equal to the net selling pr ice of the property minus the adjusted tax basis. 1 However, under Section 1031 of the IRC, real estate owners who dispose of their investment, rental, or vacati on property and reinvest the net proceeds in other like kind property are ab le to defer recognition of the capital gain realized on the sale of the relinquished property. It is important to note that a Section 1031 exchange is, strictly speaking, a tax deferral technique. The taxpayers basis in th e replacement property is set equal to the transaction price of the replacement property, minus the gain deferred on the disposition of the relinquished property. Therefore, when (if) the replacement property is subsequently disposed of in a fully taxable sa le, the realized gain will equal the deferred gain plus any additional taxa ble gain realized since the acquisition of the replacement property. However, if the subs equent disposition of the re placement property is also structured in the form of a Section 1031 exchange, the realized gain can again be deferred. 2 1 The adjusted tax basis is equal to the original cost basis (including the value of the land), plus the cost of any capital improvements undertaken since acquisition of the property, minus cumulative depreciation. 2 Tax deferral turns into permanent tax savings upon the death of the taxpayer because the basis of the property is stepped-up to its current fair market value. Thus, the taxpayers heirs can dispose of the 7

PAGE 19

8 In order for the exchanging taxpayer to avoid completely the recognition of the accrued taxable gain, he or she must acquire a property of equal or gr eater value than the relinquished property. In addition, the taxpaye r must use all of th e net cash proceeds generated from the sale of the relinquish ed property to purchase the replacement property. The transaction is taxable to the ex tent that (1) the value of the replacement property is less than the value of the relinquish ed property and (2) ther e is cash left over after the purchase of th e replacement property. Types of Tax-Deferred Exchanges There are a number of ways in which a Section 1031 exchange can be structured involving two or more of the following parties: Taxpayer : elects to relinquish his propert y via a Section 1031 exchange. Seller : owns the real estate that the taxpayer acquires as the replacement property. Buyer : purchaser of the taxpaye rs relinquished property. Qualified Intermediary : independent agent who faci litates the exchange. The qualified intermediary (QI) takes an assignment of rights in the sale of the relinquished property and the purchase cont ract for the replacement property. In short, the QI buys and then rese lls the propertie s for a fee. Although rare, the two-party exchange is the purest form of exchange. The transaction involves two partie s who simultaneously exchange (swap) properties. Title to the relinquished property is conveyed by the taxpayer to the seller and title to the replacement property is conveye d by the seller to the taxpayer. The two party exchange in depicted in Figure 1. Since the swapped properties are rarely of equal value, the party with the least valuable position will have to pay cash (or its equivalent) to the othe r party in order to balance the equity positions. Cash received as part of the transacti on must be recognized property in a fully taxable sale and not have to pay taxes on gains deferred through one or more Section 1031 exchanges.

PAGE 20

9 as a taxable gain in the year of the ex change. Thus, taxpayers exchanging into less valuable properties lose a portion of the tax deferral benefits associated with like-kind exchanges. 3 Relinquished Property Replacement Property Figure 1. Direct Exchange (Swap) Most Section 1031 transactions are delay ed exchanges that i nvolve the use of a qualified intermediary. In a delayed exchange ownership of the relinquished property is transferred to the buyer. However, the buyer of the relinquished property transfers the agreed-upon cash amount to the QI, not the taxpayer. This first phase of the delayed exchange, often referred to as the taxpayers down-leg, is depicted in the top portion of Figure 2. D i r e c t D e e d R e l i n q u i s h e d P r o p e r t y TAXPAYER BUYER INTERMEDIARY Cash D i r e c t D e e d R e p l a c e m e n t P r o p e r t y TAXPAYER SELLER INTERMEDIARY Cash FIRST PHASE (DOWN-LEG) SECOND PHASE (UP-LEG)D i r e c t D e e d R e l i n q u i s h e d P r o p e r t y TAXPAYER BUYER INTERMEDIARY Cash D i r e c t D e e d R e l i n q u i s h e d P r o p e r t y TAXPAYER TAXPAYER BUYER BUYER INTERMEDIARY INTERMEDIARY Cash D i r e c t D e e d R e p l a c e m e n t P r o p e r t y TAXPAYER SELLER INTERMEDIARY Cash D i r e c t D e e d R e p l a c e m e n t P r o p e r t y TAXPAYER TAXPAYER SELLER INTERMEDIARY INTERMEDIARY Cash FIRST PHASE (DOWN-LEG) SECOND PHASE (UP-LEG) Figure 2. Delayed Exchange with Intermediary 3 To the extent an exchanging taxpaye r must recognize a portion of the re alized gain because of the receipt of cash, the taxpayers basis in th e replacement property is hi gher and, therefore, any subsequent realized gain from the sale of the replacement property will be lower. TA XPAYER SEL LER Relinquished Property Replacement Property TA XPAYER TA XPAYER SEL LER SEL LER

PAGE 21

10 The cash paid by the buyer is parked with th e QI until the taxpayer is able to identify and close on the replacement property. Within 45 days of sale of the relinquis hed property, the taxpayer must identify the replacement property. The identification must be specific, such as the address of the property to be acquired. To allow for the possi bility that the taxpayer may not be able to come to terms with the owner of the potential replacement property, the taxpayer may designate more than one replacement property. 4 The taxpayer must acquire one or more of the identified replacement properties within 180 days of the date of the closing of the relinquished property; that is, the 45 and 180 day periods run concurrently. There are no exceptions to these time limits and failure to comply will convert the transaction to a fully taxable sale. At the closing of the replacement property, the QI transfers cash to the seller of the replacement property and the seller transfers ownership to the taxpayer. This second pha se of the delayed exchange, often referred to as the taxpa yers up-leg, is depicted in the bottom portion of Figure 2. 5 Basic Requirements of a Valid Section 1031 Exchange In general, both real and personal proper ty can qualify for tax-deferred treatment. However, some types of property are specifi cally disqualified; for example, stocks, bonds, notes, and ownership interests in a li mited partnership or multi-member limited 4 More specifically, the taxpayer can (1) identify up to three properties of any value or (2) identify more than three properties so l ong as their combined values do not ex ceed 200 percent of the value of the relinquished property. 5 If the taxpayer closes on the replacement property prior to the sale of the relinquished property, the transaction becomes a reverse exchange. This dataset does not include a large enough sample of reverse exchanges to examine empirically how the pricing of such exchanges varies from delayed exchanges and fully taxable sales. I therefore do not discuss reverse exchanges here.

PAGE 22

11 liability company. 6 Both the relinquished property and the replacement property must be held for productive use in trade or business or held as a long-term investment. Thus, personal residences and property held for sa le to consumers (i.e., dealer property) cannot be part of a Section 1031 exchange. 7 A holding period equal to or greater than one year is commonly assumed to qualify the relinq uished property as a long-term investment for the purposes of implementing a tax-deferred exchange; however, the one year rule of thumb has no basis in statutory or case law. For a transaction to qualify as a Sect ion 1031 exchange there must (1) be a reciprocal exchange (rather than a sale for cash) and (2) the exchange must involve likekind property. An exchange is clearly crea ted by the use of a QI and the required exchange documentation. Like kind means sim ilar in nature or ch aracter. In fact, virtually any real estate is lik e-kind to any other real estate. However, real property is not like-kind to personal property. Therefore, for example, a warehouse cannot be exchanged for jewelry. In addition, foreign property cannot be exchanged for U.S. property. With the development of IRS regulations concerning Section 1031, tax-deferred exchanges are also used to trade lines of business, such as such as television and radio stations, newspapers, distributorships, and franchises, including, among others, sports teams, beer distributorships, and professional service practices (McBurney, 2004) Because a line of business includes multiple classes of assets real, personal and intangible property an exchange for each class needs to be completed (McBurney and Boshkov, 2003; McBurney, 2004). 6 Since 2003, a percentage ownership interest as a tenant-in-common (TIC) is qualified property for the purposes of a Section 1031 exchange. The taxpayer, however, must be careful that the TIC has been structured to avoid its re-characterization by the IRS as a partnership for federal income tax purposes. 7 Vacation homes will only qualify if they have been rented out the majority of the year.

PAGE 23

12 Advantages of Tax-Deferred Exchanges The tax literature and popular press point to several motivations for use of Section 1031 exchanges. First, exchanges serve as an effective shelter from taxes, thereby preserving investment capital. In addition, ex changes can be used to upgrade portfolios (Fickes, 2003). By deferring taxes, the taxpaye r can also leverage appreciation and afford to acquire a larger/higher priced replacem ent property. Section 1031 exchanges can also be used to consolidate or diversify propertie s, exchange low-return properties for highreturn properties, or to substitute depreciable property for non-depreciable property (Wayner, 2005a and 2005b). Drawbacks of Tax-Deferred Exchanges Despite the advantages of tax-deferral Section 1031 exchanges have several drawbacks. First, the taxpayers basis in the replacement property is set equal to the market value of the replacement property, minus the deferred gain. Thus, the larger the amount of tax-deferral, the smaller is the de preciable basis in the replacement property and, therefore, the smaller is the allowable deduction fo r depreciation. Moreover, the larger the amount of tax-deferra l, the larger will be the realized gain if and when the replacement property is subsequently di sposed of in a fully taxable sale. A second disadvantage is that the tran saction costs (both monetary and nonmonetary) associated with initiating and co mpleting the exchange will likely exceed the costs of a fully taxable sale. The additional costs may include settlement fees, intermediary fees, and attorney prepara tion fees (Wayner, 2005b). These first two disadvantages are explicitly considered in th e conceptual model presented in Chapter 4. An additional disadvantage is that Sec tion 1031 exchanges do not allow for the recognition of a loss for tax purposes. Thus, taxp ayers will avoid using exchanges if they

PAGE 24

13 have not realized a positive capital gain. Also, unlike the proceeds from a cash out refinancing, tax-deferred exchanges do not provide a method for drawing tax-free cash out of the relinquished property. This is becau se any cash received in the year of the exchange is fully taxable. Exchanges and Price Effects If a taxpayer is successful in completing a simultaneous or delayed tax-deferred exchange, the realized tax liability will be deferred until the replacement property is subsequently disposed of in a fully taxable sale. A portion of the re alized gain will be recognized in the tax year in which the exchange occurs to the extent that the value of the relinquished property exceeds the value of th e replacement property. The present value of the income tax deferral benefit is therefore a function of the magnitude of the deferred capital gain, the expected holding period of the replacement property, and the applicable discount rate. A taxpayer entering into a tax-deferred exch ange can afford to accept an offer for the relinquished property that is lower than th e investment value he or she places on the property by an amount that is equal to, or le ss than, the present valu e of the income tax deferral benefit. That is, depending on current market conditions, in cluding liquidity, the negotiating abilities of the taxpayer and potential buyers, and whether or not potential buyers are aware the taxpayer is initiating a Section 1031 exchange, the selling taxpayer may be willing or required to share a portion of the expected tax deferral benefits with the buyer of the relinqu ished property. Since Section 1031 exchanges are a tax-deferral technique, sellers will not enter in excha nges, unless their propert y has appreciated in value and (or) has significantly depreciated. If avoiding capital gain s tax is the main motivation for participating in an exchange th en all else equal reli nquished properties will

PAGE 25

14 have higher values than propert ies that are not part of an exchange. Therefore, the net effect of discussed above factors is not clear. I expect to find that prices paid for relinquished properties will be higher or less than transaction prices in fully taxable sales, all else equal. In contrast, taxpayers face significant compliance risk when seeking to complete the second leg of a tax-deferred exchange by identifying and purchasing a replacement property within the 45and 180-day time lim its. The strict time requirements imposed by IRS regulations, in addition to the complicated nature of tax-deferred exchanges, can lead to a temporary increase in demand in order for investors to find replacement properties and close the exchanges in a timely manner. In perfect markets, a temporary increase in demand by exchange motivated investors has no effect on market pr ices because supply can instantaneously respond. However, commerci al real estate markets are known to be thin and less elastic. Theref ore, I expect that buyers cl osing on a tax-deferred exchange transaction may be willing or required to give up some of their benefits from deferring taxes and pay a premium for the replacement property relative to its fair market value in order to acquire their property within the ti me constraint. In a competitive market the amount of the price premium w ill not exceed the expected pres ent value of tax deferral, with the actual magnitude of the premium again depending on market liquidity, the negotiating abilities of the ta xpayer and potential sellers, and whether or not potential sellers are aware the taxpayer is attemp ting to complete a Section 1031 exchange.

PAGE 26

CHAPTER 3 OTHER ATYPICAL MOTIVATIONS Purchases by Out-of-State Buyers Anecdotal evidence suggests that out-of-s tate buyers, especia lly from higher-cost areas, pay more for real estate than in-state buyers, especially those residing in lower cost areas. This observation is explained by buyers anchoring to the higher values in their home area and therefore being willing to pay a premium for real estate in lower-cost areas. An example of anchori ng are sales in neighboring stat es to California. Investors from California will be more willing to offer a premium for a property in Arizona or Las Vegas, since they are anchoring on the high prices of real estate in their home state. The anchoring phenomenon is explained by behavi oral literature. Slovi c and Lichtenstein (1971) and Tversky and Kahneman (1974) were the first academics to discuss heuristics and biases. In the real estate literature, Northcraft a nd Neale (1987) presented strong evidence of anchoring in property pricing that was similar for both amateurs and real estate professionals. Other real estate studies that find ev idence for anchoring include Black and Diaz (1996), Diaz and Hansz (1997) Diaz and Wolverton (1998), and Diaz, Zhao and Black (1999). Additional evidence of anchoring in real estate is seen in appraisal smoothing. A large body of literatur e discusses smoothing in appraisal based indexes, due to appraisal values lagging true prices and being too reliant on historical prices (Geltner (1989) and Webb (1994)). Turnbull and Sirmans (1993) attribute observed out-state buyers price premiums to higher search costs. In their model, buyers with higher search costs will search less than 15

PAGE 27

16 investors with lower search costs, and therefore will tend to overpay on average. Lambson, McQueen and Slade (2004) identify th ree factors that cont ribute to observed premiums: biased beliefs, high search cost s and time pressure, namely the haste associated with out-of-state buyers. Thei r paper examines 2,854 apartment sales in Phoenix during 1990 mid 2002 and finds a 5 7 percent premium associated with purchases by out-of-state buyers. One limita tion of the Lambson, McQueen and Slade study is that it is based on one market and one property type. Therefore, it is not clear to what extent their findings carry over to other property t ypes and markets. The objective of this paper is to examine not only whether there is any price premium associated with purchases by out-of-state buyers, but also to es tablish to what extent such premiums vary accross markets and by property types. Condominium Conversion The conversion of rental properties to condominium ownership have been very popular in the recent past, induced by increa sing house prices and lagging rent levels. This trend is expected to slow or be reve rsed in 2006 given an oversupply of condos and shortage of apartments in some markets. In competitive markets the motivation of the condo converter should not ha ve an impact on sales pric es. However, as Lambson, McQueen and Slade (2004) point out apartmen t complexes trade infrequently with high transaction costs and real estate buyers have heterogeneous information (p. 86). Therefore, an investor that is buying a multifamily property with the objective of converting its units to condos may be willing to pay a premium because of the expected higher price he will net per condo sold.

PAGE 28

17 Portfolio Sales Previous studies on portfolio sales fo cus on the stock price effects from announcements of portfolio sales. A portfolio sale is defined as a transaction in which two or more unrelated properties are sold to the same buyer. Studies that are based on property sales prior to 1992 found no abnormal returns associated with portfolio sales announcements. Glascock, Davidson and Sirmans (1991) examined 51 real estate portfolio purchases prior to 1986 and find that abnormal returns are in significantly different from zero. McIntosh, Ott and Lia ng (1995) reached a similar conclusion based on a 54-transaction sample during 1968-1990, in which all of the acquirers are Real Estate Investment Trusts (R EITs). Booth, Glascock and Sarkar (1996) studied a sample of 94 portfolio acquisitions and also failed to observe any significant abnormal returns. These findings are consistent with expectations in competitive markets, since they signify that no change in shareholde r wealth results whether a ssets are acquired and sold separately or together. However, a more recent study based on th e modern post-1992 REIT market by Campbell, Petrova and Sirmans (2003) finds av erage abnormal returns for the acquirer of approximately 0.5 percent in a study base d on 209 portfolio acquisitions during 19952001. As the authors point out, these results sugg est a clear change in the regime of such transactions, post 1992, when new regulations relaxed the restrictions on ownership concentrations for Real Estate Investment Tr usts and thereby led to an increase of capital flows to REITs, especially by institutions. Campbell, Sirmans and Petrova (Ibid.) present evidence that the observed returns are relate d to the positive eff ect of reconfirming geographical focus, signaling by taking on private debt and private placement of stock with institutions.

PAGE 29

18 None of the existing studies addre sses whether properties with similar characteristics will sell at a premium (discount) if they are part of a portfolio transaction. By bundling together several properties in one transaction, buyers may enjoy acquisition economies of scale, decreased transaction costs, as well as decreased search costs. A portfolio acquisition can provi de the buyer with quick expos ure or focus in a desired geographical or property market. However, such transactions are also more complex to negotiate and complete. All else equal, a portfolio buyer may be willing to give up some of the expected benefits from acquisition a nd pay a premium to obtain a desired portfolio. The premium paid will also depend on the mark et power of the real estate portfolio buyer. Therefore, in cases involving large in fluential buyers (e.g. REIT s, or institutions) a discount associated with real estate portfolio purchase may be observed. Anecdotal evidence from the real estate professional press suggests that portf olio acquisitions by REITs are frequently the quickest and cheapes t way for REITs to acquire properties. They can consequently turn around and dispose of less attractive properties. Sale-Leasebacks In a sale-leaseback transaction, a firm sells an asset, such as real estate property or equipment, to another firm and simultaneously leases it back. Academic research focuses on tax motivation as the major source of valu e creation in corporat e sale-leasebacks (see Miller and Upton (1976), Lewellen, Long and McConnell (1976), Myers, Dill and Bautista (1976), Brealey and Yong (1980 )). Recent studies including Smith and Wakeman (1984), Alvayay, Rutherford and Smith (1995), Moyer and Krishnan (1995) and Lasfer and Levis (1998) c onfirm the importance of tax rela ted motivations to lease or buy, but also acknowledge other non-tax incentives. Leasing offers benefits to the lessor by increased non-debt tax shields through de preciation. Sale-leasebacks create value

PAGE 30

19 when there is a difference between applicable tax rates of the lesso r and lessee, namely the lessor is in a higher tax br acket, while the lessee is in a lower tax bracket. Studies on the effect of sale-leaseback announcements on the lessees and lessors share prices record a positive effect on the lessees share price (see Slovin, Sushka and Poloncheck (1990), Allen, Rutherford and Springer ( 1993), Ezzell and Vora (2001) and Fisher (2004)). Slovin, Sushka and Poloncheck (1990) conclude that the observed positive market reaction to announcements of a sale -leaseback is due to the perception of reduction of present value of expected taxes and present evidence that gains from saleleasebacks accrue only for the lessees. In similar spirit, Lewis and Schallheim (1992) assume that leasing offers the opportunity fo r transferring non-debt tax shields and posit that if the lessee firm can locat e a lessor firm who is more able to enjoy such tax shields then the buyer will pay more than they are worth to the lessee (p.498, Ibid.). According to the authors, this higher price takes the form of reduced lease payments. In this study, sale-leaseback s are viewed as another example of atypical motivation. Under the Tax Capitalization Hypothesis and consistent with the expectations by Lewis and Schallheim (1992), I hypothesize that any ex pected tax benefits by the lessor may be capitalized into a higher purchas e price. Therefore, everything else equal, properties that are part of sale-leaseback transactions will have higher sales price.

PAGE 31

CHAPTER 4 THEORETICAL MOTIVATION FOR PRICE EFFECTS AND CONCEPTUAL FRAMEWORK Theoretical Motivation for Price Effects Generally, price models assume that buyers (sellers) are homogeneous in their motivation to buy (sell). Various conditions of sale, which can be viewed as distinct motivations appear to be quite common in commercial real estate transactions. Such conditions of sale can impact transaction pr ices. Motivated buyers can create a temporary increase in demand in property markets. In perfect markets, a temporary increase in demand by motivated investors has no effect on market prices because supply can instantaneously respond. However, in commerc ial real estate markets the supply of available properties is less el astic to shocks in demand (Wheaton and Torto, 1990; Eppli and Shilling, 1995). Jones and Orr (1999) point to differe nces in elasticity across real estate markets, with inelas tic supply most severe in re tail and office properties. Therefore, equilibrium price may change in order to eliminate excess demand. In sales motivated by a tax-deferred exchange, an out-of-state buyer, a condo-converter, a saleleaseback transaction or a portfo lio sale, I expect a positive ef fect on selling price, due to increased demand based on any of these atypical motivations. This effect is presented in Figure 3. In finance theory, both the Price Pressure Hypothesis and the Imperfect Substitute Hypothesis allow for changes in equilibrium price in response to shocks in demand. 20

PAGE 32

21 Ln(Sales Price) True Price Price with Atypical Motivation Square Feet of Improvements Ln(Sales Price) True Price Price with Atypical Motivation Square Feet of Improvements Figure 3. Impact of Atypical Motivation on Price Price Pressure Hypothesis The Price Pressure Hypothesis (PPH), deve loped first by Scholes (1972) posits that if the number of shares outstanding is incr eased by a secondary offering or large block sales, investors need to be offered a sweeten er in the form of a reduced share price, so that they are willing to hold more shares. This leads to a temporary decline in share prices. Scholes (1972) tests this hypothesis by examining 1,200 secondary distributions of large block sales during 1947-1965, but finds no significan t price effect. Kraus and Stoll (1972) further develop th e PPH and posit that in an imperfect market, with few investors, trading ma y produce significant price changes if the expectations of the marginal se ller of the security are different from the marginal buyer. For example, it may be difficult for a large seller to distribute his shares at the same price. As a result, two types of distri bution effects may arise. The fi rst type is due to liquidity costs. This is a temporary effect in which the different costs of finding willing investors can move the transaction price away from th e equilibrium price. In support of the PPH,

PAGE 33

22 Kraus and Stoll find evidence for some form of distribution effect Mikkelson & Partch (1985), Hess & Frost (1982), and Harris & Gurel (1986) also find empirical evidence supporting the PPH. Imperfect Substitute Hypothesis The second hypothesis that supports price ch anges in response to excess demand is the Imperfect Substitute Hypothesis (ISH). The ISH is developed as the second distribution effect discussed by Kraus & Stoll (1972). This effect is due to different investor preferences for a given security. This is a permanent price effect, and depends on the number of investors and on the substitutability of one security for another. In public securities markets, the Imperfect Substitute Hypothesis allows for equilibrium price changes to eliminate the excess demand. Evid ence of demand induced price shocks can be found in the impact of large block trades on prices (Scholes (1972), Kraus and Stoll (1972) and Mikkelson and Parch (1985)). Section 1031 exchanges may create a si milar tax-induced demand shock. In particular, the compliance risk in delayed exchanges could create increased acquisition demand, which may force the taxpayer to share some of the expected tax-deferral benefits with the seller of the replacem ent property in the form of an increased purchase price. In addition, if the seller of the replacement propert y is aware that the ta xpayer is seeking to complete a tax-deferred exchange, the taxpayers bargaining position is clearly compromised. Tax Capitalization Hypothesis In competitive markets, the value of commerc ial real estate fully reflects the current and expected future tax treatment of depr eciable assets (Hendershott and Ling, 1984, and Ling and Whinihan, 1985). Expected increases in tax liabilities are capitalized in the

PAGE 34

23 value of the assets, resulting in lower asse t values. The use of a tax-free exchange constitutes a reverse tax-capitalizati on (Holmes & Slade, 2001). The buyer can distribute some of his expected tax benefits due to tax-deferral in order to outbid other potential buyers. With sale-leasebacks, leasing offers th e opportunity for tran sferring non-debt tax shields (Lewis and Schallheim, 1992). Unde r the Tax Capitalization Hypothesis and consistent with the expectations by Lewi s and Schallheim (1992), I hypothesize that any expected tax benefits by the lessor can be expressed in the form of a higher purchase price. Other Factors With out-of-state buyers a possible premium in price can be explained by anchoring. Evidence for anchoring in real esta te is given by Northcraft and Neale (1987), Black and Diaz (1996), Diaz and Hansz (1997) Diaz and Wolverton (1998), and Diaz, Zhao and Black (1999). A second explanati on for a possible prem ium involves higher search costs for out-of-state buyers (Turnbull and Sirmans, 1993). Time pressure to buy is a third possible explana tion for expected price premiums (Lambson, McQueen and Slade, 2004). Conceptual Framework The following framework will focus on conceptualizing the net benefits from using an exchange. Assume a taxpayer who owns an income producing property has decided that the risk-return characteri stics of her portfolio would be enhanced by disposing of the asset and reinvesting his (her) equity into a replacement prope rty located in a market with more growth potential. Assume also that the replacement property has already been identified. The first strategy available to th e taxpayer is to di spose of the existing

PAGE 35

24 property in a fully taxable sale and then use the net proceeds, along w ith additional equity capital, to acquire the replacem ent property. The second option is to take advantage of Section 1031 of the IRC and exchange out of the existing property and into the replacement property. The second strategy woul d allow the taxpayer to defer recognition of the taxable gain that has accrued on the existing property. The net present value of the sale-purchase strategy, NPVSALE t assuming all-equity financing, can be represented as n i n s nt dr s ntcgntnt i s i oio t t tk RECAP CG SCP k DEPI P ATSP NPVSALE1 ,2 ,2 2 2 ,2 2 1)1( )1( )1( ) ( (1) where: 1 tATSP = the net after-tax proceeds from the sale of the existing property at time t ; Pt 2 = the acquisition price of th e replacement property at time t ; o = the taxpayers marginal tax rate on ordinary income; I i = the expected net cash flow of the replacement property in year i of the expected nyear holding period; s iDEP,2 = allowable depreciation on the replacement property in year i conditional on a sale-purchase strategy; k = the required after-tax rate of return on unlevered equity; 2 ntP = the expected price of the replacement property in year t + n; 2 ntSC = expected selling costs on the disposition of the replacement property in year t + n; cg = the tax rate on capital gain income; s ntCG,2 = expected capital gain income on the sale of the replacement property in year t + n, conditional on a sale-purchase strategy; dr = the tax rate on depreciation recapture income; and s ntRECAP,2 = depreciation recapture income on the sale of the replacement property in year t + n, conditional on a n-year sale-purchase strategy. The first term on the right-hand-side of equation (1) represents the additional equity capital that must be invested at time t under the sale-purchase stra tegy, and is equal to the after-tax proceeds from a fully taxable sale minus the acquisition price of the replacement property at time t As is detailed below, if the pri ce of the replacement property is equal

PAGE 36

25 to the price of the existing property, then 21 ttPATSP is equal to total taxes due on the sale of the existing propert y, plus total selling costs. The second term on the right-hand-side of equation (1) represents the cumulative present value of the replacement propertys ne t cash flows from annual operations, plus the present value of the annual tax sa vings generated by depreciation. Annual depreciation, is equal to s iDEP,2 RECPER PL DEPtt s i 22 ,2)1( (2) where is the acquisition price of the replacement property, is the percentage of that represents non-depreciable land, and RECPER is the allowable co st recovery period for the replacement property. 2 tP 2 tL 2 tP 1 Since the replacement propert y is purchased with the proceeds from a fully taxable sale, the origin al tax basis of the replacement property is stepped up to equal the total acquisition price, thereby maximizing allowable depreciation deductions over the expected n -year holding period. 2 tP The third and final term on the right-ha nd-side of equation (1) represents the expected after-tax cash proceeds from the sale of the replacement property at the end of the assumed n-year holding period. Deducted from th e expected selling price of the replacement property at time t+n are the following: expected selling costs ( ), the expected capital gain tax liability ( 2 ntSC cg s ntCG,2 ), and the expected depreciation recapture tax ( dr s ntRECAP,2 ). 1 Congressional legislation has repeatedly altered the pe riod of time over which rental real estate may be depreciated. Currently, residential real property (e.g., apartments) may be depreciated over no less than 27 and 1/2 years. The cost recovery period for nonresiden tial real property (e.g., shopping centers, industrial warehouses, and office buildings) is 39 years.

PAGE 37

26 Under current federal income tax law, all taxable income from property sales must be classified as either ordinary income, capital gain income, or depreciation recapture income. The distinctions are important because capital gain income, under the tax rules in place in 2006, is subject to a maximum 15 percent tax rate. 2 In contrast, the maximum statutory rate on ordinary income is 35 percent. Assuming the taxpayers asset is classified as trade or business property and has been held for more than one year, the total taxable gain on the sale of the replacement property in year t+n has the following two components s nt ntnt s ntUNDBASIS SCPCG,2 2 2,2) ( (3) and n i s i s ntDEP RECAP1 ,2 ,2, (4), where is the un-depreciated cost basis of the replacement property at time t+n. More specifically, is equal to acquisition price of the replacement property ( ), plus any capital expenditures over the n-year holding period. s ntUNDBASIS,2 s ntUNDBASIS,2 Pt 2 3 Note that the magnitude of is conditioned upon whether the replacement property was acquired via an exchange or with a sale-purchase strategy. Also note that is the amount by which the original acquisition price of the replacemen t property (plus any subsequent capital improvements) is expect ed to increase in nominal value over the ns ntUNDBASIS,2 s ntCG,2 2 From 1997 to May 6, 2003, the maximum capital gain tax rate was 20 percent. 3 According to the IRS, a capital expenditure increas es the market value of the property. In contrast, expenditures deemed by the IRS to be operating expenses maintain, but do not fundamentally alter, the market value of the property. Capital expenditures are not depreciable in the y ear in which they are incurred. Rather, they are added to the tax basis of the property and then systematically expensed through annual depreciation deductions.

PAGE 38

27 year holding period. Also, captures the portion of the total taxable gain on sale that results from depreciation. Total taxes due on the sale of the re placement property in year t+n conditional on a sale-purchase strategy, are therefore expected to be s ntRECAP,2 s ntTDS,2 s nt dr s ntcgRECAP CG,2 ,2 (5) Next, the components of the after-tax proceeds from a fully taxable sale of the relinquished property at time t (the first term on the righthand-side of equation (1)) are examined. Note that 1111 tttcgtdr 1 t A TSPPSCCGRECAP (6) where is equal to the price of th e relinquished property and is equal to total selling costs at time t. Similar to the subsequent sale of the replacement property at time t+n the capital gain and depreciation recapture portions of the total taxable gain from the sale of the relinquished property at time t are 1 tP 1 tSC 1 111)(t tttUNDBASIS SCPCG (7) and h i i tDEP RECAP1 1 1, (8) where is the undepreciated cost basis of the relinquished property at time t and is equal to 1 tUNDBASIS 1 iDEP RECPER PL DEPhtht i 11 1)1( (9) 1 htP represents the acquisition price of the existing property when purchased h years ago, is the percentage of the orig inal acquisition price that was non1 htL

PAGE 39

28 depreciable, and RECPER is the allowable cost recovery period for the existing property. Total taxes due on the sale of th e relinquished property at time t are therefore expected to be 1 tTDS 1 1 t drtcgRECAP CG (10) The second acquisition option available to the taxpayer is to take advantage of Section 1031 and exchange into the replacement property. Th e net present value of the exchange strategy, assuming all-equity financing, can be represented as n i n e nt dr e ntcgntnt i e ioio ttttk RECAP CG SCP k DEPI PECP NPVEX1 ,2 ,2 2 2 ,2 2 1)1( )1( )1( (11) where: tEC = the total cost of exchanging out of the relinquished property and into the replacement property at time t ; e iDEP,2 = depreciation on the repl acement property in year i conditional on an exchange strategy; e ntCG,2 = the expected capital gain income on the sale of the replacement property in year t + n, conditional on an exchange strategy; and e ntRECAP,2 = depreciation recapture income on the sale of the replacement property in n years assuming an exchange at time t All other variables are as previously defined. The capital gain and recapture components of the total taxable gain on the sale of the replacement property at time t+n, conditional on an exchange strategy are e nt ntnt e ntUNDBASIS SCPCG,2 2 2,2) ( (12) and 2 1 n ,e ,e tn i i 2 R ECAPDEP (13)

PAGE 40

29 where is equal to the acquisition pr ice of the replacement property ( ) minus the taxable gain that was deferred at time t ( DEFGAIN e ntUNDBASIS,2 Pt 2 t ) by executing an exchange strategy. DEFGAIN t = + and is equal to 1 tCG 1 tRECAP e iDEP,2 ) ()1(22 ,2RECPER DEFGAIN PL DEPt tt e i (14) where is the percentage of the replacement propertys acquisition price that is nondepreciable and RECPER is the allowable cost recovery period for the existing property. Note that reducing the tax basis of the replacement property by the amount of the deferred gain insures that > Total taxes due on the sale of the replacement property in year t+n conditional on an exchange strategy at time t are expected to be 2 tL s iDEP,2 e iDEP,2 e ntTDS,2 e nt dr e ntcgRECAP CG,2 ,2 (15) The taxpayer should exchange into the repl acement property if th e net present value of the exchange strategy [equation (11)] ex ceeds the net present value of the salepurchase strategy [equation (1)]. To determine the net present value of the sale-purchase strategy, equations (7) and (8) ar e first substituted in to equation (6). Equations (2), (3), (4), and (6) are then substituted into equation (1). To determine the net present value of the exchange strategy, equations (12), (13), a nd (14) are substitute d into equation (11). Finally, subtraction of equa tion (1) from equation (11) pr oduces the following expression for the incremental NPV of the exchange strategy

PAGE 41

30 n i n e nt s nt dr i e i s io ttttk RECAP RECAP k DEP DEP TDSECSC INCNPV1 ,2 ,2 ,2 ,2 1 1)1( ) ( )1( ) ( ] [ n s nt e ntcgk CGCG )1( ) (,2 ,2 (16) The first term in equation (16), captures the immediate net benefit of tax deferral. Note that if the time t selling costs associated with the salepurchase strategy and exchange strategy are equa l, the advantage of the exchange is equal to the deferred tax liability. To the ex tent exchanges are more expensive to execute than sales, will be negative and this incremental outflow will be netted against the positive deferral benefits. ] [1 1 t ttTDSECSC 1 tTDS ttECSC1 As noted above, the tax basis of the repl acement property is reduced by the amount of the taxable gain deferred by th e exchange, which insures that > The second term in equation (16) captures the cumulative present value of the foregone depreciation deductions over the n -year holding period. However, to the extent annual depreciation deductions are reduced by an exchange, the amount of depreciation recaptured when the replacement property is sold in year t + n is reduced by an exchange. The present value of the reduced depreciation recapture taxes is reflected in the third term in equation (16). s iDEP,2 e iDEP,2 Finally, because the tax deferral associated with an exchange reduces the tax basis in the replacement property, the taxable capita l gain due on the sale of the replacement property will be larger with an exchange. The negative effects of the increased capital gain tax liability on the incremental NPV of an exchange are captured by the fourth term in equation (16).

PAGE 42

CHAPTER 5 SIMULATING THE MAGNITUDE OF PRICE EFFECTS FOR EXCHANGES As previously discussed, taxpayers face significant compliance risk when seeking to complete a tax-deferred exchange. Mo reover, the exchanging taxpayer may have compromised his or her bargaining position with potential sellers of replacement properties. As a result, the taxpayer may be forced to pay a premium for the replacement property, assuming the marginal (price determin ing) buyers and sellers in the market are not motivated by Section 1031 tax deferral benefits and compliance issues. In a competitive market, the magnitude of the price discounts accepted by sellers of relinquished properties and th e price premiums paid by acquirers of replacement properties should not exceed the incremental NPV of the exchange strategy, with the actual magnitude of the premium depending on market liquidity, the negotiating abilities of the taxpayer and other potentia l buyers and sellers, and whether or not potential buyers and sellers are aware the taxpayer is attemp ting to complete a Section 1031 exchange. Before turning to the empirical estimates of the price discounts offered by sellers of relinquished properties and the price premiums paid by purchasers of replacement properties, equation (16) is used to simulate the magnitude of INCNPV t under a number of plausible assumptions. Simulated values of INCNPV t are then divided by the price of the replacement property at time t to determine the percentage price effect. These simulations are intended to quantify the maximu m percentage price effects that are likely to be found in the subsequent empirical work. 31

PAGE 43

32 To solve equation (16) numerically, the following base-case assumptions for the parameter values are made: Price of relinquished an d replacement property: = Pt 1 Pt 2 Cost recovery period ( RECPER ): 27.5 years for residential and 39 years for nonresidential commercial properties Selling costs ( and ): 3 percent of sale price 1 tSC 2 ntSC Exchange costs ( EC t ): equal to 1 tSC Ordinary income tax rate ( o ): 35 percent Capital gain tax rate ( cg ): 15 percent Depreciation recapture tax rate ( dr ): 25 percent After-tax discount rate ( k ): 8 percent Non-depreciable portion of original tax basis ( and ): 20 percent 1 htL 2 tL The price of the replacement property is as sumed to be equal to the price of the relinquished property to abstract for any effects unequal equity positions would have on time t inflows and outflows as well as future depreciation deductions. Note that the assumed magnitude of = does not affect the numerical simulation results because INCNPV Pt 1 Pt 2 t is divided by the price of the repl acement property to produce a percentage price effect. Other key variables in the calculation of INCNPV t include the number of years since acquisition of the relinquished property, HOLD 1 the annualized rate of price appreciation since acquisition of the relinquished property, 1 and the expected holding period of the replacement property, HOLD 2 1 Table 1 represents the simulation results for residential commercial real estate. The top panel in Table 1 contains the base case simulation results. One pattern is noteworthy: the incremental value of an exchange is unambiguously positively related to HOLD 1 For example, assuming HOLD 2 = 8, HOLD 1 = 5, and 1 = 6 percent, INCNPV t is equal to 1 t It is straightforward to show that the value of INCNPV from equation (16) is not affected by the rate at which the replacement property is exp ected to appreciate in nominal value.

PAGE 44

33 2.48 percent of This implies the taxpayer could afford to pay up to a 2.48 percent premium for the replacement property, assuming they did not agree to a price discount on the sale of the relinquished property. Pt 2 As HOLD 1 increases to 10, the maximum pri ce impact rises from 2.48 percent to 4.03 percent. Assuming HOLD 1 = 20, the maximum price impact increases further to 5.33 percent. In short, the relative attr activeness of the ex change strategy is unambiguously positively related to the magnitude of the accumulated gain, all else equal. The relation between INCNPV t and 1 however, for a given HOLD 1 is less clear. For example, assuming HOLD 1 = 5, increased price appreciation produces slight increases in INCNPV t However, with HOLD 1 = 20, higher values of 1 produce lower values of INCNPV t With HOLD 1 = 10, the relation between 1 and INCNPV t is sensitive to the assumed value of HOLD 2 All else equal, the value of tax deferral increases with its duration. However, the top panel of Table 1 indicates that INCNPV t increases with HOLD 2 but at a decreasing rate. For expected hold ing periods longer than eight to ten years, INCNPV t is largely unaffected by increases in HOLD 2 and, in fact, for holding pe riods in excess of 16 years the value of INCNPV t begins to decrease. The premium price effect in Panel A ranges from 1.85 percent to 9.00 percent of value. These results clearly indicate that any price discounts or prem iums observed in the data are likely to vary depending on the magn itude of the taxpayers accumulated gain on the relinquished property. However, the size of the taxpayers accumulated taxable gain is not observable in the data set. The maxi mum price premiums displayed in Panel B of

PAGE 45

34 Table 1. Incremental NPV of Apartment Exchange as a Percent of Replacement Property Value ( HOLD1)( 1) 481216202428481216202428 52%1.852.142.272.302.282.232.1 72.012.352.512.552.552.512.47 56%2.002.482.692.752.712.632.5 32.252.843.103.183.163.113.04 510%2.112.743.033.103.062.952. 812.453.233.573.673.663.583.49 520%2.303.213.613.723.663.513. 312.793.904.384.534.514.414.27 102%3.454.014.264.334.294.204. 073.754.444.744.834.824.764.67 106%3.204.034.404.504.444.304. 123.644.665.115.245.225.135.00 1010%3.034.044.494.614.544.374. 153.574.805.355.515.495.375.22 1020%2.814.054.614.754.674.464. 192.793.904.384.534.514.414.27 202%5.806.807.257.367.297.136. 916.347.558.098.258.238.127.97 206%4.125.335.876.025.935.735. 464.766.256.917.117.076.946.75 2010%3.354.675.265.415.325.104. 814.065.666.376.596.556.406.20 282%7.048.298.859.008.918.718. 437.719.249.9110.1210.099.959.76 286%4.155.506.116.276.175.955. 654.876.537.267.487.447.297.08 2810%3.194.575.195.365.265.034. 723.935.626.376.606.566.406.19 52%1.331.601.731.761.741.691.6 31.481.811.962.002.001.971.92 56%1.471.942.152.212.172.101.9 91.722.302.552.632.622.562.49 510%1.582.212.492.562.522.422. 281.922.683.023.123.113.042.94 520%1.782.673.083.183.122.972. 772.263.353.843.993.963.863.72 102%2.923.483.733.793.753.663. 543.223.904.194.294.274.214.12 106%2.673.493.863.963.903.773. 583.114.124.564.694.674.584.46 1010%2.503.503.954.074.003.833. 613.044.264.804.964.944.834.67 1020%2.293.524.074.214.133.923. 652.954.455.115.315.285.144.96 202%5.286.266.716.826.756.596. 375.807.017.547.717.687.577.42 206%3.594.805.345.485.395.194. 934.235.716.366.566.536.396.21 2010%2.834.134.724.874.784.564. 273.525.125.826.046.005.865.66 282%6.527.768.328.468.388.177. 897.188.709.379.579.549.409.21 286%3.634.975.575.735.635.415. 114.345.996.716.936.896.746.54 2810%2.664.044.654.824.724.494. 193.405.085.826.056.015.865.65 52%1.461.932.202.332.392.402.38 1.662.222.522.662.722.742.72 56%1.892.693.153.383.473.493.45 2.233.183.683.934.034.054.03 510%2.243.303.90 4.214.334.354.30 2.693.954.614.945.075.115.08 520%2.844.365.21 5.655.835.865.78 3.495.276.236.696.886.936.89 102%2.773.714.244.514.624.644.59 3.174.284.875.165.285.315.28 106%3.154.545.325.725.885.915.84 3.745.386.256.686.856.896.86 1010%3.395.086.03 6.516.716.756.66 4.116.107.167.687.897.947.90 1020%3.715.786.95 7.557.797.837.73 4.607.048.348.979.249.309.24 202%4.766.427.367.848.048.077.99 5.477.438.488.999.209.259.21 206%4.326.357.508.088.328.368.26 5.197.588.869.489.749.809.75 2010%4.126.327.56 8.198.458.498.38 5.067.659.039.709.9810.059.99 282%5.857.949.129.729.9710.019.90 6.749.2110.5211.1611.4211.4911.43 286%4.596.858.138.789.049.098.97 5.568.229.6410.3310.6210.6910.63 2810%4.176.497.80 8.468.748.788.66 5.167.899.3510.0610.3510.4210.36 Panel A: = 15%, k=8%, ECt= Holding Period of Replacement Property Holding Period of Replacement Property Panel B: Panel D: Panel C: = 15%, k=10%, ECt= Panel F: Panel E: = 20%, k=10%, E C t = = 15%, k=10%, ECt= 1.2 = 15%, k=8%, ECt =1.2* = 20%, k=8%, E C t = cg 1 t SCcg 1 t SC1 t SCcg cg 1 t SCcg 1 t SCcg 1 t SC

PAGE 46

35 Table 1 are based on an increase in the assumed after-tax equity discount rate ( k ) to 10 percent from 8 percent. All ot her variables remain at thei r base case levels. Comparison of Panel A and Panel B demonstrates that a higher discount rate unambiguously increases the incremental value of the exchange option. This is because the va lue of tax deferral produced by the exchange is immediate. In contrast, the foregone depreciation deductions and the increased capital gain tax liability at sa le that results from the decreased tax basis both occur in subsequent years. Thus, the pres ent value of these future cash outflows is reduced by a higher discount rate. The net percen tage price benefit from exchange in this panel ranges from 2 percent to slightly above 10 percent. The calculated price impacts reported in Panel C of Table 1 assume the discount rate has been reset to 8 percent, but that the dollar costs of executing an exchange ( EC t ) are 20 percent higher than the co sts of a fully taxable sale ( ). As expected, higher upfront exchange costs reduce the maximum benef it of an exchange (relative to the base case in Panel A). However, decreases in the maximum price impacts are modest, averaging approximately one-half of a percen tage point across varying assumptions for HOLD 1 tSC 1 HOLD 2 and 1 The price benefits in this pa nel range from 1.33 percent to 8.46 percent. In Panel D, the dollar costs of executing an exchange ( EC t ) are assumed to be 20 percent higher than the cost s of a fully taxable sale ( ) and the discount rate is set to 10 percent. Hence, Panel D has the same assu mptions as Panel B, except for an increased cost associated with exchange. This simula tion represents a combination of higher after tax discount rate, which has immediate positive effect to the incremental NPV and increased cost of using an exchange which ha s a slight negative impact on the net benefit 1 tSC

PAGE 47

36 of using an exchange. Therefore, the values in this panel are lower than in Panel B, but higher than in Panel C. The maximum price impact is 9.57 percent for HOLD 1 = 28, HOLD 2 = 16 and 1 = 2 percent. Panel E of Table 1 reports the results assu ming a tax rate on cap ital gain income of 20 percent (the maximum statutory cg from 1999 to 2003). Clearly, the immediate value of tax deferral is larger the higher is cg However, the simulated price effects in Panel E are not uniformly higher than t hose reported in Panel A. This is because a higher capital gain tax rate will also increase the tax liability that results from the eventual sale of the replacement property. The longer the expect ed holding period of the replacement property, the more likely it is that the immediate tax deferral benefits associated with a higher cg will exceed the present valu e of the increased taxes due on the subsequent sale of the replacement property. This anticipated result is confirmed in panel E. That is, increasing cg from 15 percent to 20 percen t produces larger values of INCNPV t except in some cases where the magnitude of the deferred gain is small (i.e., when HOLD 1 and 1 are small) or when the expected holding period of the replacement property is relatively short. Panel F of Table 1 reports the results assu ming a tax rate on cap ital gain income of 20 percent and after tax discount rate of 10 percent. Values tend to be higher than in Panel B, where cg is set to 15 percent. However, si milar to Panel E, values are only higher for longer holding periods and apprecia tion rates. The difference between Panel E and Panel F is that in Panel F values increa se faster, which expre sses the added positive effect of the increased discount rate. The maximum price impact in this panel is 11.49 percent for HOLD 1 = 28, HOLD 2 = 24 and 1 = 2 percent.

PAGE 48

37 Overall, the maximum effect a Sectio n 1031 exchange is likely to have on observed transaction prices in a competitive residential market (whe n the marginal buyer and seller are not exchange motivated) is es timated to range from about 1 percent to approximately 11.5 percent. In Table 2 I repeat the simulation analysis for non-residential real estate. The only difference in this case is that non-residentia l real estate, such as office, industrial and retail properties, has a 39 year cost recovery period ( RECPER ). All other assumptions in the simulation analysis remain the same. Table 2 reveals results similar to Table 1. All el se equal, the value of tax deferral tends to increase with its duration. Therefore, with the longer depreciation recovery period of 39 years, the maximum net benefit from using an exchange will also be higher. However, with longer recovery period, be nefits from depreciation each year are smaller compared to if faster depreciation schedul e is used. Hence, values in Ta ble 2 will tend to be smaller than the corresponding values in Table 1 for small appreciation rate s. For appreciation rate 1 = 10 percent and higher, the faster de preciation effect is offset by higher appreciation and values in Table 2 become higher than the corresponding values in Table 1. I record a maximum benefit based on all simulations of 13.25 percent in Panel F for HOLD 1 = 39, HOLD 2 = 39 and 1 = 2 percent. There is a Ushaped relationship between INCNPVt and HOLD 2 At first INCNPVt increases with HOLD 2 it peaks at year 25 for Panels A though D and then it decreases at a slow rate. With Panels E and F, where the tax rate on capital gain income is set to 20 pe rcent there is a strict ly positive relationship

PAGE 49

38 between INCNPVt and HOLD 2 ; however, after year 20 INCNPVt increases at a decreasing rate. The premium price effects in Panel A va ry from 1.51 percent to 9.85 percent of value. In Panel B of Table 2 the assumed after-t ax equity discount rate (k) is increased to 10 percent. Comparison of Panel A and Panel B demonstrates the incremental value of the exchange is increased by as little as 0.14 percent, for holding periods of 5 years and appreciation rate of 25 percent, to more th an 1.3 percent, for holdi ng periods of 39 years for both the relinquished and the replacement property. The percentage price net benefit from exchange in this panel ranges from 1.65 percent to close to 11 percent. Panels C and D repeat the simulations of Panels A and B, with costs of executing an exchange ( EC t ) set to be 20 percent hi gher than the costs of a fully taxable sale ( ). Higher up-front exchange costs again redu ce the maximum benefit of an exchange. Maximum price differences in both panels ar e about 55 percentage points lower than the corresponding values in Panels A and B ( 9.30 percent in Panel C vs. 9.85 percent in Panel A, and 10.40 percent in Panel D vs. 10.96 percent in Panel C). 1 tSC Panels E and F of Table 2 report the re sults based on the same assumptions as Panels A and B, but assuming a tax rate on capital gain income of 20 percent (the maximum statutory cg from 1999 to 2003). As with the residential simulations, the simulated price effects in Panels E and F ar e not uniformly higher than those reported in Panels A and B. A higher capital gain tax ra te is also associated with increased tax liability when the taxpayer eventually sells the replacement property in an ordinary sale. With a higher capital gains tax the added benefit of using an exchange is larger for higher appreciation rates and longer holding periods. The maximum price impact in

PAGE 50

39 Table 2. Incremental NPV of Non-Residentia l Exchange as a Percent of Replacement Property Value ( HOLD1)( 1) 5102025303951020253039 52%1.511.812.022.032.021.991.651.992.212.232.232.20 56%1.872.452.842.872.852.782.132.803.223.253.243.20 510%2.152.953.503.533.513.412.523.444.014.064.043.99 520%2.653.834.644.694.654.513.194.555.405.465.445.36 102%2.853.473.893.913.893.823.133.844.284.324.314.26 106%3.084.104.794.844.814.683.544.715.455.505.495.42 1010%3.234.515.395.455.415.253.815.296.226.296.276.18 1020%3.425.056.176.246.195.994.176.047.227.307.287.16 202%4.865.986.736.786.756.615.376.657.457.517.497.41 206%4.165.696.736.806.756.564.866.617.717.797.777.66 2010%3.855.566.736.806.756.544.636.597.837.927.897.77 302%6.187.658.658.718.678.496.858.549.609.679.659.55 306%4.396.147.337.417.357.145.197.208.468.558.538.40 3010%3.835.676.937.016.956.724.676.788.118.208.188.05 392%6.928.629.789.859.809.597.699.6510.8810.9610.9410.82 396%4.316.157.417.497.437.205.157.268.598.698.668.53 3910%3.745.616.896.976.916.684.596.748.098.198.168.03 52%0.971.271.471.481.471.43 1.111.441.661.671.671.65 56%1.331.902.292.322.302.23 1.592.252.662.692.682.64 510%1.622.412.952.982.962.86 1.982.893.463.503.483.43 520%2.123.294.094.144.103.96 2.653.994.844.904.884.80 102%2.322.923.343.363.343.27 2.593.293.733.763.753.70 106%2.543.554.244.294.254.13 3.014.164.894.944.934.86 1010%2.693.974.844.904.854.70 3.284.745.665.735.715.62 1020%2.894.515.615.695.635.44 3.635.496.666.746.726.60 202%4.335.436.186.236.206.06 4.846.106.896.956.936.85 206%3.635.156.186.246.206.01 4.326.067.157.237.217.10 2010%3.325.026.186.256.205.99 4.096.047.277.367.337.21 302%5.657.108.108.168.127.94 6.317.999.049.119.098.99 306%3.855.596.786.866.806.59 4.656.657.907.997.977.84 3010%3.305.136.376.456.396.17 4.136.237.557.647.627.49 392%6.388.089.239.309.259.04 7.169.1010.3210.4010.3810.26 396%3.785.616.866.936.886.65 4.616.718.038.138.107.97 3910%3.205.076.336.426.366.13 4.056.197.537.637.607.47 52%1.291.772.172.242.272.27 1.482.012.422.482.502.51 56%1.962.873.633.763.813.82 2.313.334.114.224.274.27 510%2.493.744.804.985.055.05 2.984.395.455.615.675.68 520%3.425.266.827.087.197.20 4.146.217.798.028.118.12 102%2.523.484.294.434.484.49 2.893.974.804.924.964.97 106%3.344.946.296.516.606.61 3.975.767.127.327.407.41 1010%3.895.907.607.898.008.01 4.686.938.658.909.009.02 1020%4.607.159.319.679.819.82 5.608.4610.6310.9611.0811.10 202%4.406.137.607.857.957.96 5.087.038.518.738.818.83 206%4.787.169.189.519.659.66 5.718.3910.4210.7210.8310.86 2010%4.957.629.8910.2610.4210.43 6.009.0011.2811.6211.7511.77 302%5.697.989.9310.2510.3810.39 6.599.1711.1211.4111.5211.54 306%5.298.0310.3410.7310.8810.90 6.369.4311.7712.1112.2512.27 3010%5.178.0410.4710.8811.0411.05 6.299.5211.9712.3312.4712.50 392%6.479.1311.3811.7511.9011.92 7.5110.5012.7613.1013.2313.25 396%5.418.2910.7211.1311.2911.30 6.549.7712.2212.5812.7212.75 3910%5.188.1010.5810.9911.1611.17 6.329.6112.1012.4712.6112.64 = 15%, k=8%, ECt= = 15%, k=10%, ECt= = 15%, k=8%, ECt=1.2* = 15%, k=10%, ECt= 1.2* = 20%, k=8%, ECt= = 20%, k=10%, ECt= Panel C: Panel A: Holding Period of Replacement Property Panel B: Panel D: Panel E: Panel F: Holding Period of Replacement Property cg1 t SCcg1 t SC1 t SCcg1 t SC1 t SC1 t SCcg1 t SCcg1 t SC 1 t SCcg1 t SC

PAGE 51

40 Panel E is 11.92 percent for HOLD 1 = 39, HOLD 2 = 39 and 1 = 2 percent, while the maximum benefit from exchange in Panel F is 13.25 percent for HOLD 1 = 39, HOLD 2 = 39 and 1 = 2 percent. To summarize the results in Table 2, an increase in the after ta x-cost of equity by two percent, all else the same, is associated with a maximum increase in the incremental NPV from completing an exchange of slightly over 1.3 percent. Increase, in the dollar costs associated with an exchange, relative to the cost of comple ting a taxable sale, is associated with approximately a 0.5 percent de crease in NPV, all else the same, in the worst case. A capital gain tax rate of 20 percent, all else the same is associated with, at most, approximately a 4.5 percent higher allowed price premium. Maximum price benefits vary between 8.5 and 11.5 percen t for apartments and between 9.3 and 13.3 percent for non-residential real estate. The next chapter describes the data and empirical methodology used to measure the size of exchange premiums and discounts actu ally observed in commercial real estate markets over the 1999 to 2005 period.

PAGE 52

CHAPTER 6 DATA AND METHODOLOGY Data Property sales data is obtained from CoStar Group, Inc. The CoStar Comps Professional database includes historical in formation on over 1.2 million confirmed commercial real estate tran sactions from 1999 through the first half of 2005. The CoStar database includes all sales in excess of $250,000 in more than 40 major U.S. markets. 1 Land, mobile homes, and special use propertie s are excluded from this analysis. The initial sample contains 270,415 confirmed sa les in five property markets: office, industrial, apartment, retail and hotel/motel. CoStar Comps Professional has a separate attribute field that identifies whether the sa le represents an exchange. Therefore, any missing exchange identification is due to a l ack of information, rather than CoStars failure to report the type of transaction. An observation is eliminated from the sample if it could not be determined whether the sale was a part of a Section 1031 exchange. This further reduces the sample to 158,196 observations. CoStar Comps Professional also contains descriptive information on the type of exchange (e.g. taxpayers sale of relinquished property, si multaneous exchange, reverse exchange, etc.) in detailed notes. Based on the manual inspection of these notes, each exchange property sale is placed in to one of the following categories: 1. Sellers relinquished property in delayed (Starker) exchange 2. Buyers replacement property in delayed (Starker) exchange 1 The CoStar product used in this analysis is CoStar Comps Professional (www.costar.com/products/comps/). 41

PAGE 53

42 3. Sellers relinquished and buyers replacement property in two separate transactions 4. Direct exchange (swap) 5. Sellers relinquished propert y in reverse exchange 6. Buyers replacement property in reverse exchange 7. Reverse exchange (type not confirmed) 8. Exchange into Tenants-in-Common 9. Other tax-deferred exchange exchange wh ich cannot be categorized in any of the above types or where CoStar wa s unable to confirm its type. To assure reliability of the data, CoStar requires agents to physically inspect the site and record a variety of property characteristics and tran saction details. I therefore exclude sales not confirmed by CoStar. In addition, I exclude all transactions with recorded sales price below $250,000. CoStar covers comprehensively only transactions that are above this threshold, although in some cases brokers do report smaller transactions. For the sake of consistency such smaller transactions are eliminated from the sample. The final sample has 124,830 transa ctions which facilitates a comprehensive empirical investigation of th e Section 1031 exchange market, as well as other atypical motivations, not possible with previous data sets. Of the 124,830 usable sales transactions, 23,989 (or 19 percent) involved the use of a Section 1031 exchange. Table 3 summarizes the number of transact ions by year and pr operty type. The year 1999 contains fewer transactions than year 200 0. Also in 2005 there is only data for half of the year through June 2005. A substantial increase in the use of excha nges in commercial real estate markets has been discussed in the popular press (e.g., McLinden, 2004). However, the data necessary to support a comprehensive analysis has not been availa ble. Table 3 breaks down the sample by exchange and non-excha nge transactions. The table reveals that exchanges as percentage of all sales are ve ry stable over the 1 999-2005 sample period. For example, between 1999 and 2000, the tota l number of commercial real estate

PAGE 54

43 Table 3. Description of Size of Exchange Market Property Type 1999 2000 2001 2002 2003 2004 Jun-05 Total Exchange 587 2011 1774 2089 2096 2016 636 11209 Apartment Nonexchange 1291 4754 4505 4683 4728 4996 1381 26338 All 1878 6765 6279 6772 6824 7012 2017 37547 Exchange 222 757 640 630 671 749 231 3900 Industrial Nonexchange 1172 4248 4210 4437 4469 4711 1308 24555 All 1394 5005 4850 5067 5140 5460 1539 28455 Exchange 180 675 587 575 631 671 241 3560 Office Nonexchange 1028 3512 3430 3388 3519 3993 1074 19944 All 1208 4187 4017 3963 4150 4664 1315 23504 Exchange 252 904 755 897 1023 958 294 5083 Retail Nonexchange 1372 4836 4839 5300 5389 5186 1249 28171 All 1624 5740 5594 6197 6412 6144 1543 33254 Exchange 16 47 41 37 43 43 10 237 Hotel/ Motel Nonexchange 88 338 328 301 334 359 85 1833 All 104 385 369 338 377 402 95 2070 Exchange 1257 4394 3797 4228 4464 4437 1412 23989 Total Nonexchange 4951 17688 17312 18109 18439 19245 5097 100841 All 6208 22082 21109 22337 22903 23682 6509 124830

PAGE 55

44 exchanges grew from 1,257 to 4,394; however the exchanges share of all verified CoStar transactions remained in the 18-20 percent range. Notably, the year of 2003, which is the year in which the maximum statutory capital gain rate was decreased from 20 percent to 15 percent, was not associated with a decrease in the number of exchanges. In fact, the number of exchanges increased from 4,228 in the previous year to 4,464. A decrease in the maximum capital gain rate as the simulati on analysis suggests can possibly make an exchange a less attractive alternative if th e tax deferral is the main motivation to participate in such a transaction. Inspection of Table 3 reveals that exchange s are more frequently used in apartment markets. Of the 37,547 verified apartmen t transactions in the sample, 11,209 (approximately 30 percent) involved the use of an exchange. Moreover, this percentage has remained remarkably stable during 1999 2005. Among other property types, exchanges generally account for 10-18 percent of all transactions. Tables 4 through 6 present breakdowns of property sales for all 46 markets covered by CoStar for apartment, office and retail properties, respectively. Markets are sorted alphabetically to make it easier for the reader to locate his or her market of interest. The number of transactions for each market a nd the percentage of total transactions by property type are presented. The tables show that there is a substantial variability in terms of transactions observed in different markets. Apartments Table 4 reveals that the largest apartment market is Los Angele s with 25 percent of all sales. New York City is the second larges t market with 12 percent of all transactions. The smallest market (Charlo tte) contains only 12 usable sales observations. The table reveals that relatively few mark ets account for a major share of all apartment sales. For

PAGE 56

45 example, more than half of a ll transactions are concentrated in less than 10 markets and the top 20 markets account for 86 percent of all sales. Table 4 also contains the percentage of exchanges observed in each market, as well as the breakdown of the type of exchanges observed: relinquished, replacement, relinquished and replacement, direct swaps, a nd other types. Table 4 clearly shows that the use of exchanges varies substantially ac ross the major markets. Importantly, in eight markets (Marin-North SF Bay Area, Portland, Reno, Sacramento, San Diego, San Francisco, San Jose and Seattle) exchange s represent the dominant form of property transaction. Interestingly, all of these markets are located in the Western U.S., and most are in California. This is consistent with anecdotal evidence that exchanges are more common on the West Coast. One factor cited for the wide r use of Section 1031 exchanges in the Western US is that exchanges are related to the real esta te booms in the West in the 1960s, 1970s and 1980s, which made investors more entrepre neurial (McLinden, 2004). Another possible explanation is the rapid appreciation of real estate in major metropolitan areas in California in the last few year s. Anecdotal evidence suggests that homes have appreciated at an annual rate of 20 percent in Southern California over the la st five years. In the other major markets tracked by CoStar, exchange s have been much less prominent. For example, over the 1999-2005 time period, only 1.6 percent of apartment sales in New York involved an exchange. The distribution of exchange types of s hows that the number of relinquished and replacement exchanges by markets is quite si milar. This is expected, since for each exchange involving a relinquished property, there should be at least one replacement

PAGE 57

46 Table 4. Description of Apartment Property Sales by Markets Name Apts. All % of Total Apt. Exch. % Exch. Reliq. Exch. Repl. Exch. Both Direct Other Los Angeles 9450 25.17% 3574 37.82% 1139 1348 514 24 549 New York City 4586 12.21% 73 1.59% 20 27 5 11 10 San Diego 2448 6.52% 1419 57.97% 477 490 360 6 86 Chicago 1991 5.30% 377 18.94% 162 114 41 18 42 Seattle 1686 4.49% 900 53.38% 207 349 99 3 242 Phoenix 1442 3.84% 143 9.92% 31 68 17 0 27 Oakland 1398 3.72% 645 46.14% 256 200 152 1 36 Miami 1102 2.93% 81 7.35% 30 42 4 1 4 San Francisco 1029 2.74% 521 50.63% 275 137 94 4 11 Denver 1002 2.67% 473 47.21% 141 189 106 5 32 Portland 748 1.99% 516 68.98% 87 163 102 1 163 Ft. Lauderdale 733 1.95% 85 11.60% 22 46 5 3 9 Riverside/San Bernardino 712 1.90% 283 39.75% 77 108 66 3 29 Tucson 615 1.64% 212 34.47% 34 120 51 3 4 Washington, DC 579 1.54% 31 5.35% 10 10 1 3 7 San Jose 569 1.52% 287 50.44% 122 83 63 4 15 Dallas/Fort Worth 546 1.45% 129 23.63% 22 43 8 2 54 Colorado Springs 527 1.40% 38 7.21% 22 8 5 2 1 Boston 506 1.35% 64 12.65% 20 38 1 1 4 Tampa 497 1.32% 37 7.44% 10 20 2 2 3 Sacramento 421 1.12% 261 62.00% 71 104 73 3 10 Cincinnati/Dayton 408 1.09% 53 12.99% 21 24 4 1 3 Atlanta 398 1.06% 24 6.03% 6 8 0 5 5 Detroit/Toledo 350 0.93% 44 12.57% 24 12 0 6 2 Las Vegas 336 0.89% 125 37.20% 22 82 12 0 9 Fresno 316 0.84% 107 33.86% 23 54 13 1 16 Cleveland/Akron 297 0.79% 20 6.73% 14 5 1 0 0 North SF Bay Area 296 0.79% 183 61.82% 67 55 50 1 10 Houston 287 0.76% 55 19.16% 12 34 2 2 5 West Palm Beach 277 0.74% 27 9.75% 7 13 4 0 3 Stockton/Modesto 274 0.73% 121 44.16% 44 35 29 3 10 New Jersey 249 0.66% 8 3.21% 3 4 0 0 1 Orlando 236 0.63% 16 6.78% 2 9 1 0 4 Austin 227 0.60% 32 14.10% 10 7 4 1 10 Philadelphia 201 0.54% 19 9.45% 14 1 0 0 4 Minneapolis 170 0.45% 65 38.24% 21 26 10 2 6 Baltimore 132 0.35% 13 9.85% 2 10 1 0 0 Columbus 126 0.34% 55 43.65% 18 24 11 0 2 Jacksonville 125 0.33% 7 5.60% 0 3 1 2 1 Ventura 106 0.28% 44 41.51% 18 11 11 1 3 St Louis 46 0.12% 10 21.74% 5 2 1 1 1 Kansas City 40 0.11% 10 25.00% 5 5 0 0 0 Salt Lake City 20 0.05% 7 35.00% 3 2 2 0 0 San Antonio 16 0.04% 3 18.75% 0 1 0 0 2 Reno 15 0.04% 8 53.33% 1 6 1 0 0 Charlotte 12 0.03% 4 33.33% 0 3 1 0 0

PAGE 58

47 exchange. However, if the taxpayer fails to identify replacement properties in a timely fashion, or close the exchange according to the guidelines issued by the IRS, no replacement exchange takes place. Also, it is common for the replacement property to be of a different type than th e original (relinquished pr operty). Finally, replacement properties may also be located in different markets than the original properties. Table 4 shows that direct exchanges ar e quite rare. Out of 9,450 sales in Los Angeles only 24 transactions represented direct swaps. This is also the largest number of direct exchanges recorded for any market. The small number of direct swaps is not surprising given how difficul t it is to match the require ments of both taxpayers and complete a direct swap of properties. For the purposes of the regression analysis th at follows in a later chapter, I exclude all direct exchanges to eliminate the possible bias that their inclusion may induce in the model. Other exchanges, which mostly in clude exchanges whose type has not been confirmed, can introduce similar problems as with direct exchanges; hence, these transactions are also eliminated from the samp le. I also exclude sales that are associated with any other special condition that is not of interest for this study. Examples of such special conditions are sales that are part of an auction, bank ruptcy or sales that involve building contamination, natural disaster damage, or the threat of cont amination. In total, CoStar delineates more than 30 such unusual conditions, which potentially have an effect on observed transaction prices. I therefore eliminate all obser vations that contain a sale condition which is not analyzed by this study. Th e only special conditio ns that I allow are related to condominium conversions and portfol io sales. Since, as noted previously, the number of sales in the top 20 apartment markets accounts for 86 percent of all apartment

PAGE 59

48 sales, I focus in the empirical analysis on property sales in th ese markets only. In addition, I exclude markets for which there is not a sufficient number of exchanges to generate meaningful statistical tests. These excluded markets are Miami, Ft. Lauderdale, Washington, DC, Dallas, and Colorado Springs. This leaves us with 15 markets, which represent 75 percent of a ll apartment sales. The final sample has 23,640 apartment transactions over the 1999-2005 time period. Office Properties Table 5 presents information on office property sales by markets and shows that in this sample the largest office market is Los Angeles with 8.5 percent of all sales. Phoenix is the second largest market with 7.2 percent of all transactions, while Washington, DC is the third largest market with 6.8 percent of all office transactions. The smallest market (San Antonio) contains only 22 sales observations. The table reveals a different picture than with apartments. First, there is less concentration of transactions in the largest markets. Consequen tly, the top 20 markets account for 75 percent of all sales. Second, th e distribution of exchanges across markets is also quite different. Although the use of exchanges varies substantially across major metropolitan areas, in none of the markets do exchanges represent the dominant form of property transaction. 2 The observation that markets located in the Western U.S. have the highest number of exchanges remains unchange d. The distribution of relinquished and replacement types of exchanges also varies across the sample. In approximately one half of the markets, replacement exchanges are either equal to or as much as twice the number of relinquished exchanges. In some markets, sharp contra sts are observed. For example, 2 Salt Lake City is the only market that has a share of exchanges that is above 50 percent, but this market can be ignored since it only has 26 transactions. Therefore, percentages may be biased due to the small sample.

PAGE 60

49 Table 5. Description of Office Property Sales by Markets Name Office All % of Total Office Exch. % Exch. Reliq. Exch. Repl. Exch. Relq. & Repl. Exch. Direct Exch. Other Exch. Los Angeles 2001 8.51% 446 22.29% 158 202 33 3 50 Phoenix 1683 7.16% 126 7.49% 12 84 9 0 21 Washington D.C. 1600 6.81% 120 7.50% 30 57 6 10 17 Chicago 1163 4.95% 116 9.97% 29 59 4 11 13 Atlanta 1073 4.57% 36 3.36% 6 16 0 8 6 Seattle 1039 4.42% 340 32.72% 111 149 28 3 49 Denver 902 3.84% 241 26.72% 64 128 34 4 11 San Diego 858 3.65% 282 32.87% 63 152 43 2 22 New York City 835 3.55% 42 5.03% 14 13 2 6 7 Tampa 714 3.04% 49 6.86% 14 24 3 7 1 New Jersey 698 2.97% 34 4.87% 6 13 2 10 3 Detroit/Toledo 639 2.72% 46 7.20% 12 20 0 9 5 Miami 628 2.67% 28 4.46% 7 13 5 2 1 Philadelphia 625 2.66% 30 4.80% 11 10 1 3 5 Ft. Lauderdale 580 2.47% 41 7.07% 8 25 2 3 3 Orlando 571 2.43% 26 4.55% 5 13 1 4 3 Dallas/Fort Worth 540 2.30% 68 12.59% 9 39 4 0 16 Boston 506 2.15% 19 3.75% 4 7 0 3 5 Oakland 495 2.11% 167 33.74% 49 71 34 0 13 Baltimore 430 1.83% 25 5.81% 7 16 0 0 2 Colorado Springs 416 1.77% 38 9.13% 10 17 4 4 3 West Palm Beach 402 1.71% 33 8.21% 7 23 0 0 3 Portland 395 1.68% 159 40.25% 61 57 12 0 29 Sacramento 394 1.68% 159 40.36% 26 89 38 1 5 Las Vegas 392 1.67% 117 29.85% 15 80 15 0 7 Tucson 379 1.61% 60 15.83% 11 36 8 4 1 Riverside/San Bernardino 373 1.59% 142 38.07% 28 80 22 1 11 Houston 344 1.46% 39 11.34% 7 30 0 0 2 Cleveland/Akron 299 1.27% 20 6.69% 3 6 3 3 5 San Jose 277 1.18% 65 23.47% 23 27 10 0 5 Cincinnati/Dayton 275 1.17% 9 3.27% 1 6 1 1 0 Austin 266 1.13% 19 7.14% 1 11 1 0 6 North SF Bay Area 261 1.11% 102 39.08% 31 51 18 1 1 Jacksonville 249 1.06% 7 2.81% 2 5 0 0 0 San Francisco 212 0.90% 69 32.55% 34 25 7 1 2 Fresno 200 0.85% 59 29.50% 9 33 9 0 8 Columbus 200 0.85% 63 31.50% 19 36 6 2 0 Stockton/Modesto 136 0.58% 31 22.79% 6 17 4 0 4 Minneapolis 104 0.44% 16 15.38% 1 9 2 3 1 Ventura 94 0.40% 18 19.15% 3 7 3 0 5 St Louis 65 0.28% 7 10.77% 1 5 0 1 0 Kansas City 61 0.26% 11 18.03% 3 7 0 0 1 Charlotte 50 0.21% 8 16.00% 2 4 0 0 2 Reno 32 0.14% 9 28.13% 1 6 2 0 0 Salt Lake City 26 0.11% 14 53.85% 5 7 1 0 1 San Antonio 22 0.09% 4 18.18% 2 2 0 0 0

PAGE 61

50 replacement exchanges in Phoenix outnumber re linquished exchanges by a factor of 7. In Las Vegas, replacement exchanges are 5 times more frequent than relinquished property sales and in Sacramento this ratio is e qual to 3.4. A possible explanation for such differences is the combination of several properties used to complete a replacement exchange, as well as replacement property sales being completed outside of the particular market or with a different property type. Tabl e 5 reveals that direct exchanges are also rare for office properties. The largest numb er of swaps is observed in Chicago, where only 11 transactions represent direct exchanges. As with apartments, I exclude all direct and other types of exchanges from the sample in the regression analysis. I also ex clude transactions asso ciated with special conditions that are not sale-leasebacks or portf olio sales. As noted previously, the office sales in the top 20 office markets represent 75 percent of all office transactions. However, after elimination from the sample of sales wi th the characteristics described above, eight markets out of the largest 20 markets do not have a sufficient number of exchanges to generate statistically meaningful tests. Th ese markets are: Atlanta, New York City, Northern New Jersey, Detroit, Miami, Philadelphia, Orlando, Boston and Baltimore. I include in the sample the next largest markets that have a sufficien t number of exchanges: Sacramento, Las Vegas, Tucson, and Riverside/San Bernardino. Therefore, in the empirical analysis I focus on the office propert y sales in the largest 15 markets that also have a sufficient number of replacement and relinquished exchanges to generate any results that are statistically meaningful. The sales in these 15 markets represent 56 percent of the office real estate market. The final sample has 8,871 office transactions during the 1999-2005 time period.

PAGE 62

51 Retail Properties Table 6 presents a breakdown of retail prop erty sales by markets. New York City with 3,669 retail sales (11 per cent of all sales) is the la rgest market, followed by Los Angeles with 3,235 retail transactions (9.7 percent of all sales), and Chicago with 2,796 observations (8.4 percent of all sales). The sm allest retail market is Reno, which only has 9 sales recorded. Similarly to the office sales, there is less concentration of transactions in the largest markets. The top 20 markets account for 77 percent of all sales. The use of exchanges in retail transactions also varies substantially across markets. Similarly to the office transactions, tax defe rred exchanges are not as popular as with apartment sales. In none of the 46 CoStar markets do exchanges outnumber nonexchanges. The largest percentage of excha nges is observed in Portland, where this type of transaction represents more than 40 percent of all sales. 3 Once again exchanges are more frequently observed in the Western United States. There is substantial variation in the di stribution of relinquish ed and replacement types of exchanges across the sample. In 10 of the retail markets relinquished exchanges outnumber the replacement exchanges. In mo re than 50 percent of the sample the replacement exchanges are less than twice as frequent as relinquished exchanges. Finally, the replacement exchanges in 10 markets are more than 3 times higher than the relinquished exchanges 4 Notable markets, in which replacement exchanges significantly outnumber relinquished exchanges include Tu cson, with a ratio of replacement to relinquished property sales equal to 5.6. La s Vegas has a ratio of 5.3, Riverside 4.9, Dallas /Forth Worth 4.2, and Houston 3.8. Table 6 illustrates also the rareness of 3 In Reno 44 percent of all sales are exchanges; however this percentage is not reliable since it is based on only 9 observations. 4 Reno and San Antonio are not included in this analysis, based on their small number of observations.

PAGE 63

52 Table 6. Description of Reta il Property Sales by Markets Name Retail All % of Total Retail Exch. % Exch. Reliq. Exch. Repl. Exch. Relq. & Repl. Exch. Direct Exch. Other Exch. New York City 3669 11.03% 60 1.64% 8 20 4 12 16 Los Angeles 3235 9.73% 775 23.96% 247 360 61 12 95 Chicago 2796 8.41% 381 13.63% 138 158 30 29 26 Seattle 1546 4.65% 470 30.40% 127 228 40 11 64 Phoenix 1443 4.34% 145 10.05% 22 66 8 5 44 Atlanta 1323 3.98% 90 6.80% 11 37 7 7 28 Detroit/Toledo 1068 3.21% 61 5.71% 24 19 2 7 9 Denver 1035 3.11% 299 28.89% 82 158 43 3 13 Tampa 1017 3.06% 43 4.23% 4 27 3 3 6 Dallas/Fort Worth 966 2.90% 165 17.08% 19 79 8 2 57 San Diego 918 2.76% 317 34.53% 82 163 47 2 23 Boston 853 2.57% 50 5.86% 14 21 2 3 10 Washington, DC 841 2.53% 43 5.11% 16 11 3 2 11 Miami 819 2.46% 41 5.01% 10 20 3 2 6 Orlando 799 2.40% 35 4.38% 6 20 2 3 4 Oakland 739 2.22% 192 25.98% 63 91 27 0 11 Ft. Lauderdale 733 2.20% 47 6.41% 14 23 5 4 1 Houston 661 1.99% 121 18.31% 21 80 8 1 11 Tucson 584 1.76% 106 18.15% 14 79 11 2 0 Riverside/San Bernardino 578 1.74% 207 35.81% 27 133 30 3 14 Las Vegas 561 1.69% 199 35.47% 27 144 18 1 9 West Palm Beach 551 1.66% 43 7.80% 12 26 2 2 1 Portland 535 1.61% 220 41.12% 67 95 19 4 35 Philadelphia 520 1.56% 34 6.54% 4 12 0 4 14 San Francisco 492 1.48% 175 35.57% 81 56 35 1 2 Cleveland/Akron 474 1.43% 27 5.70% 6 13 0 5 3 New Jersey 461 1.39% 14 3.04% 3 6 0 2 3 Cincinnati/Dayton 454 1.37% 34 7.49% 4 24 2 1 3 Colorado Springs 444 1.34% 39 8.78% 13 22 2 2 0 Baltimore 413 1.24% 22 5.33% 11 5 3 1 2 North SF Bay Area 356 1.07% 118 33.15% 38 58 17 2 3 Sacramento 308 0.93% 117 37.99% 26 68 22 0 1 Fresno 291 0.88% 40 13.75% 10 19 2 1 8 Jacksonville 285 0.86% 14 4.91% 3 8 0 1 2 San Jose 274 0.82% 80 29.20% 30 29 15 1 5 Columbus 256 0.77% 91 35.55% 23 56 11 0 1 Austin 246 0.74% 28 11.38% 12 9 3 0 4 Stockton/Modesto 180 0.54% 35 19.44% 10 9 4 4 8 Minneapolis 146 0.44% 22 15.07% 9 6 2 4 1 Ventura 95 0.29% 33 34.74% 8 15 8 1 1 St Louis 95 0.29% 11 11.58% 3 5 0 2 1 Kansas City 64 0.19% 6 9.38% 1 5 0 0 0 Charlotte 52 0.16% 10 19.23% 7 3 0 0 0 Salt Lake City 49 0.15% 12 24.49% 6 5 1 0 0 San Antonio 20 0.06% 7 35.00% 0 7 0 0 0 Reno 9 0.03% 4 44.44% 0 3 1 0 0

PAGE 64

53 direct exchanges in retail sales. Once ag ain, the largest number of direct swaps is observed in Chicago, where a total of 29 such transactions occurred. As with apartments and office properties, I exclude all direct exchanges and other types of exchanges from the regression sa mple. I also eliminate all transactions associated with other special conditions that are not condominium conversions, saleleasebacks, or portfolio sales. The number of sales in th e top 20 retail markets accounts for 77 percent of all retail sales. However, after elimination of th e above observations, seven of the largest 20 markets do not have a sufficient number of exchanges to generate statistical meaningful tests. These markets include New York City, Atlanta, Tampa, Boston, Washington, DC, Miami and Orlando. Theref ore, in order to have a sample of 15 markets I include the next largest market s that have also a sufficient number of exchanges. These are Las Vegas and San Fran cisco. Hence, in the empirical analysis I focus on the property sales in the largest 15 ma rkets that also have a sufficient number of replacement and relinquished exchanges to ge nerate any results th at are statistically meaningful. The sales in these 15 markets repr esent 52 percent of the retail real estate market. This yields a final sample of 12,015 office transactions over the 1999-2005 time period. Research Methodology I use standard hedonic regression techniques to assess the influence of tax-deferred exchanges, as well as other sale conditions on observed transaction prices. Standard hedonic models include the log of the transaction price or rent as the dependent variable and a set of independent variables that capture the site, structural, and location characteristics of the property. For exampl e, Frew and Jud (2003) regress the observed sale price on a number of independent variab les, including the square footage of rental

PAGE 65

54 space, land area, age, and number of units. In an apartment rent prediction model based on 4,500 apartment complexes in eight markets, Valente et al. (2005) use the log of asking rent as the dependent variable. For re gressors they use square footage per unit, number of floors, property age, submar ket dummies, and year of sale dummies. Building age, age squared, the square footage of improvements, building foot print, lot size, number of units, and number of floors, are some of the most common structural characteristics used in commercial property price or rent equations (see, for example, Colwell, Munneke and Trefzger 1998, and Saderion, Smith and Smith, 1993). The choice of functional form is also very impor tant in order to ensure that the model is correctly specified. Weirick and Ingram (1990) provide an excellent analysis of various approaches to functional forms in hedonic re gressions, when the dependent variable is selling price. In particular, the authors compare three standard functional forms: A linear model A semi-log model which uses the logarithmic transform of the dependent variable (selling price) A log-linear model, which uses logarith mic transforms of both the dependent variable as well as independent variables As Weirick and Ingram (Ibid.) point out, the linear form has serious deficiencies from a market theory standpoint. Such models force the value of an extra square foot of improvement for a 2,000 sq. ft. property to be th e same as the value of an extra square foot for a 10,000 sq. ft. property. The semi-log a nd log-linear models take into account nonlinearities in the data. In addition, by usi ng quadratic transformations of explanatory variables (such as square footage and lot si ze) I can capture property value relationships that are concave or convex in cer tain characteristics (Ibid.).

PAGE 66

55 Recently, the focus of the residential hedoni c pricing literature has shifted to the proper control of location. One of the most impo rtant papers in this area is Clapp (2003), who presents a semi-parametric method for valuing residential location and includes latitude and longitude as expl anatory variables. Case, Cla pp, Dubin and Rodriguez (2004) use a second order longitude-l atitude expansion to contro l for location, as well as a number of demographic characteristics, such as percent Black, Hispanic, etc. Finally, by using geographic coordinates, Fik, Ling and Mulligan (2003) fully account not only for the absolute location of th e home, but also for relative location in a metropolitan market. The authors use a complete variable interactive approach to model the log of sale price as a function of structural characteris tics, discrete location dummies, location dummies and {x, y} coordinates intera cted with structural characteristics, interaction terms between structural characteri stics (e.g. age*sq. ft.), interacted structural characteristics interacted with location dummies and triple interaction terms of location dummies, geographic coordinate s and structural characteristics. This fully interactive specification allows Fik et al. (2003) to effectively estimate separate price surfaces for identified sub-markets, rather than constrai n the estimated coefficients on structural characteristics to be constant across submarke ts with the price surf ace shifted up or down by location dummies only. In ad dition, interacting sub-market dummies with absolute location allows Fik et al. (2003) to capture discontinuities or structural shifts that occur as the price surface crosses submarket boundaries. Table 7 provides definitions for all variab les used in the hedonic regressions. The dependent variable is LNPRICE the natural logarithm of the sale price. An advantage of using the log of price is that less weight is given to extr eme values than when using

PAGE 67

56 untransformed prices. I also divide square footage of improvement and land square footage by 1000 to keep size ranges consistent. Following Weirick and Ingram (1990), I use a semi-log model with quadratic transforms for square footage in thousands ( SQFT ), and land square footage in thousands ( LANDSQFT ). With this semi-log form, unit price per unit change in the characteristic is given by simply multiplying the estimated coefficient by the observed selling price. To quantify the effect of Section 1031 exchanges, condominium conversions, saleleasebacks, portfolio sales, and out-of-stat e buyers on sale prices, I use a stepwise estimation technique and estimate the follo wing model separately for each of the identified 15 markets in the apartment sales sample )17( 2 2 2 _43 2 2005 2000 18 17 16 15 14 13 3 2 12 11 10 9 8 7 6 5 4 3 2 10 P s s s n nn i i i mSMDUM YR PORTSALE CONDOCONV CONDO SUBSIDIZED SENIOR BUYEROUT CONDITION UNITS FLOORS PARKING LANDSQFT LANDSQFT SQFT SQFT AGE AGE REPL RELQ EXRELQ EXREPL LNPRICE I use a similar model to estimate the effect of tax-deferred exchanges, condominium conversions, sale-leasebacks, por tfolio sales, and out-of-state buyers on property sale prices in offi ces and retail transactions )18( 2 2 2 _43 2 2005 2000 14 13 12 3 2 11 10 9 8 7 6 5 4 3 2 10 P s s s n nn i i i mSMDUM YR PORTSALE ACK SALELEASEB BUYEROUT CONDITION FLOORS PARKING LANDSQFT LANDSQFT SQFT SQFT AGE AGE REPL RELQ EXRELQ EXREPL LNPRICE In the estimation of the hedonic pricing equation, I include a dummy variable (EXRELQ) that quantifies the extent to which transaction prices ar e lower (higher), all else equal, if the seller of the property is a taxpayer initiating the downleg portion of a

PAGE 68

57 delayed exchange (i.e., if the property is a relinquished propert y). I also include a dummy variable (EXREPL) that quantifies th e extent to which transaction prices are higher, all else equal, if the buyer of the property is a taxpayer completing the upleg portion of a delayed exchange (i.e., if the sa mple property is a replacement property). Finally, I include a dummy variable ( RELQ_REPL) that quantifies the extent to which prices are higher, all else equal, if the sa me property is used as both the relinquished property for the seller and the replacement property for the buyer. The first exchange involves the front end of a 1031 transaction for the seller, who now needs to find a replacement property in order to complete the exchange. The second exchange involves the back end of an exchange in which the ta xpayer has already sold his property and uses this sale to acquire a replacement propert y thereby completing his own exchange. Lambson, McQueen and Slade (2004) find evidence that out-of state buyers pay price premiums for apartment complexes in Phoenix, which the authors associate with possible higher search costs and anchor ing. I therefore use a dummy variable BUYEROUT to control for any effects related to whether the buyers pr incipal residence is out-of-state. Two dummy variables ( CONDOCONV and PORTSALE) are also used in the apartment regressions to quantify the extent to which prices are higher, all else equal, when the apartments are purchased with the in tention to convert them to condos, or when the transaction is part of a portfolio sale. In the office and retail regr essions, dummy variables ( SALELEASEBACK and PORTSALE ) are used to quantify the price premiu m associated with sale-leasebacks or portfolio sales.

PAGE 69

58 Table 7. List of Regression Variables Dependent Variable: LNPRICE Natural logarithm of the sale price Exchange Variables: EXRELQ Binary variable set equal to one if transaction represents sale of a relinquished property EXREPL Binary variable set equal to one if transaction represents purchase of a replacement property RELQ_REPL Binary variable set equal to one if transaction represents both sale of a relinquished property and purchase of a replacement property Building Characteristics : AGE Age of the building(s) in years SQFT Total improvements square footage in thousands FLOORS Number of floors UNITS Number of units CONDITION i Physical condition of the propert y based on inspection. The categories include below average, average, and above average. The omitted category is average. PARKING Number of parking spaces SUBSIDIZED Binary variable set equal to one if property use is subsidized multi-family SENIOR Binary variable set equal to one if property use is senior multi-family CONDO Binary variable set equal to one if property use is multi-family condominium Year Dummies: YR i Yearly time periods from 1999 thr ough 2005. Each year is included as a binary variable except 1999, which is suppressed. Site and Location Characteristics : LANDSQFT Log of s quare footage of land in thousands X Latitude of property Y Longitude of property SDUM i Binary variable signifying the subm arket in which the property is located, as defined by CoStar {X k ,Y l } X,Y polynomial, where k,l >0, k+l<=3 Deal Characteristics: BUYEROUT Binary variable set equal to one if buyer lives out of state CONDOCONV Binary variable set equal to one if the transaction was motivated by condoconversion SALELEASEBACK Binary variable set equal to one if transaction was part of sale-leaseback PORTSALE Binary variable set equal to one if tr ansaction was part of portfolio sale

PAGE 70

59 I use several variab les to account for the relations hip between selling price and property structural characteris tics. First, I expect a negative relation between age ( AGE) and price and a positive coefficient on age squared ( AGE2 ). This expectation reflects the frequently observed quadratic relation betw een price and age. A vintage effect is sometimes observed, which is related to high pr ices for very old properties. I expect the coefficients on SQFT, LANDSQFT, PARKING, FLOORS and UNITS to be positive. The variable CONDITION i controls for building condition. I specify average condition as the control group. With residential real estate, 79 percen t of apartments are reported to be in average condition, 14 percent of the apartment properties are categorized by CoStar as being in above av erage condition, and 7 pe rcent are labeled as below average. In the office sample, 66 percent of the properties are classified to be in average condition, 32 percent are above average, and only 3 pe rcent are in below average condition. Finally, with retail properties 70 percent are in average condition, 22 percent are above average, and 8 percent are in below average condition. I control for the effects of time by includi ng dummies for each year in the sample with 1999 as the base year for comparison. In the apartment regression s, I also determine whether the use of the apartm ents is primarily as senior housing, subsidized housing, or multifamily condominiums. The comparison group, which represents 98 percent of the sample, is all other multifamily apartments. Finally, I include dummy variables to c ontrol for differences across submarkets within each major market. These submarkets are defined by CoStar. There are 405 unique submarkets in the apartment sample. In the largest market, Los Angeles, 42 submarkets are identified by CoStar. In the second largest market, New York, 30 different

PAGE 71

60 submarkets are defined. In the smallest of the 15 apartment markets, Sacramento, there are 21 submarkets. In the office sample there are 488 distinct submarkets and that these submarkets largely overlap with the submarkets defined in the apartment sample. There are 51 submarkets present in the Los Angeles office sample, 35 submarkets distinguished in the Phoenix area and only 6 submarkets identified in Tucson, the smallest of the 15 studied markets. The retail sample contains 491 different s ubmarkets. Los Angeles is once again the largest market in this sample with 60 subm arkets, Chicago, the second largest market in the sample, has 37 submarkets, and San Franci sco, which is the smallest market in the sample with only 339 observations, has 33 submarkets defined. The estimated models differ for residentia l and non-residential (office and retail) properties. For each property type I run the specified model by market In order to avoid the effect of outliers in the data, I winsorize all continuous dependent variables in the regressions at the top and bottom one per cent of the distribution. The winsorising procedure takes the non-missing values of a c ontinuous variable sorted in ascending order and replaces its one percent highest and lowest values by the next value counting inwards from the extremes. The only exceptions are FLOORS PARKING and UNITS where I winsorize at the top and bottom 0.5 percent of the distribution, to account for the narrow distribution. Longitude and latitude coordinates are not winsorized. The next chapter presents summary statistics of apartment transaction data by market and the regression results from the models specified.

PAGE 72

CHAPTER 7 RESULTS FOR RESIDENTIAL REAL ESTATE This chapter focuses on the results from the empirical analysis on the apartment market. An additional benefit of the anal ysis on apartment markets is the relative simplicity and homogeneity of apartment leases, which simplifies modeling relative to office and retail properties. Summary statistics for the variables of inte rest are presented in Table 8. The first two columns present a summary of the data at the aggregate level, while the remaining columns present statistics for each of th e 15 markets studied. The average apartment complex in the sample is 49 years old, contai ns 23,034 square feet of improvements, is built on 40,162 square feet of land area, has 27 units, 2.5 floors, 29 parking spaces and sold for $2,194,040. With an average age of 79 years properti es in New York and Boston tend to be much older than in other markets across the country. Phoenix has the newest apartments with an average age of 27 years. The apartm ent buildings in the sample tend to be the largest in Phoenix with an average size of 76,688 sq. ft. Oakl and is the market with the smallest average size of apartment buildings both in terms of s quare footage (12,267) and number of units (15). Approximately 13 percent of the transactions involve the purchase of a replacement property to finalize an exchange; 12 percent i nvolve the sale of the relinquished property in an exchange; and eight percen t involve the sale of relinquish ed property that is also the replacement property of the buyer in a separate exchange. 61

PAGE 73

62 Table 8. Summary Statistics of Apartment Data by Markets Table 8. Summar y Statistics of A p artment Data b y Markets All Boston Chicago Denver Los Angeles New York Apartments Obs 23640 Obs 400 Obs 1585 Obs 782 Obs 7761 Obs 4067 Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. PRICE 2,194,040 4,687,458 2,373,855 5,558,665 1,769,994 4,413,919 3,329,894 7,490,368 1,633,174 3,190,086 1,932,527 4,289,929 AGE 48.81 26.46 78.73 29.67 58.96 25.62 41.98 21.45 42.75 20.41 79.51 17.61 SQFT 23,034 47,387 18,303 41,339 23,116 46,243 37,569 77,415 15,132 25,928 16,829 34,333 LANDSQFT 40,162 115,064 34,552 121,501 35,967 127,081 76,477 188,721 19,308 49,358 5,888 25,033 UNITS 26.99 47.45 21.13 44.46 26.69 42.53 41.34 71.88 18.80 26.80 19.31 31.72 FLOORS 2.47 1.45 3.05 1.24 2.90 1.85 2.40 1.48 2.06 0.74 3.95 1.96 PARKING 29.51 80.43 18.00 56.20 19.85 70.96 57.06 133.35 18.11 40.59 0.58 11.50 X 37.29 4.26 42.36 0.23 41.89 0.16 39.77 0.14 34.05 0.13 40.74 0.08 Y -107.55 18.27 -71.15 0.25 -87.76 0.17 -105.01 0.10 -118.29 0.14 -73.93 0.05 Binary Variables EXREPL 0.13 0.08 0.05 0.20 0.15 0.01 EXRELQ 0.12 0.05 0.08 0.14 0.13 0.00 RELQ_REPL 0.08 0.00 0.02 0.12 0.09 0.00 EXCH 0.32 0.13 0.16 0.46 0.37 0.01 CONDITION_BA 0.07 0.09 0.06 0.06 0.05 0.14 CONDITION_A 0.79 0.82 0.87 0.80 0.86 0.53 CONDITION_AA 0.14 0.08 0.07 0.13 0.08 0.34 BUYEROUT 0.07 0.06 0.02 0.13 0.01 0.04 MULTIFAMILY 0.98 0.95 0.98 0.97 0.99 1.00 SENIOR 0.00 0.00 0.00 0.00 0.00 0.00 SUBSIDIZED 0.01 0.01 0.01 0.02 0.00 0.00 CONDO 0.01 0.03 0.01 0.01 0.00 0.00 CONDOCONV 0.01 0.08 0.01 0.01 0.00 0.00 PORTSALE 0.02 0.02 0.02 0.06 0.01 0.04 YR2000 0.17 0.09 0.20 0.22 0.18 0.04 YR2001 0.17 0.16 0.30 0.17 0.17 0.15 YR2002 0.19 0.18 0.18 0.18 0.22 0.19 YR2003 0.19 0.04 0.18 0.17 0.19 0.24 YR2004 0.19 0.38 0.08 0.18 0.16 0.30 YR2005 0.06 0.12 0.02 0.06 0.05 0.07

PAGE 74

63 Table 8. Continue d Oakland Phoenix Portland Riverside/San Bernardino Sacramento Apartments Obs 1176 Obs 1291 Obs 490 Obs 593 Mean Std. Dev. Mean Std. Dev. Me an Std. Dev. Mean Std. Dev. Obs 370 Mean S td. De v PRICE 1,557,993 3,049,442 5,215,020 8,142,932 2,275,221 4,588,447 3,859,097 7,191,817 3,180 ,281 5,283,3 7 AGE 49.04 21.41 27.35 13.46 36.57 22.36 28.08 14.77 33.0 SQFT 12,267 24,835 76,688 96,327 32,946 56,284 46,825 70,886 3 0 16.21 8,826 57,13 2 LANDSQFT 20,988 52,797 179,486 235,606 80,334 161,134 126,363 203,940 9 3,007 153,49 9 UNITS 15.43 23.90 85.55 93.84 36.26 57.03 53.65 73.32 46.3 FLOORS 2.15 0.75 1.86 1.09 2.05 0.95 1.79 1.37 1.9 PARKING 16.29 41.72 133.63 172.03 51.43 103.14 82.47 134.03 0 60.44 3 0.36 67.89 107.5 1 X 37.82 0.10 33.49 0.08 45.51 0.07 34.03 0.21 38.6 Y -122.19 0.12 -112.02 0.13 -122.66 0.13 -117.22 0.40 Binary Variables 0 0.07 121.43 0.13 EXREPL 0.14 0.05 0.28 0.17 0.2 6 EXRELQ 0.19 0.02 0.15 0.12 0.1 RELQ_REPL 0.12 0.01 0.22 0.10 0.1 EXCH 0.46 0.08 0.64 0.39 0.6 CONDITION_BA 0.08 0.06 0.07 0.02 0.0 CONDITION_A 0.89 0.83 0.75 0.84 0.8 CONDITION_AA 0.03 0.11 0.18 0.14 0.1 BUYEROUT 0.01 0.59 0.23 0.03 0.0 MULTIFAMILY 0.99 0.96 0.97 0.98 0.9 SENIOR 0.00 0.00 0.01 0.01 0.0 SUBSIDIZED 0.00 0.01 0.01 0.00 0.0 CONDO 0.00 0.03 0.01 0.00 0.0 CONDOCONV 0.00 0.01 0.01 0.00 0.0 PORTSALE 0.00 0.03 0.01 0.02 0.0 YR2000 0.30 0.15 0.17 0.14 0.2 YR2001 0.10 0.16 0.25 0.13 0.1 8 8 2 3 0 7 3 9 0 1 0 0 1 4 6 YR2002 0.13 0.16 0.15 0.25 0.22 YR2003 0.19 0.18 0.10 0.21 0.13 YR2004 0.12 0.22 0.23 0.13 0.12 YR2005 0.05 0.10 0.05 0.06 0.06

PAGE 75

64 Apartments Obs 2031 Obs 796 Obs 471 Obs 1281 Obs 546 Mean Std. Dev. Mean Std. Dev. Mean St d. Dev. Mean Std. Dev. Mean Std. Dev. Table 8. Continue d San Diego San Francisco San Jose Seattle Tucson PRICE 2,456,796 5,202,395 1,685,894 2,231,469 1,821,914 3,851,114 2,615,822 5,353,083 2,400,949 4,261,516 AGE 32.31 16.59 68.57 26.86 42.82 15.79 35.49 23.64 30.05 15.37 SQFT 19,124 37,755 7,913 8,770 9,650 19,492 31,382 58,257 44,434 71,118 LANDSQFT 38,008 94,553 6,579 10,138 20,088 46,560 68,933 153,127 111,491 176,153 UNITS 23.45 38.97 9.54 11.08 13.27 25.14 33.41 54.85 57.67 79.91 FLOORS 1.98 0.56 2.61 0.97 1.89 0.39 2.40 0.96 1.78 2.40 PARKING 30.98 66.50 5.91 11.23 14.53 30.31 47.80 100.41 83.36 133.18 X 32.82 0.17 37.71 0.11 37.32 0.08 47.56 0.24 32.24 0.04 Y -117.12 0.10 -122.40 0.08 -121.93 0.11 -122.31 0.11 -110.93 0.05 Binary Variables EXREPL 0.21 0.13 0.16 0.24 0.19 EXRELQ 0.21 0.29 0.23 0.15 0.06 RELQ_REPL 0.16 0.10 0.10 0.08 0.09 EXCH 0.57 0.52 0.48 0.47 0.33 CONDITION_BA 0.04 0.04 0.04 0.07 0.21 CONDITION_A 0.85 0.89 0.91 0.73 0.72 CONDITION_AA 0.11 0.07 0.05 0.20 0.07 BUYEROUT 0.03 0.01 0.02 0.10 0.43 MULTIFAMILY 0.95 1.00 0.99 0.96 0.97 SENIOR 0.01 0.00 0.00 0.01 0.02 SUBSIDIZED 0.00 0.00 0.00 0.02 0.00 CONDO 0.03 0.00 0.00 0.01 0.01 CONDOCONV 0.04 0.00 0.00 0.01 0.00 PORTSALE 0.00 0.04 0.00 0.01 0.04 YR2000 0.20 0.37 0.31 0.15 0.12 YR2001 0.19 0.08 0.06 0.18 0.12 YR2002 0.18 0.09 0.09 0.18 0.17 YR2003 0.17 0.11 0.12 0.17 0.21 YR2004 0.17 0.16 0.13 0.21 0.23 YR2005 0.05 0.05 0.13 0.04 0.12

PAGE 76

65 Approximately seven percent of all buye rs reside out-of-st ate. There are 1,749 transactions in which the buyer was out of stat e. Phoenix is the market with the highest percentage of out-of-state buyers. They re present 59 percent of the 1,291 observations. Other markets of potential interest when quan tifying the effect of having an out-of-state buyer include: Tucson with 43 percent out-o f-state buyers, Portland with 23 percent, Denver with 13 percent, S eattle 10 percent, and New York 4 percent. Approximately two percent of all sales were part of a portfolio sale; there are 488 portfolio sales in the sample. Denver is th e market with the highest percentage of portfolio sales; they represent 6 percent of all sales. The other markets with a large enough number of portfolio sales to generate statistical significance are: New York City, 4 percent of all sales; San Francisco, 4 percent of sales; Phoenix, 3 percent of all observations; and Chicago, 2 per cent of sales. Portfolio sale s and out-of-state buyers in Los Angeles are only 1 percent each, but ther e are 7,761 observations in this market. Slightly over 200 properties, or one percen t of all sales were were purchased by condo converters. There are only two markets in which I coul d look for any price effects from condo-conversion motivation: San Diego with 77 condo conversions, and Boston with 34 such transactions. Table 9 performs t-tests for differences be tween the mean price of apartments in a control group that contains all properties not associated with any conditions and the groups of properties that are the subject of study. These ar e properties that represent a replacement exchange only; properties that ar e associated with a relinquished exchange only; properties that served as both the replacement and relinquished property; purchases by out-of-state buyers only; purchases by c ondo converters; and portfolio sales not

PAGE 77

66 assoc ns fest for Apartment Properties Apartments Observations Mean Value Standard iated with any conditions. For example, the replacement exchanges group contai all properties in which EXREPL is equal to 1, there are no conditions of sale ( CONDOCONV and PORTSALE are both equal to zero) and the buyer is not from out-o state. Table 9. Differences in Mean Prices of Control Sample and Identified Groups of Inter of Sales Price Error T-test Value Control Group 14,383 1,703,610 30,440 EXREPL 2,786 1,878,481 59,843 -2.36* BUYEROUT 1,288 7,004,433 295,063 -39.36* CONDOCONV 133 7,545,770 989,659 -17.68* PORTSALE 310 4,650,359 352,059 -13.79* EXRELQ 2,609 1,749,449 57,314 -0.61 RELQ_REPL 1,687 1,707,154 48,233 -0.04 The r h a s not statistically significantly di fferent from the average price in the ich esults show that replacement apartment ex changes are associated, on average, wit 10.3% price premium and the pr ice difference is statistical ly significant. Relinquished exchanges and transactions a ssociated with both replacemen t and relinquished exchange have average prices that are control group. Sales that are completed by out-of-state buyers are associated with a price wh is four times higher than the average price of properties in the comparison group. The results are similar with respect to sales to condo converters, which ar e associated with an even higher premium, which is also statistica lly significant. Finally, I also observe that portfolio sales are, on average, 173% more e xpensive than transactions in the comparison

PAGE 78

67 sample. Therefore, I reject the null hypothesis that the diffe rence between the mean price in the comparison group and the mean price observed with replacement excha of-state buyers, condo conversi ons and portfolio sales corres pondingly are equal to zero. I was no nges, outt able to reject the null hypothesis fo r sales involving relinqui shed exchanges and Next, I present the results from estimating equation (17) for each of the 15 markets in Table 10. ach oressiotepwise decide which of the submarket dummies to l final ml other dt variables are not subject to the procedure. I use a robust estimation method to account for potential heteroskedas erefored p-va djusted d resu Table 1 the esefficie s butes are predict statisti ficantf the ot surprising and is speci fic to higher density areas in which more n ix two different exchanges. This result is consistent with our expectations. I perform e f the reg ns using a s estimation procedure to eave in the odel. Al ependen ticity; th all reporte lu es are a values. The reporte lts in 0 show th at timated co nts on the tructural attri of the ed si gn and ically sign in most o market regressions. PARKING tends to be positive and insignificant. However, in two markets, San Diego and New York City, it is negative and significant. It also has a negative sign in the Chicago and Los Angele s regressions, with p-values of 0.14 and 0.16 correspondingly. This finding is n parking is usually associated with apartm ent communities that ar e distant from the centers of the city a nd hence tend to be cheaper. The coefficient on UNITS is positive i all regressions but one, and is significant in 6 of the 15 models. In Phoenix, UNITS reverses sign and becomes nega tive and significant at less than one percent level. Phoen is the market in which the largest average transactions are observed. The average

PAGE 79

68 apartment property in the sample has 86 units and 76,688 sq. ft. of improvements. In contrast, the average apartment property so ld in the 15 markets has 27 units and 23,034 sq. ft. of improvements. The estimated coefficient on the variable of interest, EXREPL is positive and significant in 12 of the 15 regressions. This provides evidence that buyers of replacement er tained in the regression sample. t t and positi estate d erved positive coefficients pr ovide evidence that sellers enter in tax delay hen properties are paying statistically significant price premiums in the majority of the larg markets con The coefficient estimates of EXRELQ are generally positive a nd significant in 7 ou of the 15 regressions. However, in only th ree of the regressions are the coefficient estimates on relinquished exchanges higher than the coefficient estimates of replacement exchanges. Casual observation shows that relinquished exchanges have significan ve coefficients in markets that have seen more appreciation in residential real than others. In addition, it is important to re member that if the main motivation to enter into an exchange is to postpone capital gain taxes, properties that are part of relinquishe exchanges will tend to be ones that have seen more capital appreciation than other markets. Hence, all else the same, they will tend to be more expensive properties. Therefore, the obs ed exchanges when their properties have appreciated in value. This positive effect on price may offset any other possible ne gative effects related with relinquished exchanges and thus I can not make any conc lusions about the magnitude of the price impact related to a sale being part of a relinquished exchange. With replacement exchanges, there are no su ch issues; and therefore all else equal the coefficient on EXREPL can be directly associated with a price premium paid w

PAGE 80

69 Table 1 t Dummies by Apartment Markets Market Boston Chicago Denver Angeles York City Oakland Phoenix Portland 0. Regression Statistics for OLS Model with Structural Characteristics and SubmarkeLos New Observations 0 EXREPL 86 400 1585 782 7761 4067 1176 1291 490.139 0.107 0.160 0.076 0.094 0.071 -0.020 0.0 EXRELQ -0.145 0.137 0.140 0.043 0.081 0.050 0.000 0.087 RELQ_REPL 0.756 0.260 0.192 0.093 -0.005 0.113 -0.108 0.113 0.98 0.00 0.18 0.01 AGE AGE2 .000 0.71 0.10 0.21 0.00 0.03 0.14 0.72 0.03 SQFT 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 SQFT2 LANDS LANDSQFT2 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 PARKING 0.001 -0.001 0.000 -0.001 -0.002 0.000 0.000 0.000 99 FLOOR 085 0.06 UNITS 0.01 0.17 0.02 0.00 0.07 0.79 0.00 0.33 COND 0.22 0.00 0.00 0.00 0.69 0.01 0.00 0.58 COND 0.174 BUYER 7 22 0.00 0.32 SENIOR -0.003 0.024 0.607 -0.419 -0.634 0.257 0.394 0.310 SUBSIDIZED -0.536 0.174 0.060 -0.069 -0.364 -1.110 -0.015 0.277 0.10 0.01 0.00 0.00 0.43 0.03 0.67 0.06 0.23 0.00 0.00 0.00 0.15 0.07 1.00 0.13 0.00 0.00 0.00 0.00 -0.008 -0.005 -0.007 -0.009 -0.003 0.000 -0.010 -0.008 0.12 0.05 0.04 0.00 0.10 0.96 0.00 0.02 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0 0.021 0.030 0.029 0.039 0.028 0.043 0.023 0.027 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 QFT 0.003 -0.001 0.000 0.000 0.010 0.000 0.000 0.002 0.03 0.62 0.63 0.87 0.00 0.89 0.37 0.12 0.02 0.18 0.31 0.20 0.00 0.35 0.95 0.12 0.48 0.14 0.22 0.16 0.05 0.80 0.82 0.S -0.036 0.024 0.041 0.044 0.081 0.118 0.035 0.0.26 0.23 0.08 0.02 0.00 0.00 0.14 0.008 0.002 0.002 0.006 0.003 0.001 -0.001 0.002 ITION_BA -0.101 -0.202 -0.194 -0.089 -0.008 -0.094 -0.232 -0.035 ITION_AA -0.057 0.167 -0.018 0.116 0.094 0.099 0.064 0.51 0.00 0.73 0.00 0.00 0.20 0.20 0.01OUT 0.051 0.065 0.038 0.016 0.040 -0.206 0.109 -0.040.66 0.65 0.47 0.77 0.30 0. 0.99 0.93 0.00 0.13 0.00 0.05 0.17 0.36 0.20 0.25 0.60 0.67 0.08 0.25 0.84 0.00

PAGE 81

70 Table 10. Continued Market Boston Chicago Denver Los Ang New Yo C Oakland Phoenix Portland eles rk ity CONDO 0.154 0.186 0.307 0.171 -0.4-0.007 0.103 0.030 84 0. 0. 0.0.0.0ONV -00.-00.0.0.0.00.0.0.0.0.0.0.0E 0.0.-0-0 0.-0-0.0.0.0. 0.0.0-00.0.0.-00.0.-0.0.0.0.0.0.0.0-00.000.0.0.-0.0.0.0.0.0.0.0 0.0.0.0.0.0.0.-0.0.0.0.0.0.0.0 0.0.0.0.0.0.0.00.0.0.0.0.0.0.00.0.0.0.0.0.0.00.0.0.0.0.0.0.00.0.0.0.0.0.0.00.0.0.0.0.0.0.0eported) 1413121312121310.0.0.0.0.0.0.0 00000.0.0. 36 0. 09 00 0. 05 00 90 12 .84 CONDOC .038 098 .038 044 093 516 046 .006 71 21 79 86 03 04 47 .99 PORTSAL 509 158 .104 .012 072 .042 0.601 01 02 16 82 62 48 .00 YR2000 .089 145 067 032 .038 134 136 0.045 58 01 34 12 63 00 04 .58 YR2001 .005 270 .338 .142 040 457 179 0.091 97 00 00 00 60 00 01 .24 YR2002 112 364 261 295 139 627 262 0.028 43 00 00 00 07 00 00 .73 YR2003 332 541 279 523 373 657 279 .095 06 00 00 00 00 00 00 .27 YR2004 394 622 220 753 507 802 443 .146 01 00 00 00 00 00 00 .08 YR2005 559 691 381 911 641 858 539 .343 00 00 00 00 00 00 00 .00 SDUMi (not r CONST .276 .295 .631 .132 .079 .517 .372 2.858 00 00 00 00 00 00 00 .00 R-squared .88 .80 .90 .83 83 85 93 0.89 All standard errors are adfor po heteroicity. s are d unde mates justed ten tial skedast P-value repo rte r the coeffici ent esti and ar e in bold and italics. Dependent Variable C of price PL variequaf transacesents sale placem Q variable resents sale of a relinquished property; REPL ary vset eq one if action nts bo of rehed p of emety. A ge odingars; Age squared; SQF re f e FT re footage of tota emeed; LANDSQFT Square footage of land ; LANDSQFT2 Square fo PARKING Parking, definedber of parking spac OORS ber o umnits ;TIO sicalon opertyn in. The i ad abgeitted y is a BUY Binable sto one if buyer lives out of s NIOR ary vaet equne if p use is multi-; SUBSIDIZED B et equal pr op is subsidized muy; CO Binale se onety use is multi-family condo; CONV variequaif tr anas pndo conversion; PORTSALE Binary varit equal to one if t Y early through 2005. Each year ed ay vacepthichessedi Briableng the submarket in which the property is as d CoStar brokers. is LNPRI E Log selling ; EXRE Binary if transac able set l to one i tion repr of a re RELQ_ ent property Bin EXREL ariable Binary ual to set equal to one trans tion rep th sale represe linquis roperty and purchase ootage of total improv a replac ments; SQ nt proper 2 Squa GE A f the buil l improv (s) in ye nts squar AGE2 T Squa otage of land squared; as num es; FL Num f floors ; UNITS N nclude below average ber of u verage an CONDI ove avera N i Phy The om conditi categor f the pro verage; based o EROUT spection ary vari categories et equal tate; SE Bin riable s al to o roperty senior family inary variable s to one if minium erty use NDOCO lti-famil able set NDO l to one ry variab saction w t equal to art of co if proper Binary able se ransaction was part of portfolio sale; R n Y time periods from 1999 is includ s a binar lo cated, riable ex efined by 1999, w is suppr ; SDUM inary va signifyi

PAGE 82

71 Table 10. Continued Market Riverside Bernardino c sco sle /Sa n Sa ramento San Diego San Franci San Jo e Seatt Tucson Observat ions 593 1 1 XREPL 0.098 0. 8 1 5 4 370 2031 796 47 128 546 E 158 0.075 0.07 0.10 0.05 0.04 0.03 0 3 XRELQ 0.120 0. 8 4 6 1 0.02 .01 7 9 ELQ_REPL 0.234 0. 115 2 4 4 0.00 0 5 2 GE -0.012-0 0 02 02 5 0.01 0 5 6 GE2 0.000 0. 0 0 0 0 0.33 0 7 9 QFT 0.028 0. 4 3 5 1 0.00 0 0 0 QFT2 0.000 0. 1 0 0 0 0.00 0 0 0 ANDSQFT 0.001 0. 1 5 0 2 0.21 0 4 7 ANDSQFT2 0.000 0. 0 0 0 0 0.08 0.29 0.36 0.09 0.17 0.64 0.54 0.000 -0.001-0.0010.001 -0.002 0.001 001 47 3 8 LOORS -0.0010. 2 7 8 3 0.98 0 0 00 00 SUBSIDIZED 0.358 0.002 -0.226 -0.181 -1.378 0.102 0.400 0.00 0.98 0.17 0.20 0.00 0.14 0.00 CONDO -0.040 0.174 0.341 4.651 0.325 0.424 0.61 0.01 0.00 0.00 0.00 0.06 0.00 0.00 0.01 0.0 0.0 0.28 E 184 0.030 0.06 0.00 0.01 0.03 0.00 0.19 0 0.8 0.5 0.61 R 182 0.087 0. 0.08 0.09 0.12 .00 0.00 0.00 0.0 0.0 0.02 A .009 -0.002 0.00 -0.0 -0.0 -0.00 .04 0.25 0.90 0.7 0.3 0.13 A 000 0.000 0.00 0.00 0.00 0.00 .34 0.81 0.51 0.9 0.6 0.60 S 028 0.030 0.09 0.05 0.02 0.02 .00 0.00 0.00 0.0 0.0 0.00 S 000 0.000 -0.00 0.00 0.00 0.00 .00 0.00 0.00 0.0 0.0 0.00 L 002 0.001 -0.00 0.00 0.00 0.00 .43 0.25 0.90 0.1 0.6 0.05 L 000 0.000 0.00 0.00 0.00 0.00 PARKING 0. 0. 0.18 0.03 0.79 0.3 0.1 0.00 F 143 0.112 0.07 0.13 0.15 -0.00 .04 0.01 0.0 0. 0. 0.60 UNITS 0.000 0.001 0.007 0.001 0.020 0.002 0.001 0.81 0.43 0.00 0.80 0.02 0.14 0.47 CONDITION_BA -0.061 -0.153 -0.101 -0.186 -0.113 -0.140 -0.163 0.53 0.11 0.02 0.00 0.06 0.00 0.00 CONDITION_AA 0.047 0.048 0.175 0.180 0.140 0.107 0.055 0.38 0.47 0.00 0.00 0.01 0.00 0.45 BUYEROUT -0.013 0.057 0.019 -0.017 -0.019 -0.015 0.190 0.89 0.74 0.78 0.83 0.79 0.75 0.00 SENIOR 0.172 0.151 -0.888 0.061 0.248 0.18 0.04 0.00 0.80 0.00

PAGE 83

72 Table 10. Continued Market Riverside /San Bernardino Sacramento San Diego S F SST an rancisco an Jose eattle ucson CONDOCONV -0 -0 0. 0.14 .338 .061 160 6 0.12 0.82 0.02 0.E --0-00.00.0.0.00.0.-0-00.0.0.00.0.0.-00.0.0.0.0.0.00.0.0.0 0.0.0.00.0.0.0 0.0.0.00.0.0.000.0.00.0.0.0eported) 112131210.0.0.0 000. 19 PORTSAL 0.037 -0.798 0.248 .156 .550 111 .154 0.70 0.00 0.16 00 15 35 .16 YR2000 0.106 0.147 0.052 208 108 .007 .128 0.19 0.07 0.27 00 00 89 .16 YR2001 0.258 0.221 0.211 321 347 022 .024 0.00 0.01 0.00 00 00 68 0.035 .80 YR2002 0.343 0.474 0.445 376 363 077 0.00 0.00 0.00 00 00 14 .71 YR2003 0.472 0.620 0.664 386 375 148 .074 0.00 0.00 0.00 00 00 01 .41 YR2004 0.726 0.855 0.907 454 397 163 .127 0.00 0.00 0.00 00 00 00 .15 YR2005 0.998 0.877 0.842 .510 370 299 .309 0.00 0.00 0.00 00 00 00 .00 SDUMi (not r CONST 12.528 12.704 3.363 .940 .050 .804 2.869 0.00 0.00 0.00 00 00 00 .00 R-squared 0.91 0.91 0.85 .84 .92 89 0.92 All standard errors are adjusted for poten tial heteroskedasticityues are d undereffici enates a in P-val repo rte the co t estim nd ar e bold and italics. Dependent Ve is LNPR Log ofg price; PL Bvariablual if t presents sale of m ent pro XRELQ y variaqual toransaresents sale of a relinquished property; RELQ_REPL Binarble set o one if ction rets bothof reled property and purchase of a repnt propert Age ouilding(ears; AG ge sq SQFT re footage of total improvement 2 Squae of to vemened; LANDSQFT S ; LANDSQFT2 Square footaged squared; ING P definedber of parking spaces ORS er of floors ; UNITS Number o ; CONDIT Physicaition of operty bn insphe c i averagove ave omittry is a YE Binale s one if buyer lives out of state; R Binarble set eqone if p use is sulti-fa UBSIDIZED Binary variable set equal to on y use dized muily; CO Binary e set e one ifty u ndomin NDOCO nary vart equaf tr ansas pado con; PORT ariableual to one saction was part of portfolio sale; YR arly time periods from 1999 through 2005. Each year is incs a binary le except 1999, which iressed; S Binaiable sng t the prlo cated, d by CoStar brokers. ariabl ICE sellin EXRE inary e set eq to one ransaction re a replace perty. E Binar ble set e one if t ction rep y varia y. AGE equal t f the b trans a s) in y presen E2 A sale uared; inquish Squa laceme s; SQFT re footag ta l impro ts squar quare footage of land of lan f units PARK ION arking, l cond as num the pr ; FLO ection. T Numb ategories irage. The ased o nclude below average e and ab ed catego verage; BU ROUT ry variab et equal to SENIO e if pr opert y varia is subsi ual to lti-fam roperty NDO enior m variabl mily; S qual to proper se is multi-family co SALE Binary v ium; CO NV Bi iable se l to one i action w rt of con nversio set eq luded a if tran variab n Ye DUM s supp i ry var ignifyi he submarket in which operty is as define

PAGE 84

73 the sale is part of a replacement exchange. The coefficient on RELQ_REPL also is ositive and inekcisien ts the ced price effect of relinquished nd replacent exchange vation. Thitude coefte nd larn thicie placement exchanges. The coeff on the ble ind ng tha buyet-ofBUYEROUT istically conom sign nt in j o ma Ph and Tucson. The also mts of s ic intwhen zing rice premiums associated with out-of-state buyers. In Phoenix 59 percent of sales were to out-state buyers, while in Tucs t. With conium conversions there are only two maf po inte iego ad Boston. I fid that the coeffici ent on the variable indicating a purchase by condo conve NDO s economicalstat sigt in go. The coefnt is alsoifi canew Yity aklan e dummy coefficients reveal substantial price appr eciation over the seven year study period. Finally, although submarket dummy variables ar e not reported in th e regression tables, the model fit is improved significantly by th e use of the submarke t location controls. p significant 13 out of th 15 mar ets. As dis ussed, th coeffic t represen ombin a em moti e magn of this ficient s to be ger tha e coeff nt on re icient varia ica ti t the r is ou state, statis and e ically ifica ust tw rkets oenix ese ar arke peci f erest analy the p of on their share was 44 percen domin rkets o tential rest San D n n a rter, CO CONV i ly and istically nifican San Die ficie sign t in N ork C nd Oa d. The coefficient on the variable indicating a portfolio sale, PORTSALE, is statistically significant in 3 out of the 5 markets of interest Chicago, New York City and San Francisco. However, contrary to our intuition the coefficient on PORTSALE in San Francisco is negative, rather than positive. The estimated year dummies, with 1999 as the omitted year, are generally positive and significant. Moreover, the magnitude and significance of the tim

PAGE 85

74 In OLS Model II, I add controls for ab solute location by using a third expansion of the propertys latitude and longitude coordinates. This expansion yiel additional explanatory variables. order ds nine following by about 0.01) the fit of the mode e REPL is positi result s lly positive and significant in 7 out of the 15 markets 1 The resulting model specification has the form: Table 11 presents the regression results for Model II. Controlling for absolute location, in addition to relative location, impr oves slightly ( ls. Adjusted R-squared for the market regressions range from 0.81 to 0.92. Th regression results are virtually unchange d from the results reported for Model 1. Therefore, I will only discuss results rela ted to the key variables of interest. The coefficient on the variable indicating a replacement exchange, EX ve and significant in 12 of the 15 regressions. This reconfirms the previous that buyers of replacement prope rties are paying statistically significant price premium in the majority of the sample markets. The coefficient on the variable indicating a relinquished exchange, EXRELQ remains genera regressions. The size of coefficient estimates remains largely unchanged. The coefficient on RELQ_REPL is again positive and significant in 13 out of the 15 and its magnitude remains larger than the coefficient on replacement exchanges. 1 I also experiment with a model where I control for differences in the effects of key structural characteristics across submarkets by interacting age and the square footage of improvements with the su market dummies. The Third Model yields a slightly better fit R-squares are increased by 1-5 percent b across markets. Howeve r, it sacrifices even more degrees of freedom. In addition, coe fficients on age and square footage, due to the presence of interaction vari ables, are mostly insignificant and hard to interpret. Therefore, I do not report the results from the third model in the dissertation. 2 _3 7 6 5 4 3 2 )19( ),( 2 200 43 2 2005 2000 18 17 16 15 14 13 2 12 11 10 9 8 10 mEXREPL LNPRICE r k r l lk P s s s n nn i i iUNITS FLOORS PARKING LANDSQFT LANDSQFT FT lkYX SMDUM YR PORTSALE CONDOCONV CONDO SUBSIDIZED SENIOR BUYEROUT CONDITION SQ SQFT AGE AGE REPL RELQ EXRELQ

PAGE 86

75 Table 11. Regr Dummies and Longitude, Latitude Coordinates by Apartment Markets Market Boston Chicago Denver Oakland Phoenix Portland ession Statistics for OLS Model w ith Structural Characteristics, Submarket Los Angeles New York City Observations 400 1585 782 7761 4067 1176 1291 EXREPL 0.145 0.099 0.162 0.076 0.108 0.070 -0.020 4900.086 0.09 0.02 0.00 0.00 0.35 0.03 0.66 0.06 EXRELQ -0.131 0.128 0.137 0.046 0.046 0.054 0.000 0.08 0.28 0.00 0.00 0.00 0.45 0.05 1.00 0.13 RELQ_REPL 0.754 0.234 0.191 0.092 -0.010 0.115 -0.106 0.11 0.00 0.00 0.00 0.00 0.97 0.00 0.19 0.02 AGE -0.008 -0.005 -0.006 -0.010 -0.002 -0.001 -0.010 -0.00 0.13 0.08 0.05 0.00 0.21 0.60 0.00 0.02 AGE2 0.71 0.16 0.22 0.00 0.23 0.29 0.72 0.03 SQFT SQFT2 LANDSQFT 0.002 0.000 0.000 0.001 0.009 0.000 0.000 0.002 0.13 LANDSQFT2 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 PARKING 0.001 -0.001 0.000 -0.001 -0.002 0.000 0.000 0.000 FLOORS -0.039 0.024 0.042 0.043 0.072 0.114 0.035 0.086 UNITS 002 0.33 COND 0.25 0.00 0.00 0.00 0.34 0.01 0.00 0.54 COND 2 BUYER 0.33 SENIOR 0.270 0.41 SUBSI 8 1 8 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.021 0.029 0.029 0.039 0.029 0.043 0.023 0.027 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.04 0.95 0.68 0.60 0.00 0.78 0.36 0.02 0.34 0.33 0.30 0.00 0.40 0.97 0.13 0.43 0.19 0.21 0.17 0.11 0.90 0.81 0.99 0.23 0.19 0.07 0.02 0.00 0.00 0.14 0.06 0.009 0.002 0.002 0.006 0.003 0.001 -0.001 0.0.01 0.19 0.02 0.00 0.08 0.85 0.00 ITION_BA -0.093 -0.191 -0.199 -0.086 -0.019 -0.091 -0.232 -0.040 ITION_AA -0.063 0.160 -0.015 0.114 0.071 0.089 0.064 0.170.47 0.00 0.78 0.00 0.00 0.25 0.20 0.01 OUT 0.056 0.072 0.035 0.013 0.038 -0.222 0.110 -0.046 0.63 0.61 0.50 0.80 0.31 0.18 0.00 -0.017 -0.002 0.548 -0.418 -0.677 0.241 0.398 0.91 0.99 0.01 0.12 0.00 0.08 0.17 DIZED -0.531 0.223 0.049 -0.074 -0.328 -1.060 -0.018 0.245 0.20 0.13 0.66 0.66 0.13 0.27 0.81 0.03

PAGE 87

76 Table 11. Continued Chicago Denver Los New York Oakland Phoe Market Boston An nix Portland geles City CONDO 0.140 0.171 0.298 0.154 -0.441 0.018 0.103 0.028 0.41 0.11 0.00 0.07 0.01 0.74 0.12 0.85 0 ORTSALE -R2000 R2001 R2002 R2003 R2004 R2005 3 Y2 X2 DUMi (not reported) CONDOCONV 0.035 0.028 0.045 0.041 1.415 0.049 .006 0.74 0.74 0.75 0.87 0.00 0.44 0.99 P 0.514 0.149 0.098 0.005 0.086 0.145 0.041 0.605 0.01 0.02 0.18 0.92 0.04 0.33 0.50 0.00 Y 0.087 0.142 0.074 0.033 0.050 0.135 0.138 0.049 0.60 0.01 0.30 0.10 0.53 0.00 0.03 0.55 Y 0.003 0.272 0.351 0.146 0.034 0.468 0.180 0.095 0.98 0.00 0.00 0.00 0.65 0.00 0.01 0.23 Y 0.119 0.366 0.271 0.298 0.129 0.636 0.263 0.032 0.41 0.00 0.00 0.00 0.09 0.00 0.00 0.70 Y 0.356 0.535 0.287 0.529 0.365 0.665 0.279 0.094 0.04 0.00 0.00 0.00 0.00 0.00 0.00 0.28 Y 0.401 0.642 0.233 0.759 0.506 0.812 0.443 0.143 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.09 Y 0.552 0.694 0.393 0.916 0.645 0.864 0.540 0.337 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Y 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.54 0.00 0.18 0.00 0.00 0.00 0.96 0.97 X 0.000 0.000 0.000 0.000 0.000 0.000 0.00 0.71 0.24 0.00 0.42 0.73 Y 0.000 0.000 0.08 0.60 S CONST 7.226 13.126 33.875 33.261 62.993 9.494 22.172 27.363 0.34 0.09 0.03 0.00 0.00 0.66 0.05 0.16 R-squared 0.88 0.81 0.90 0.83 0.83 0.85 0.93 0.89 All standard errors are adjur pot eterosicity. es are r undereffici eates sted fo en tial h kedast P-valu epo rted the co nt estim and ar e in bold and italics. Dependeble RIC of price; L Bariabual if t ale acem pert Q variaqual toransapresents sale of a relinquished property; REPL ry vaset eq one if ction rets bo of rehed p chase of a replacement property. A ge oildingars; A ge s SQF re f mprovem FT2 re footage of tota emened; LANDSQFT Square footage of land ; LANDSQFT2 Square foo land ; PARKING Parking, defined ber of parking space ORS ber o NITS Numbits ;ION sicalon oferty inshe iw average, av abage. ittedy is a YE Binable sto one if buyer lives out of sta IOR ry varet equne if pruse is sulti-f SUBSIDIZED B equal tr op is subsidized mu; COND y setone ty use is multi-family condo CONV varequal tr anss p PORTSALE Binary varia if tron w of porale; YR arly time periods from 1999 rough 2005. Each year is inclued as ry variab which is SDUMi Biniable signing e submarket in which the property is lo cated, as defined by CoStar brokers; X latitude of property; Y longitude of property. nt Varia of a repl is LNP ent pro E Log y. EXREL selling Binary EXREP ble set e inary v one if t le set eq ction re to one ransaction represents s ELQ_R Bina riable ual to trans a presen th sale linquis roperty and pur ootage of total i GE A f the bu l improv (s) in ye ts squar GE2 A quared; T Squa ents; SQ Squa tage of squared as num s; FLO Num f floors ; U nclude belo er of un erage and CONDIT ove aver i Phy The om conditi categor the prop verage; BU based on ROUT pection. T ry varia categories et equal te; SEN Bina iable s al to o operty enior m amily; inary variable set o one if p minium; erty use NDOCO lti-family iable set O Binar to one if variable action wa equal to art of condo conversion; if proper Binary ble set equal to one d ansacti le except 1999, as part tfolio s suppressed; n Ye a bina ary var ify th th

PAGE 88

77 Table 11. Continued Market Rive /S r a c nto Fro side n Sa Bernardino rame San Diego San ancisc San Jose Seattle Tucson O bservations 593 372 6 0.10.0.0 0 031 796 471 1281 546 EXREPL 0.09 58 072 76 0.099 0.053 0.045 0. 03 0.00 0.0.0 122 0.1002 0.00 0.EPL 5 0.10.00 0.00 0.012-0.0-001 0.04 0.000 0.00.36 0.33 0.028 0.00.00 0.00 0.000 0.00.-00 0.00 0.FT 001 0.00.17 0.46 0.DSQFT2 000 0.00.08 0.32 0.ING 0.000 -0.0-00.0 0.52 0.18 0.0.8 RS 01 0.147 0.10 0.137 0.58 -003 97 0.04 0.01 0.0 00 00.000 0.001 0.00 0.00.020 0.02 0.1 .77 0.ONDITION_BA .0560.1-0-058 0.10 0. 00 2 0.00 0.04 0.28 EXRELQ 0. 82 .028 0.062 0.002 0.013 0.034 0. 23 0.01 0.96 0.67 0.59 RELQ_R 0.22 82 087 0.109 0.084 0.090 0.130 0. 00 0.00 0.04 0.02 0.02 AGE -0. 09 .002 0.001 -0.002 -0.002 -0.006 0. 21 0.75 0.74 0.38 0.07 AGE2 0. 00 000 0.000 0.000 0.000 0.000 0. 78 0.66 0.99 0.72 0.39 SQFT 0. 28 030 0.093 0.053 0.026 0.021 0. 00 0.00 0.00 0.00 0.00 SQFT2 0. 00 000 0.001 0.000 0.000 0.000 0. 00 0.00 0.00 0.00 0.00 LANDSQ 0. 02 001 0.001 0.005 0.000 0.002 0. 24 0.84 0.15 0.63 0.05 LAN 0. 00 000 0.000 0.000 0.000 0.000 0. 35 0.33 0.18 0.65 0.59 PARK 01 .001 01 -0.002 0.001 0.001 04 4 0.33 0.20 0.00 FLOO -0.0 9 0.070 1 .0 0. 0.00 0 .0 .58 UNITS 7 01 0 00 0 45 0.00 0.72 0.02 0.15 0.33 C -0 59 .100 .182 0.112 -0.139 -0.145 0. 02 0.00 0.06 0.00 0.00 CONDITION_AA 0.046 0.047 0.173 0.150 0.139 0.107 0.054 0.39 0.49 0.00 0.00 0.01 0.00 0.46 BUYEROUT -0.021 0.047 0.016 -0.008 -0.017 -0.012 0.197 0.83 0.77 0.81 0.92 0.81 0.81 0.00 SENIOR 0.175 0.156 -0.891 0.061 0.248 0.18 0.03 0.00 0.80 0.00 SUBSIDIZED 0.375 0.021 -0.191 -0.235 -1.388 0.085 0.446 0.00 0.86 0.28 0.10 0.00 0.24 0.00 CONDO -0.057 0.170 0.365 4.640 0.348 0.401 0.48 0.02 0.00 0.00 0.00 0.07

PAGE 89

78 Table 11. Continued Market Riverside /San Bernardino Sacrame nto San Diego San Francisco San Jose Seattle Tucson CONDOCONV 0.345 0.066 0.152 0.143 0.07 0.81 0.02 0.19 P ORTSALE R2000 R2001 R2002 R2003 R2004 R2005 3 Y2 X2 eported) 0.026 0.807 0.252 0.160 0.555 0.093 0.124 0.79 0.00 0.15 0.00 0.15 0.43 0.27 Y 0.111 0.144 0.046 0.208 0.107 0.008 0.118 0.17 0.07 0.33 0.00 0.00 0.88 0.19 Y 0.261 0.217 0.209 0.317 0.351 0.024 0.028 0.00 0.01 0.00 0.00 0.00 0.65 0.77 Y 0.345 0.474 0.441 0.366 0.368 0.078 0.037 0.00 0.00 0.00 0.00 0.00 0.14 0.69 Y 0.474 0.620 0.664 0.379 0.381 0.149 0.071 0.00 0.00 0.00 0.00 0.00 0.01 0.43 Y 0.724 0.852 0.907 0.463 0.399 0.167 0.129 0.00 0.00 0.00 0.00 0.00 0.00 0.14 Y 1.002 0.876 0.848 0.506 0.374 0.306 0.319 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Y 0.000 0.000 0.000 0.000 0.000 0.001 0.000 0.64 0.89 0.00 0.00 0.10 0.00 0.03 X 0.001 0.000 0.000 0.000 0.000 0.007 0.000 0.67 0.73 0.02 0.00 0.28 0.00 0.07 Y 0.001 0.009 0.68 0.00 SDUMi (not r CONST 3.215 28.361 -9.089 18.833 -8.118 20.801 28.200 0.70 0.36 0.60 0.45 0.69 0.07 0.04 R-squared 0.91 0.91 0.85 0.85 0.92 0.89 0.92 All standard errors are adjusted for poten tial heteroskedasticitylues are d undereffici enates a in P-va repo rte the co t estim nd ar e bold and italics. Dependent Vas LNPR Log og price L Bariabqual if t ale of a re ent prop RELQ ry variaqual toransaresents sale of a relinquished property; RELQ_ Binaryble set to one if action rets bothof reled propertychase of a replacement property. Age of ing(s) in ars; AGE2 T f Squar of to ovemeed; LANDSQFT Sotag ; LANDSQFT2 Square footage of land squared; NG P, definedber of parking spaces ORS er of floors Number of Physical of y he c iw average, average e averageittory is averge; BUYE Binarle seto one if buyer lives out of state; SE Binary variable set e one if propy use is sulti-fa UBSIDIZED Binar one if pr y use is ily; COND Binary one if prop use is multi-family condominium DOCON ary vt equaf tr anss pado con; PORTSALE Binary variable sel to one action was part of portfolio sale; YR arly time periods from 1999 through 2005. Each year is inclua binary variable except 1999, which S Binaable signng tet in which the prop cated, as defined by CoStar brokers; tude of; Y de of property. riable i placem ICE erty. EX f sellin Bina ; EXREP ble set e inary v one if t le set e ction rep to one ransaction represents s REPL varia equal trans presen sale inquish and pur ootage of total improvements; AGE e footage the build ta l impr ye nts squar Age squared; SQF quare fo Square e of land SQFT2 PARKI arking as num the propert ; FLO Numb ; UNITS nclude belo units ; CONDITION and abov i condition ed categ based on inspec ROUT tion. T y variab ategories t equal The om a NIOR qual to ert enior m mily; S y variable set equal to opert ; CON subsidized m V Bin ulti-fam ariable se O l to one i variable set equal to action wa rt of con erty nversio t equa if trans n Ye DUM ded as erty is lo is suppressed; X lati i property ry vari longitu ifyi he submark

PAGE 90

79 The coeff icient on BUYEOUT remins positive and significant in Phoenix and ucson. The coeft on COa e fica The coeffon POLE o mical and statiscally sant in ago, New Yty, Poand ancisRTS alstive significant in Bosut neg and s cant intland acram How these results will e discussed since are based on too fserva. Based on OLS Model II, the percentage sa les price changes co rrespg to estimated (statistically significant) coefficients for EXREPL EXQ, RREP ONDOCONV, BOUT PORTSALE for each markerese Ta. Percentageeffect alcusing = 100*{exp( x ) ons ith Halvorsen amqui0). Iequat s the e ef sae (PRICE) of the pe of a n (e.g. replaceexcha XR relinquishment exchange ( EXRELQ ), etc s EXR EXRELQ, or sothe condition. The percen being part of a replacement property hange ranes from 5.43 ercent (in Sea effect of the transaction bei ng part of relinquished property exce on sprice R a T ficien CONDO NV rem ins positiv and signi ant in S n Diego. icient RTSA is econ ly ti ignific Chic ork Ci rtland San Fr co. PO ALE is o posi and ton, b ative igni fi Por nd Sa ento. ever not b they ew ob tions ondin the REL ELQ_ L, C UYER and t are p nted in ble 12 price s are c lated u 100* g 1} c istent w nd Pal st (198 n this ion g i relativ fect on le pric resenc cond itio ment nge ( E EPL ) ); x i EPL or ome r tage price effect of the sale exc g p t tle) to 17.58 percent in Denver, while the hang ales is between 4.73 percent in Los Angeles to 20.02 percent in Sacramento. The price premium associated with a sale being part of both a replacement and relinquished exchange ranges from 8.81 percent in San Jose to 26.26 per cent in Chicago. Although the coefficient on relinquished property is harder to interpret, because of the issu es discussed, it is clear that replacement exchanges have a positive effect on price. Moreover, in the majority of the markets (12 out of 15) this price effect is statistically and eco nomically significant.

PAGE 91

80 In comparison, the simulation analysis disc ussed in Chapter 5 predicts that when the holding period, both for the relinquished and replacement property, does not exc years, the observed incremental NPV from a 1031 exchange is 1 4 percent. With a 10 year holding period, the incremental NPV in creases to 4 8 percent, depending on the assumptions for the other key variables. Finally, with a 20 year holding period incremental value from using an exchange ranges from 4.8 to 10 pe rcent. Therefo eed 5 re, partic les result d a nt positive impact on price in San Francisco. ue ipants in tax-delayed exchanges that ha ve a short-term investment horizon need to be very careful, since the w ealth that they lose in th e form of a higher replacement property price may offset, in whole or in pa rt, the gain from the deferment of taxes. Out-of-state buyers are associated with a price premium of 11.61 percent in Phoenix and 21.77 percent in Tucson. Condomini um conversions are associated with a price premium of 16.37 percent of sales price in San Diego. Finally, portfolio sa in 8.99 percent higher prices in NYC, a 16.10 per cent increase in price in Chicago an 14.79 perce Robustness of Results Omitted Variables Issues As noted previously there is a concern that coefficient estimates may be biased d to omitted variables. The magnitude of the coefficients of the variable representing a relinquished exchange ( EXRELQ ) clearly shows that properties that become part of relinquished exchanges may have some ex traordinary characteristics which are not captured by the variables in the model. I follow the Haurin (1988) and Glower, Haurin and Hendershott (1998) approach to form a variable that controls for unusual

PAGE 92

81 TableBoston 40015.55% Oakland11767.28%5.53%12.24% Riverside/San Sacramento37017.12%20.02%19.96% San Di Tucson 546 13.91%21.77% 12. Marginal Effects for Significant Coefficients for Variables of Interest MarketObs EXREPLEXRELQRELQ_REPLBUYEROUTCONDOCONVPORTSALE Chicago158510.41%13.65%26.36% 16.10% Denver 78217.58%14.65%21.00% Los Angeles77617.90%4.72%9.63% New York City4067 8.99% Phoenix1291 11.61% Portland4908.96% 11.69% Bernardino 59310.03%13.01%25.23% ego20317.44% 9.09% 16.37% San Francisco7967.86%6.39%11.55% -14.79% San Jose47110.41% 8.81% Seattle 12815.43% 9.46%

PAGE 93

82 characteristics of a property. Each propert ys atypicality is computed based on th e hedonic equations coefficients, and the fo rmula provided in Glower, Haurin and Hendershott (1998) %100* ) exp() exp(*SP hbahba ATYPi ii ii (20) The property atypicality measure, ATYP is presented as percentage of selling price ( SP ); are the physical characteristics, are the mean values of these characteristics, and a and bi are the intercept and the slope estimates from the hedonic regression (Ibid.). The hedonic equation used has the following form (21) Location characteristics are excluded. The f undamental value of a property is presented by the value predicted from a hedonic price e quation based on a sample of apartment properties sold across all mark ets that are not associated with any conditions, completed with out-of-state buyers, or part of an exchange. Next, I include the atypicality measure in the model specified by the equation The estimated coefficients from this model are not different from the ones )22( ),( _00 43 2 19 13 3 2 10 r k r l lk lk P s s s mYX SMDUM ATYP BUYEROUT REPL RELQ EXRELQ EXREPL LNPRICE ih ih 2 2 212 11 10 3 2 9 8 7 6 5 4 3 2 10CONDO SUBSIDIZED SENIOR CONDITION UNITS FLOORS PARKING LANDSQFT LANDSQFT SQFT SQFT AGE AGE LNPRICEi i i m estimated with OLS Model II. This leads us to conclude that OLS Model II is well specified.

PAGE 94

83 Endogeneity Issues There is a concern that the variable representing relinquished exchanges, EX is not exogenous wit RELQ h respect to selling pri ce. For example, an omitted factor such as extreme price appreciation, whic h is correlated with both the probability for a sale to be art of a relinquished exchange as well as selling price, cou relationship between the variab le representing that the sale is part of a relinquished exchange an a estimates the endogenous variable as a function of all exogen ous variables. In the second stage it regresses the dependent va riable (in this case, natural log of price) on all variables and includes the residuals of the endogenous variable. The second stag estimates, when performing this procedure, are not different from the coefficients estimated in the original model. Also the DWH test shows that there is no severe bias in the OLS model estimates. Therefore, the result s from the OLS model appear to be robust to endogeneity concerns. p l d cause a significant d the selling price. To address these concerns, I perform a DurbinWu Hausmn (DWH) test for endogeneity. I estimate the base OLS Model II using two-stage least squares regressions, where EXRELQ is an endogenous variable. The DWH test first e coefficient

PAGE 95

CHAPTER 8 RESULTS FOR OFFICE PROPERTIES attention on results from the empirical analysis on the comm er than s s. erty rnardino, Sacrament o, San Diego, Seattle, Tampa, Tucson and Washington, DC. Los Angeles is the larg est market represented with 1,491 sales, followed by Phoenix with 995 observations. Th e smallest market is Tucson with 264 office transactions. This chapter focuses ercial real estate office market. There are 8,871 observations in the office property sample used for regression analysis. Summary statistics for the variables of interest are presented in Table 13. Statistics at the aggregate data level ar e presented in the first two columns. With office properties, the average property price is about 2.5 times high apartments. The mean sales price is $5,353,894 and the standard deviation of price i $9,734,759. Office transactions are, on average, mu ch larger than our apartment sale The average office property sold in the sa mple has 41,718 square feet of improvements and is built on 78,952 square feet of land. Also, the office properties in our sample are newer than apartments, averaging 29 years of age. A large portion of the office properties are classified as being in excellent co ndition (32 percent), 66 percent are in good condition, and only 3 percent ar e in below average condition The average office prop has 2.43 floors and 100 parking spaces. There are fifteen markets represented in this sample including: Chicago, Dallas/Forth Worth, Denver, Fort Lauderdal e, Las Vegas, Los Angeles, Oakland, Phoenix, Riverside/San Be 84

PAGE 96

85 Table 13 also presents summary statistics of regression variables by markets. I notice that the largest average transactions ar e observed in the Washington, DC area with an average sales price of $14,800,000, which is price of $5,353,894 for lso has the largest avera hich is he e sale a sepa rate exchange. This is in contrast to the excha d property lyzed nt a significantly higher than the average the entire office sample Washington, DC a ge size office property in the sample w ith a mean square footage of 93,519, w also significantly higher than the average squa re footage size of properties in the sample, 41,718. In contrast, the smallest average tran sactions based on both price and square footage of improvements, are observed in Tu cson. The average office property in Tucson had a selling price of only $1,713,384 and is just 15,370 sq. ft. in size Office properties sold in Oakland and Los Angeles tend to be the oldest with average ages of 34.82 and 37.96 years, respectively. The market with the newest office properties is Las Vegas. T mean office building age in Las Vegas is onl y 14.54 years with a standard deviation of 15.89. Approximately 12 percent of the transact ions in the office sample involve the purchase of a replacement property to complete an exchange; 6 percent involve th of a relinquished property; and 3 percent re present both a sale of relinquished property and purchase of a replacement property in nge distribution in apartment markets, where replacement and relinquishe exchanges were approximately equal in number. The retail data, which will be ana in the next chapter, also contain a much higher number of replacement exchanges than relinquished exchanges. This leads us to the conclusion that with offices and retail properties it is often the case that more than one property is invol ved in the replaceme exchange.

PAGE 97

86 Table 13. Statistics of Markets Summary Office Data by Table 13. Summary Statistics of Office Data by Markets h Dallas/ Wo Denve aa Ls All Cicago Fortrth r Ft. Luderdle as Vega Offic b 758 Obs 4 s16 Vari Mea Dev. MeanSt Mea Std. e Os 8871 Obs able n Std.Mean Std. Dev. 333 Obs587 Obs 55 Ob 3 d. Dev.n Std. Dev. MeanStd. Dev. Mean Dev. PRIC 39 014 00 600,00018 70,33, 6 55,30 AGE 25 2.5 66 21.17 3.0 1 SQFT 1, 769 03 816 11 ,7 17 ,3 92 LANDSQFT 8, 148 63,0 13962 29 6,4 204 ,0 4 FLOORS 90 89 1. PARKING 00. 11 29 4 79. 8 3 54 1900,0 4,26,1,6778 255,151 5,53,656 2,982,23 367,3 17.42 30.46 2824.4 14.98 14.54 15.89 03,38 39,709 672530,5 48,961 2077 42,8 188,98 3,531 1657 83,7 117,995793 93,72 2.87 2.42 1.2.18 1.67 410.87 239.7 111.41 16991.8 154.71 73.14 113.3 E 5,35,894 ,734,759 6,44,911,1,000,746, 8.728 34.80 27. 47184,6 60,316 1,901 69, 79520,54 95,382 1713, 2.43 2.14 2.81 2.58 2. 10470.6 104.22 185.12 181. X 36 5.4 41 21 32.88 0. 1 Y 02 .3 21 6.96 -1 13 6 Binary Vari 0.12 39.75 16 26.4 0.08 36.13 0.09 0.21 05.01 0.-80.19 0.07 -115.10.10 6.2 0.93 0. -15.7164 -87.90 0.-9 ables EXREPL 0.07 05 0. 0.12 0.07 0.17 0. 22 EXRELQ 0.02 08 0. RELQ_REPL 0.01 04 0. EXCH 0.11 29 0. CONDITION 0.04 03 0. CONDITION 0.38 62 0. CONDITION 0.58 35 0. BUYEROUT 0.25 16 0. SALELEAS 0.05 05 0. PORTSALE 0.04 02 0. YR1999 0.05 05 0. YR2000 0.18 22 0. YR2001 0.13 15 0. YR2002 0.16 17 0. YR2003 0.23 15 0. YR2004 0.19 17 0. YR2005 0.05 04 0. 0. 0.02 04 0. 0.00 04 0. 0.07 31 0. 0.01 05 0. 0.80 22 0. 0.19 73 0. 0.08 38 0. 0.04 03 0. 0.03 03 0. 0.04 00 0. 0.26 18 0. 0.17 13 0. 0.16 20 0. 0.14 17 0. 0.19 22 0.08 0. 10 0.06 0.03 0.03 0.01 0.21 0.10 _BA 0.03 0.03 _A 0.66 0.70 _AA 0.32 0.28 0.16 0.11 EBACK 0.04 0.03 0.02 0.04 0.04 0.04 0.17 0.16 0.17 0.26 0.18 0.16 0.18 0.17 0.20 0.17 0.06 0.04

PAGE 98

87 T Phoenix Riverside /San Bernardino Sacramento able 13. Continued Los Angeles Oakland Office Obs 1491 Obs 366 Obs 995 Obs 278 Obs 302 Variable Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. PRICE 5,054,013 9,195,275 2,689,489 5,631,661 4,317,847 7,569,004 2,311,880 3,180,745 2,735,394 4,929,805 AGE 34.82 19.74 37.96 22.33 19.42 12.66 20.92 14.65 23.86 19.19 SQFT 36,902 70,484 18,779 40,404 35,792 58,350 22,592 28,558 23,125 40,889 LANDSQFT 46,773 91,730 36,530 79,888 103,293 152,106 68,407 97,102 71,421 134,544 FLOORS 2.47 2.04 1.95 1.40 1.75 1.35 1.79 0.84 1.56 0.90 PARKING 92.56 166.00 45.01 99.50 102.60 1 68.64 76.83 106.87 71.71 119.26 X 34.05 0.14 37.80 0.13 33.50 0.10 34.01 0.22 38.62 0.08 Y -118.28 0.19 -122.12 0.15 -111. 99 0.13 -117.31 0.31 -121.37 0.14 Binary Variables EXREPL 0.11 0.17 0.07 0.25 0.25 EXRELQ 0.08 0.10 0.01 0.07 0.07 RELQ_REPL 0.03 0.09 0.01 0.07 0.12 EXCH 0.22 0.36 0.09 0.39 0.43 CONDITION_BA 0.02 0.04 0.02 0.01 0.01 CONDITION_A 0.74 0.88 0.71 0.83 0.60 CONDITION_AA 0.24 0.08 0.27 0.17 0.39 BUYEROUT 0.04 0.01 0.30 0.06 0.03 SALELEASEBACK 0.04 0.02 0.02 0.03 0.05 PORTSALE 0.01 0.00 0.02 0.02 0.00 YR1999 0.05 0.07 0.03 0.03 0.04 YR2000 0.13 0.20 0.20 0.12 0.22 YR2001 0.14 0.13 0.14 0.14 0.17 YR2002 0.24 0.17 0.15 0.22 0.22 YR2003 0.21 0.22 0.20 0.24 0.18 YR2004 0.20 0.15 0.21 0.15 0.10 YR2005 0.05 0.05 0.08 0.10 0.07

PAGE 99

88 Table 13. Continued San Diego Seattle Tampa Tucson Washington Office Obs 620 Obs 692 Obs 528 Obs 264 Obs 886 riable Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Va PR SQ P X Y ICE 6,503,274 10,300,000 3,991,226 7,427,023 3,382,267 7,114,803 1,713,384 4,228,926 14,800,000 15,900,000 AGE 24.00 16.93 32.58 24.04 26.63 18.95 20.73 20.87 34.37 30.95 FT 43,849 74,519 27,122 47,905 35,616 66,332 15,370 37,476 93,519 108,583 LANDSQFT 81,662 134,771 59,428 102,161 92,706 152,931 42,841 85,756 114,823 199,195 FLOORS 2.46 1.85 2.10 1.58 2.07 1.91 1.48 1.49 4.70 2.99 ARKING 119.78 180.12 68.52 122.62 87.46 163.62 47.23 115.41 168.58 220.65 32.87 0.16 47.55 0.24 27.92 0.16 32.25 0.09 38.92 0.10 -117.15 0.09 -122.29 0.12 -82.54 0.23 -110.93 0.05 -77.14 0.16 Binary Variables E XREPL 0.20 0.17 0.03 0.13 0.05 E R EX C C C SA Y Y Y Y Y Y Y XRELQ 0.08 0.11 0.02 0.02 0.02 ELQ_REPL 0.06 0.04 0.00 0.03 0.01 CH 0.34 0.32 0.06 0.17 0.08 ONDITION_BA 0.01 0.02 0.04 0.05 0.04 ONDITION_A 0.71 0.67 0.47 0.55 0.60 ONDITION_AA 0.28 0.31 0.49 0.41 0.36 BUYEROUT 0.10 0.09 0.14 0.13 0.44 LELEASEBACK 0.05 0.07 0.03 0.04 0.04 PORTSALE 0.03 0.02 0.04 0.02 0.06 R1999 0.03 0.06 0.05 0.03 0.03 R2000 0.15 0.13 0.18 0.09 0.16 R2001 0.18 0.18 0.20 0.14 0.17 R2002 0.19 0.14 0.18 0.19 0.13 R2003 0.19 0.18 0.13 0.19 0.18 R2004 0.19 0.26 0.21 0.27 0.25 R2005 0.07 0.05 0.04 0.11 0.07

PAGE 100

89 Approximately 16 percent of office propert y buyers reside ou t-of-state; out of 8,871 office sales 1,418 involved an out-of-state buyer. This is in sharp contrast to apartment sales where only 7 percent of sales were comple ted by out-of-state buyers. The markets with the highest share of out-of-state buyers are: Washi ngton, DC, 44 percent; Las Vegas, 38 percent; Phoenix, 30 percent; and Dallas/Forth Worth, 25 percent. Other markets of potential interest when quantifying the price e ffect of out-of-state buyers include Denver with 16 percent out-of-stat e buyers; Chicago, 11 percent; Tampa, 14 percent; Tucson, 13 percent; San Diego,10 percent; Seattle, 9 percent; Fort Lauderdale, 8 percent; and Los Angeles, 4 percent. Approximately 4 percent of a ll sales are part of sale-l easeback transactions; there are 342 sale-leaseback transactions in the o ffice sample. The market with the highest percentage of sale-leasebacks is Seattle, wh ere this type of transaction represented 7 percent of all sales. Other markets of potenti al interest when quantifying the effect saleleasebacks have on sales price include Los Angeles (56 observations), Washington, DC (32 observations) and San Diego (30 observations). On average, portfolio sales comprise onl y 2 percent of all sales in the office sample; therefore there are onl y 222 portfolio sales in the sa mple. Washington, DC is the market with the highest percentage of portf olio sales, where such transactions are observed in 6 percent of all sales. The onl y other market where I observe a sufficient number of portfolio sales for the purposes of quantifying poten tial price premiums associated with the motivation to bundle seve ral transactions together is Chicago, which has 30 portfolio sales.

PAGE 101

90 Table erest for Office Properties of Sales Price Error e 14. Differences in Mean Pr ices of Control Sample and Identified Groups of Int Office Observations Mean Value Standard T-test Valu Control Group 5,780 3,472,731 96,485 EXREPL 856 3,842,241 202,277 -1.41 EXRELQ 454 3,027,999 261,805 1.26 RELQ_REPL 240 2,824,935 201,544 1.36 239 3,768,159 476,300 -0.61 PORTSALE 79 12,500,000 1,585,380 -10.69 BUYEROUT 1,044 16,200,000 478,294 -41.92 SALELEASEBACK I perform differences of means tests to ex amine differences in selling prices for variables of interest at the aggregate leve l. Table 14 presents the results. The null hypothesis is that the average observed price of the proper ties in the comparison grou which contains sales that have no conditions and are typically motivated, is equal to the average price of the sample composed of replacement property sales (EXREPL group). Similar hypotheses are formed with respect to the rel p, inquished prope rties (EXRELQ), the samp es are prices are not si gnificant. Out-of-state buyer purchases and portfolio sales are associated with very high prices, which are also statistically le containing sales that are part of two separate exchanges (RELQ_REPL), purchases by out-of-state buyers (BUYER OUT), sale-leasebacks (SALELEASEBACK) and portfolio sales (PORTSALE). Table 14 sh ows that replacement office exchang associated, on average, with a 10.6% price premium, but the price difference is not statistically significant. Reli nquished exchanges and properties that are part of two different exchanges have lower average prices than the control sample, but the differences in the mean

PAGE 102

91 significantly different from the average pric e of properties in th e control group. Since portfolio sales is, as their namts ly involve larg er transactions than with single sales. Sale-leaseb nificantly higher average prices than thos e of ntrol gr I perform regressbased on equatr each o5 o43 2 2005 2000 14 13 12 3 2 9 8 2 1 these statistics do not control for size, th e reason for the extreme average value of e sugges that they simp ack transactions are associated with insig the co oup. Next ions estim ating ion (18) fo f the 1 ffice markets )18( 2 2 2 _3 11 10 7 6 5 4 0 P s s s n nn i i iSMDUM YR PORTSALE ACK SALELEASEB BUYEROUT CONDITION LANDSQFT AGE EXRELQ EXRE mLANDSQFT SQFT FLOORS PARKING SQFT AGE REPL RELQ PL LNPRICE ith Similarly to the previous chapter, each of the regressions is estimated using stepwise regression, in which I allow the pr ocedure to select which of the submarket dummies to leave in the final model, based on their contribution to the fit of the model. All other dependent variab les are not subject to the stepwise procedure. Table 15 reports coefficient estimates fo r each variable by market. The reported results are based on regressions in which standard errors are adjusted to account for potential heteroskedasticity. P-values are reported below the coefficient estimates. Rsquared varies by market from 84 percent (Las Vegas and Riverside / San Bernardino) to 92 percent (Washington, DC m odel). Although the R-squares suggest very good fits of the models, they are driven primarily by the strong relationship between price and square footage, and price and lot size It is important to note that in general, non-residential commercial real estate is harder to value th an residential commercial real estate. W office, industrial and retail prop erties, lease structures can be very complicated and vary

PAGE 103

92 significantly across properties. Sale price can be viewed as the sum of the present value of the income that the property is produci ng each year. However, ne t rents are hard determine; they are often determined by comp licated rent step-up pr ocedures. Price also depends on the credit quality of the current tenant, the terms remaining on the current leases (for example 2 years vs. 15 years), th e pro to bability of the current tenant renewing the le therefore any potential factors related to the lease structures of the properties are not captured in the current model. The reported results in Table 15 reveal that the estimated coefficients on the structural attributes are of the predicted si gn and statistically signi ficant in most of the marke ree of n ARKING is positi ee ase, the cost of re-ten anting the property, etc. Such information is not available in this data set and t regressions. The coefficient on AGE is negative and significan t in all but th the market models. The coefficient on AGE2 is positive, as expected, and significant i 10 out of the 15 markets. The coefficients on SQFT, SQFT2, LANDSQFT, and LANDSQFT2 are all positive and highly significant. The coefficient on P ve, but significant in only three of the markets. The coefficient on FLOORS is positive and significant in all but two of the regression models. The coefficient on CONDITION_BA has negative estimated coefficients, which are also significant in thr out of the 15 regressions. The coefficient on CONDITION_AA is positive and significant in approximately 50 pe rcent of the models. The estimated coefficient on the primary variable of interest representing replacement exchange, EXREPL is positive and significant in 14 out of the 15 regressions. Also the estimated coefficients on EXREPL tend to be much larger than with the apartment models. The lowest estimated coefficient is in Oakland, 0.101, while the

PAGE 104

93 Table 1 Dummies by Office Markets Market Chicago Fort Denver LauderLas Vegas Angeles Oakland Phoenix 5. Regression Statistics for OLS Model with Structural Characteristics and Submarket Dallas/ Worth Ft. dale Los Observations 758 333 587 455 316 1491 366 995 EXREP 2 L 0.304 0.263 0.260 0.296 0.250 0.202 0.101 0.19 0.00 0.01 0.00 0.00 0.00 0.00 0.15 0.00 EXREL RELQ_ 63 AGE -0.015 -0.029 -0.024 -0.023 -0.020 -0.015 -0.019 -0.019 0.00 0.00 0.00 0.00 AGE2 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.09 SQFT 0 SQFT2 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 LANDS 0.00 0.00 0.00 0.00 0.00 0.00 0.16 0.00 LANDS 0.000 00 PARKI 00 .06 FLOOR 088 0.00 CONDITION_BA -0.190 -0.122 -0.101 0.148 -0.151 -0.149 -0.126 -0.185 CONDITION_AA BUYER 17 0.211 0.148 0.478 0.207 0.42 0.00 0.31 0.01 0.00 0.05 0.10 0.00 SALELEASEB 0.292 0.82 0.20 0.28 0.41 0.14 0.18 0.84 0.00 PORTS -0.054 YR2000 YR2001 -0.110 -0.234 0.160 -0.035 0.085 0.066 -0.096 0.29 0.16 0.09 0.77 0.20 0.53 0.31 Q 0.116 0.101 0.095 0.360 0.211 0.079 0.039 -0.250 0.26 0.61 0.21 0.06 0.06 0.05 0.62 0.06 REPL 0.020 0.192 0.289 0.341 0.208 0.263 0.279 -0.1010.94 0.38 0.00 0.00 0.08 0.00 0.00 0. 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.31 0.00 0.00 0.019 0.020 0.024 0.019 0.018 0.028 0.026 0.020.00 0.00 0.00 0.00 0.00 0.00 0.00 0.000.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 QFT 0.004 0.003 0.002 0.005 0.006 0.003 0.003 0.005 QFT2 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.00 0.00 0.00 0.00 0.00 0.02 0.72 0.NG 0.000 0.000 0.000 0.000 0.001 0.000 0.001 0.00.71 0.55 0.62 0.33 0.09 0.38 0.35 0S 0.071 0.053 0.058 0.164 0.167 0.076 0.130 0.0.00 0.12 0.00 0.00 0.00 0.00 0.00 0.07 0.39 0.21 0.33 0.16 0.03 0.41 0.08 0.059 0.033 0.115 -0.011 0.075 0.075 0.069 0.063 0.24 0.65 0.03 0.87 0.27 0.02 0.37 0.18 OUT 0.063 0.371 0.072 0.2 ACK -0.026 0.227 0.099 0.110 0.282 0.092 -0.016 ALE 0.004 -0.143 -0.033 -0.190 0.762 0.527 0.97 0.32 0.80 0.25 0.00 0.00 0.76 -0.187 -0.121 -0.035 -0.153 0.002 -0.028 -0.029 -0.009 0.07 0.46 0.70 0.19 0.99 0.68 0.76 0.92

PAGE 105

94 Table 15. Continued Dallas/ Market Chicago W Denver Ft. LauderLas Vegas Los As Oakland Phoenix Fort orth dale ngele YR2002 -0.037 0.092 0.048 0.169 0.125 -0.053 -0.153 0.004 0.72 0.37 0.33 0.98 0.60 0.01 0.22 0.58 YR2003 0.192 -0.184 0.034 0.205 0.025 0.379 0.248 0.071 0.07 0.27 0.73 0.09 0.81 0.00 0.01 0.45 YR2004 0.173 -0.029 0.242 0.303 0.291 0.552 0.374 0.321 0.10 0.85 0.01 0.01 0.00 0.00 0.00 0.00 YR2005 0.454 -0.081 0.050 0.517 0.370 0.751 0.376 0.384 0.00 0.72 0.66 0.00 0.00 0.00 0.00 0.00 CONS T not reported) -squared 13.838 13.647 12.772 13.591 12.927 14.268 13.012 13.210 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 SDUMi ( R 0.90 0.91 0.89 0.87 0.84 0.88 0.87 0.89 Ard errors are adjupoterosit y. P are repder thcients and bold and italics. Depe ndent Variable is LE Lselling EXREP nary varet eque if transaction represents sale olacemerty LQ variablual to oansactresent a relinquished property; REPL variaquaif trans presesale uishe property and purchase of a rment py. AG e of tding(s) s; AGE2 squar FT footage of total improvemen FT2 Square foottota l improvements squared; LANDSQFT Square footage of L quare fooand PARKING Parfined ar of paces RS Number of floors ; CONDITIONi P condition of the p based on inspectio nategoride be e and above average. The omtegoryrage; ROUT ary varit equal tf buyerout of S Binarle seo on tiart of back E y variable set equal tonsaction w of posale; eriods f99 thro05. Ear is inc a binary variable except 1999, which is suppressed; S Biniable sig the submarket in which the prop l ned by CoSrs. ll standa sted for ntial hete kedastic -values orted un e coeffi estimate ar e in NPRIC ent prop og of EXRE price; Binary L Bi e set eq iable s ne if tr al to on ion rep f a rep s sale of LQ_RE Binary ble set e l to one a ction re nts both of relinq d eplace ts; SQ ropert E Ag age of he buil in year Age ed; SQ Square land ; ANDSQFT2 S tage of l squared; king, de s numbe arking sp ; FLOO hysical itted ca roperty BUYE The c able se es inclu o one i low average, lives averag state ; is ave Bin ALELEASEBACK one if tra y variab t equal t e if tr ansac on was p sale-lease ; PORTSAL Binar as part rtfolio YR n Yearly DUM time p ary var rom 19 gnifyin ugh 20 ch yea luded erty is as i ocated, as defi tar broke

PAGE 106

95 TM Bernardin cra D e a g able 15. Continued arket Riverside /San o Sa mento San iego S attle Tamp Tucson Washin ton Oons 30 6928 6 5 0.230.0.176 280 4 214 bservati 278 2 620 52 264 88 E XREPL 0.20 5 266 0. 0.30 0. 0.00 0.000.00 .02 01 2 0.030.0.114 .002 4 109 80 0.0.03 99 40 EPL 7 0.380.0.393 865 9 72 0.000.00 .00 46 5 -0.0-0-0.017.014 5 03 0.0.00 00 0 0.000.000.000 000 0 000 0.72 0.49 0.00 0.00 0.01 .20 0.77 0.50 0.47 0.31 0.17 0.08 0.03 0.90 FLOORS 0.122 0.157 0.051 0.073 0.093 0.023 0.076 0.01 0.00 0.01 0.02 0.00 0.58 0.00 CONDITION_BA 0.088 0.111 -0.139 -0.097 -0.003 -0.032 -0.233 0.85 0.80 0.50 0.43 0.97 0.77 0.08 CONDITION_AA 0.095 0.136 0.160 0.157 0.014 0.143 0.132 0.23 0.05 0.00 0.00 0.74 0.06 0.00 BUYEROUT -0.206 -0.024 -0.040 0.082 0.102 0.117 0.105 0.12 0.89 0.53 0.30 0.13 0.23 0.02 SALELEASEBACK 0.409 -0.094 -0.148 0.051 0.128 0.173 0.228 0.10 0.33 0.08 0.44 0.33 0.11 0.01 PORTSALE -0.091 3.739 -0.059 -0.225 0.113 0.064 0.064 0.61 0.00 0.58 0.09 0.36 0.79 0.35 YR2000 0.138 0.051 -0.091 0.053 -0.320 0.034 0.138 0.36 0.66 0.43 0.50 0.00 0.76 0.13 YR2001 0.234 0.133 0.085 0.033 -0.123 0.025 0.210 0.14 0.28 0.45 0.66 0.25 0.77 0.02 0 .00 0 0.00 0. E XRELQ -0.02 3 139 0 -0.00 0. R 0.81 0. 02 0. 0.98 0. ELQ_R 0.110.25 2 0 262 .00 0.0 -0.000.96 -0.10. A GE -0.00 1 0 .020 -0 -0.00 -0. 00512 A 0.28 0.0.00 00 0. 0.13 00 0.0. GE2 0 0 0. 0.0 SQFT 0.041 0.033 0.030 0.032 0.024 0.049 0.022 0.00 0.00 0.00 0.00 0.00 0.00 0.00 SQFT2 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.00 0.00 0.00 0.00 0.00 0.00 0.00 LANDSQFT 0.002 0.001 0.001 0.002 0.004 0.005 0.002 0.16 0.25 0.17 0.06 0.00 0.00 0.00 LANDSQFT2 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.38 0.38 0.32 0.61 0.00 0.00 0.01 PARKING 0.001 0.000 0.000 -0.001 0.000 0.002 0.000

PAGE 107

96 Table 15. Continue Market Bernardino Sacramento Dieg Seattle Tampa Tucson Washington d Riverside /San San o YR2002 0.367 0.218 0.113 0.160 -0.153 0.018 0.301 0.02 0.07 0.33 0.03 0.14 0.85 0.00 YR2003 0.454 0.276 0.240 0.181 0.045 -0.060 0.302 0.00 0.02 0.03 0.02 0.71 0.54 0.00 YR2004 0.677 0.317 0.389 0.236 0.120 0.146 0.418 0.00 0.02 0.00 0.00 0.26 0.07 0.00 YR2005 0.747 0.530 0.609 0.428 0.296 0.107 0.523 0.00 0.00 0.00 0.00 0.09 0.32 0.00 CONST 11.735 13.410 14.151 12.887 12.900 12.579 13.722 0.00 0.00 0.00 0.00 0.00 0.00 0.00 SDUM eporte i (not d) d r R-square 0.84 0.85 0.90 0.87 0.88 0.86 0.92 Aerrors are adjustetential hetsticit y.s tedcnt estim ar e in bold and italics. Depe ndent Vais LNPRICE of sellie; E Biable set equal to t sents sale of ent prop RELQ y vset oneaction ts sale of a operty; RELQ Binary set eqe i tionntse of red property and purchase of a repnt property Age ofildi yea 2 uared; SQFT Square f provement Square of tota emuar DSQ quare ff land ; L uare footagd squared; G Pefumarkes; FL Number of floors ; CONDITIONi Phyndition of the property bainsp The ries below a, average a age. The omittory is ave YERO nae sto yer livf state ; S CK Binary t equal ttr ansac p-l; PE Biable set equal to one if transaction was portfolio sn Yea perom 1roug Each ycluded as a ble except 1999is suppres Mi Briifyubm whichperty is l fined by CoStar ll standard d for po eroskeda P-value are repor under the oefficie ates and riable Log ng pric XREPL inary var one if ransaction repre relinquished pr a replacem _REPL erty. EX variable Binar ual to on ariable f transa c equal to represe if trans both sal represen linquishe laceme AGE the bu ng(s) in rs; AGE Age sq ootage of total im ANDSQFT2 Sq s; SQFT2 e of lan footage PARKIN l improv arking, d ents sq ined as n ed; LAN ber of p FT S ing spac ootage o OORS sical co sed on ectio n. catego include verage nd above aver ALELEASEBA ed categ variable se rage; BU o one if UT Bi tion was ry variabl art of sale et equal easeback one if bu ORTSAL es out o inary var part of ale; YR rly time iods fr 999 th h 2005. ear is in binary varia ocated, as de which brokers. sed; SDU inary va able sign ing the s arket in the pro

PAGE 108

97 highest coefficient is estimated in the regression based on sales in the Chicago This provides strong evidence that buyers of replacement office properties are paying statistical ificant price prem xamined in this study. The coefficient estima EXRE e e mand sant in models. In th the mo XREL eg annet (Ph) the n EXRELQ is negatively significant. In a ates of EXRELQ are substantially sm than coefficient estiman EXREPL. These findings are generally consistent with our expectation for EPL to bociateh any nificant positivium. In addition, the recorded positive coefficients present ve exchange. The coefficient on RELQ_REPL is positive and significant in 9 out of the 15 markets. However, the magnitude of this coeffi cient tends to be larger than the coefficient on replacement exchanges. The coefficient on the variable re presenting out-of-state buyers, BUYEROUT is positive and significant in multiple markets. The coefficient on BUYEROUT is significant in all four of the markets with highest shares of buyers that are out of state: Washington, DC (44 percent); Las Vegas (38 percent); P hoenix (30 percent); a nd Dallas/Forth Worth metropolitan area, 0.304. ly sign iums in the 15 office m arkets e tes of LQ ar positiv in 12 rkets a ignific 5 ree of dels E Q is n ative, d in o marke oenix coefficient o ddition, the coefficient estim aller tes o EXR not e ass d wit sig e prem evidence that relinquished exchanges are prop erties with higher cap ital appreciation, and tend to be more expensive. Therefore, such positive effects partially offset any negati effects related to relinquished exchanges. As previously note d, no such issues exist with replacement exchanges, and therefor e all else equal the coefficient on EXREPL can be directly associated with a price premium paid when the sale is part of a replacement

PAGE 109

98 (25 percent). The coefficient on BUYEROUT is also significant in Fort Lauderd Los Angeles, which have 8 and 4 percent of out-of-state buyers, respectively. These findings add to the evidence that out -of-state buyers pay price premiums. The coefficient esti ale and mate on the variable representing sale-leaseback transactions, SALE ed io sales. In addition, in the tw o markets where portfolio sales are numerous, Wash is LEASEBACK is generally positive and significant in three markets: Phoenix, Riverside / San Bernardino and Washi ngton, DC. In one market, San Diego, SALELEASEBACK is significant but negative. The negati ve coefficient of the variable in Seattle is contrary to our exp ectation. One possible explanatio n is that the result may be influenced by an outlier, since it is base d on only 30 observations. Generally, the results present evidence supporting our hypothesis that sale-leaseb acks are associated with higher transaction prices. The estimated coefficient on PORTSALE is positive and significant in three markets (Las Vegas, Los Angeles, and Sacram ento) and negatively significant in Seattle. However, none of these markets have a large number of portfolio sales from which to draw any meaningful conclusi ons about observed price increa ses (decreases) associat with portfol ington, DC and Chicago, no significant results are observed. Therefore, there only weak evidence that portfolio sales of office properties are a ssociated with price premiums. The estimated year dummies present a mixed picture that is quite different than with apartment properties. Sa les in years 2000 and 2001 tend to have lower prices than sales in 1999. The dummy coefficients show th at in 2000, prices in 9 markets were lower than prices, all else equal, in 1999. In 2001, dummy variables in 5 markets are negative,

PAGE 110

99 although insignificant. The year 2002 marks the beginning of a trend of substantial price appreciation. This trend is continue d throughout 2003-2005, with year dummies significant in 9, 13 and 12 markets, respectivel y. One exception to th is gen eral trend in price but they are r e xpansion rather than a simple linear form is entered in the equat Table 16 presents the regression results for Model II. Controlling for absolute location in addition to relative location does not lead to any signi ficant changes in the estimated coefficients relative to OLS Mode l I. Moreover, the fitness of the models improves only modestly with an increase in R-sq uared ranging from half to one percent. With respect to the key variables of interest, the results are re-confirmed. The coefficient on the variable indicating that the sale is part of a replacement exchange, 2 2000 2 6 5 4 3 2 10 appreciation after 2002 is notable. Dallas/Forth Worth is the only market in which all year dummies are negative, and ther efore no price apprec iation is observed. Submarket dummy variables are not reported in the regression outputs, the main instrument used to control for lo cation within the metr opolitan area and their inclusion improves dramatically the fit of the models. Next, I examine OLS Model II, which adds variables that control for the absolute location of the property by using a third orde r expansion of the la titude and longitude coordinates. The third orde ion to effectively draw a price surf ace based on location. The resulting model specification has the following form )20( ),( 2 2 2 _, 43 2005 14 13 12 3 11 10 9 8 7 00 rr lk lk P s snn i iYX SMDUM YR PORTSALE ACK SALELEASEB BUYEROUT CONDITION FLOORS PARKING LANDSQFT LANDSQFT SQFT SQFT AGE AGE REPL RELQ EXRELQ EXREPL LNPRICE kl s n i m

PAGE 111

100 EXREPL is positive and significant in 14 of the 15 regressions. This provides additional evidence that buyers of replacement properties are paying statistically significant premiums in the majority of the markets. Re sults with respect to relinquished exchanges EXRELQ and the combination of relinquish ed and replacement exchanges, RELQ_REPL remain the same. price , There is slight improvement in the P-val n ide / ed fo r Phoenix, Washington, DC and San Diego. Results remain largely unchanged also for the variable representing portfolio sales, PORTSALE Based on OLS Model II, the percentage sa les price changes co rresponding to the estimated and statistically significant co efficients for replacement exchanges ( EXREPL ), relinquished exchanges ( EXRELQ ) combined relinquished and replacement exchanges ( RELQ_REPL), purchases by out-of-state buyers ( BUYEROUT ), sale-leaseback ( SALELEASEBACK ) and portfolio sales ( PORTSALE) for each market are presented in Table 17. Table 17 shows that the percen tage price effects associat ed with being part of a replacement property exchange ranges from 18.93 percent in Seattl e to 36.09 percent in The coefficient on the variable representing whether the buyer is from out of state, BUYEROUT remains positive and significant in Wa shington, DC, Las Vegas, Phoenix Dallas/Forth Worth, Fort Lauderdale and Los Ange les. ues, which are reduced by approximately one percent. The magnitude of the coefficient on the variable representing whet her the sale is part of a sale-leaseback transaction, SALELEASEBACK is decreased in most markets by about 0.01 0.05. Whe additional controls are added the significance of this coefficient disappears in Rivers San Bernardino, but remains unchang

PAGE 112

101 Chicago. The average price premium paid when a property is used to complete a replacement exchange is 28.3 percent. This premium is larger than any predicted exchange benefits, as indicated by the simu lation analysis in Chapter 5. In comparison, the simulation analysis predic ts that when the holding period, both for the re and replacement property, does n linquished ot exceed 5 years, observed incremental NPV from a 1031 eriod arger than the expected tax bene fit, especially if they have a short-term inves ociated ement and relinquished exchange ra nges from 28.66 percent in San Diego to 48.27 exchange is at most 4 percent. With 10-year holding periods, incremental NPV ranges from 2.92 percent to 8.46 percent (whe n assuming an annual appreciation of 20 percent, and 20 percent capital gain tax rate). With 20-year holding periods the maximum incremental value from using an exchange is below 12 percent. Finally, with 39-year holdi ng period, which is also the depreciation recovery p for non-residential commercial real estate the maximum net benefit from using an exchange, based on our simulation analysis, is below 14 percent. Therefore, based on an average prem ium of 28.3 percent for the replacement exchanges, the results indicate that invest ors pay a premium in exchanges that is significantly l tment horizon. In three of the markets I also observe pr emiums related to relinquished exchanges. The implied percentage effects of relinqui shed exchanges on selling price are, on average, 11.98 percent and they are also signif icantly smaller than the effects ass with replacement exchanges. The price premium associated with a pr operty being part of both replac percent in Sacramento. The average price premium associated with this type of transaction is 35.38 percent.

PAGE 113

102 Table 16. Regression Statistics for OLS Model wMarket Dallas/ Worth Ft. dale ith Structural Characteristics, Submarket Dummies and Longitude, Latitude Coordinates by Office Markets Chicago Fort Denver LauderLas Vegas Los Angeles Oakland Phoenix ObservEXREPL ations 758 333 587 455 316 1491 366 995 0.31 0.27 0.26 0.30 0.25 0.20 0.09 0.21 EXREL 0.25 0.70 0.20 0.04 0.07 0.03 0.64 0.06 RELQ_ 0.93 0.23 0.00 0.00 0.07 0.00 0.00 0.60 AGE 00 0.00 0.00 AGE2 00 0.00 0.00 SQFT 00 SQFT2 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 LANDSQFT 01 00 LANDS 0.00 0.00 0.01 0.00 0.00 0.00 0.02 0.68 0.00 PARKI 0.65 0.72 0.75 0.38 0.12 0.33 0.42 0.06 FLOOR 0.06 0.06 0.16 0.17 0.07 0.13 0.08 CONDITION_ 0.06 CONDITION_AA 0.05 0.04 0.12 -0.01 0.06 0.08 0.07 0.07 BUYEROUT 0.06 0.36 0.09 0.21 0.22 0.15 0.46 0.20 SALELEA PORTS YR2000 -0.03 -0.16 0.00 -0.03 -0.04 -0.01 0.07 0.38 0.78 0.17 0.97 0.65 0.71 0.94 0.00 0.01 0.00 0.00 0.00 0.00 0.18 0.00 Q 0.12 0.08 0.10 0.39 0.20 0.08 0.04 -0.25 REPL 0.02 0.29 0.31 0.32 0.21 0.26 0.29 -0.11 -0.02 -0.03 -0.02 -0.02 -0.02 -0.02 -0.02 -0.02 0.00 0.00 0.00 0.00 0.00 0.0.00 0.00 0.00 0.00 0.00 0. 0.00 0.00 0.00 0.00 0.38 0.00 0.00 0.03 0.02 0.02 0.02 0.02 0.02 0.03 0.03 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0. 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.0.00 0.00 0.00 0.00 0.00 0.00 0.14 0.QFT2 0.00 0.00 0.00 0.00 0.00 0.00 0.00 NG 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 S 0.07 0.00 0.10 0.00 0.00 0.00 0.00 0.00 0.00 BA -0.19 -0.11 -0.08 0.16 -0.15 -0.13 -0.13 -0.19 0.06 0.43 0.33 0.29 0.16 0.07 0.39 0.29 0.62 0.02 0.93 0.38 0.02 0.37 0.12 0.42 0.00 0.23 0.01 0.00 0.04 0.11 0.00 SEBACK -0.03 0.18 0.11 0.11 0.27 0.09 -0.03 0.26 0.77 0.33 0.24 0.38 0.18 0.20 0.72 0.00 ALE 0.00 -0.17 -0.02 -0.17 0.79 0.53 -0.040.98 0.22 0.86 0.27 0.00 0.00 0.82 -0.19 -0.14

PAGE 114

103 Table 16. Continued Dallas/ Ft. Market Chicago W Denver LauderLas V Los As Oakland Phoenix Fort orth dale egas ngele YR2001 -0.11 0.17 0.08 0.07 -0.10 -0.24 -0.04 0.29 0.15 0.07 0.75 0.20 0.53 0.26 YR2002 -0.03 -0.15 0.10 -0.02 0.05 0.17 0.13 -0.06 0.76 0.36 0.30 0.88 0.60 0.01 0.21 0.49 YR2003 0.19 -0.19 0.05 0.19 0.03 0.38 0.25 0.05 0.07 0.25 0.58 0.12 0.78 0.00 0.01 0.53 YR2004 0.17 -0.03 0.25 0.29 0.30 0.55 0.37 0.32 0.10 0.82 0.01 0.02 0.00 0.00 0.00 0.00 YR2005 0.46 -0.10 0.05 0.51 0.37 0.73 0.36 0.38 0.00 0.66 0.65 0.00 0.00 0.00 0.01 0.00 X3 0.01 -0.03 0.00 0.06 0.00 -0.03 0.00 0.00 0.33 0.15 0.96 0.27 0.48 0.04 0.96 0.89 Y3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.34 0.14 0.93 0.27 0.49 0.05 0.95 0.94 XY2 0.00 0.01 0.00 -0.02 -0.01 0.01 0.00 0.00 0.33 0.15 0.94 0.27 0.48 0.05 0.95 0.87 CONST 18.73 72.70 -0.91 -3.60 -41.76 -8.45 19.56 58.87 0.04 0.02 0.95 0.90 0.28 0.53 0.69 0.00 R-squared 0.90 0.91 0.89 0.87 0.85 0.88 0.87 0.89 All standard errors are adjus teroskedasticit y. P-v repnder tficientates an n ted for poten tial he alues are orted u he coef estim d ar e i b Depe ndent Variable is LE Lelling REP ary v set eqone if t resents sale olacemerty. LQ Bariableual toaprese of a relinquished property; REPL variaequal t transa epreseth salenquish ppurchase of a rment py. AG e of theng(s) i; AGE e squ QFT re f al improvemen FT2 Square foot l improvements squared; LANDSQFT Sqotage ; LANDSQFT2 Square foot s; PARKING Parkiined asr of p space ORS Number oi Pcondthe p based on inspectio ntegolude verag a e omegoryage; y variaaif bus out SALELEASEBACK Binaryle setto one sactionart of ssebac TSALE y variable set e ion w of posale; riods f99 thr005. Ear is included as a t 1999, which is suppressed; S Binale sig the submarket in which the pris lated, as defined by CoStaers; X de ofty; Y de of py. old and italics ransaction rep NPRIC ent prop og of s EXRE price; EX inary v L Bin set eq ariable one if trans ual to ction re f a rep nts sale LQ_RE Binary ble set o one if ction r nts bo of reli ed roperty and ootage of tot eplace ts; SQ ropert E Ag age of tota buildi n years 2 Ag ared; S uare fo Squa of land age of land quared ng, def numbe arking s; FLO f floors ; CONDITION nd above average. Th hysical itted cat ition of is aver roperty BUYEROUT The ca ble set equ ries inc l to one below a yer live e, average of state ; Binar variab equal if tr an was p ale-lea k; POR Binar qual to one if transact binary variable excep as part rtfolio YR n Yearly DUM time pe ry variab rom 19 nifying ough 2 ach ye i proper operty oc r brok latitu longitu ropert

PAGE 115

104 TRiverside Bernardin raSaSe TT able 16. Continued Market o Sac /San mento n Diego attle ampa ucson W ashington Observa tions 302 69252XREPL 1 0.23 0.0.170.0. 278 620 28 64 886 E 0.2 27 28 30 0.21 0.00 0.00 00.0.0.XRELQ 2 0.04 0.0.0.0.0 0.78 0.0.0.0.ELQ_REPL 5 0.39 0.0.0.-0-3 0.00 0.0.0.0.GE 0 -0.0-0-0-0-07 0.05 0.0.0.0.GE2 0 0.00 0.0.0.0.1 0.68 0.0.0.0.QFT 4 0.03 0.0.0.0.0 0.00 0.0.0.0.QFT2 0 0.00 0.0.0.0.0 0.00 00.0.0.ANDSQFT 0 0.00 0.0.0.0. 0.34 000.0NDSQFT2 0 0.00 0.0.0.0.1 0.44 0.0.0.0.0.11 0.10 0.11 0.19 0.97 0.57 0.41 0.12 0.33 0.01 SALELEASEBACK 0.38 -0.11 -0.15 0.06 0.11 0.17 0.24 0.12 0.27 0.07 0.38 0.39 0.13 0.01 PORTSALE -0.16 3.72 -0.07 -0.23 0.11 0.07 0.07 0.31 0.00 0.52 0.08 0.39 0.77 0.34 YR2000 0.09 0.04 -0.11 0.05 -0.31 0.03 0.14 0.56 0.72 0.34 0.53 0.00 0.75 0.13 .00 00 02 00 0.01 E 0.0 14 11 00 04 0.10 0.8 02 03 98 80 0.44 R 0.1 25 38 86 .02 0.18 0.1 00 00 00 92 0.46 A 0.0 1 .02 .02 .02 .01 0.00 0.5 00 00 00 05 0.14 A 0.0 00 00 00 00 0.00 0.9 00 00 01 08 0.72 S 0.0 03 03 02 05 0.02 0.0 00 00 00 00 0.00 S 0.0 00 00 00 00 0.00 0.0 .00 00 00 00 0.00 L 0.0 00 00 00 00 0.00 0.14 .25 .08 00 .00 0.00 LA 0.0 00 00 00 00 0.00 0.3 40 63 00 00 0.02 PARKING 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.38 0.47 0.35 0.17 0.10 0.03 0.95 FLOORS 0.12 0.16 0.05 0.07 0.09 0.02 0.08 0.01 0.00 0.01 0.02 0.00 0.62 0.00 CONDITION_BA 0.06 0.11 -0.13 -0.09 -0.01 -0.02 -0.23 0.90 0.81 0.55 0.44 0.94 0.87 0.08 CONDITION_AA 0.10 0.14 0.16 0.15 0.02 0.15 0.13 0.17 0.04 0.00 0.00 0.73 0.05 0.00 BUYEROUT -0.18 -0.01 -0.04 0.07

PAGE 116

105 Table 16. Continued Market Riverside /San Bernardino Sacramento San Diego Seattle Tampa Tucson Washington YR2001 0.19 0.14 0.06 0.03 -0.11 0.00 0.21 0.20 0.28 0.57 0.70 0.29 0.97 0.02 Y R2002 R2003 R2004 R2005 3 3 Y2 ONST -squared 0.32 0.21 0.10 0.16 -0.15 0.02 0.30 0.03 0.10 0.39 0.04 0.17 0.85 0.00 Y 0.39 0.27 0.22 0.19 0.06 -0.06 0.30 0.01 0.03 0.05 0.02 0.65 0.57 0.00 Y 0.64 0.33 0.37 0.23 0.13 0.14 0.42 0.00 0.02 0.00 0.00 0.23 0.11 0.00 Y 0.71 0.53 0.59 0.43 0.31 0.12 0.52 0.00 0.00 0.00 0.00 0.07 0.26 0.00 X -0.01 -0.05 0.01 -0.02 -0.02 0.77 0.01 0.07 0.27 0.40 Y 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.29 0.78 0.01 0.06 0.28 0.39 0.97 X 0.00 0.00 0.01 -0.01 0.01 0.01 0.00 0.30 0.78 0.01 0.07 0.28 0.40 0.07 C 44.42 59.64 30.05 6.54 16.15 59.63 61.82 0.00 0.22 0.33 0.77 0.33 0.03 0.06 R 0.85 0.85 0.91 0.87 0.88 0.87 0.92 A rors are adjustedtential heteasticit y. P are re undeefficimates a in ll standard er for po rosked -values ported r the co ent esti nd ar e bold and italics. Depe ndent Va LNPRIC of selling EXRE yle sete if transacpresents sale of aement prop XRELQ variabequalif tran represee of a relinquished property; RELQ Binary v set equal if tran reproth selinquis property and purchase of a repnt property Age of tding(s A Age s; SQFT re footage of total improvements; 2 Squaree of tota l improvements squared; LANDSQFT footaged ; L footaged squared; NG Parfinedber og sp OORS er of floors ; CONDITIONi Phyndition of the property basespecti categories include below averagage and above average. The omitteory is aver YEROUT ary vaet equne if bves out ; S inary va set equal t tr ansactiopart oeaseb RTSA inary variable set equal to one if transaction was portfolio sn Yearlyeriod1999 2005 year is included as a binary variable except 1999, which is suppress UMi Biniable ing the submarket h the pr is l d by CoStar X latitude of property; Y tude o. riable is replac E Log erty. E price; Binary PL Binar le set variab to one equal to on saction nts sal tion re _REPL ariable to one sa ction esents b ale of r hed laceme SQFT AGE footag he buil ) in years; GE2 quared Square Squa of lan ANDSQFT2 Square of lan PARKI king, de as num f parkin aces; FL Numb sical co d categ d on in Bin o n. The riable s e, aver of state age; BU al to o uyer li ALELEASEBACK B riable o one if n was f sale-l ack; PO LE B part of ale; YR ed; SD time p ary var s from signify through Each in whic operty ocated, as define brokers; longi f property

PAGE 117

106 Another strong result in the office regressions involves transactions in which the buyer is from a out-of-state. Purchases by out-o f-state buyers are associated with a price premium of at least 11.25 per cent in Washington, DC and up to 43.14 percent in Dallas. n average, theng priceium associated with out-of-state buyers is 23.5 percent. six out of thearkets, efficien the vle in ut-of-state hastical a nd economic significance. This re reserong ence at there is a sicant price pium associated with purchases by out-of-state buyers. Sale-leases are assed with ce discount in San Diego of 14 percent, a price premium of 27 percent in Washington, DC, and a premf 30ent in Phoenix. As noted earlier, the result in Sa go isd on30 oations oly and may be driven by an outlier. Generall y, the sale-leaseback ixed evidence on theence of premiu disc. Finally, with respect to portfolio sales no statistically significa nt results are found s, I was notto s uppo hypothof a p premassod with O selli prem In 15 m the co t on ariab dicat ing that the buyer is o statis sult p nts st vide th gnifi rem back ociat a pri ium o perc n Die base just bserv n results pres ent m exist pr ice ms or ounts Thu able rt the esis ric e ium ciate portfolio transactions. In the next chapter summary statistics and results based on regressions for the commercial retail real estate properties sample are presented.

PAGE 118

107 Table 17. Marginal Effects for Significant Coeffi cients for Variables of Interest in Office RELQ_ REPL BUYER OUT SALE BACK PORT SALE Chicago75836.09% Dallas/ Fort Worth Regressions MarketObs EXREPLEXRELQ LEASE 33331.62% 43.14% Denver58730.20% 36.96% Ft. Lauderdale45534.95% 23.66% Las Veg Sacra 18.93%11.99% Tamp Washington88623.81% 11.25%26.74% as31628.08% 23.99% Los Angeles149122.40%8.76%29.76%16.52% Oakland366 33.25% Phoenix99523.50% 22.44%29.95% Riverside /San Bernardino 27823.70% mento30226.04% 48.27% San Diego 62030.38%15.18%28.66% -14.00% Seattle 692 a 52832.43% Tucson26434.46%

PAGE 119

CHAPTER 9 RESULTS FOR RETAIL PROPERTIES This chapter presents an analysis of the re tail property data and analyses the results from the empirical analysis There are 12,015 observations in the final retail property sample. Table 18 presents the summary statis tics for the variables of interest. Mean values and standard deviations are reported at the aggregate level as well as by market. The first two columns identify the characteristics of an average retail property sale in the full sample, which includes all 15 retail mark ets studied. The average retail selling price is $2,436,535, which is much less than the average office property price of $5,353,894. Based on transaction size as well as square footage of improvements, the observations in the retail sample appear to be more similar to the observations in the apartment sample. The average retail property in the sample ha s 19,327 sq. ft. of improvements compared to 23,034 sq. ft. for the apartment properties sample. However, retail properties include as much as three times more land than improveme nts. The average retail property is built on 77,171 sq. ft. of land. Retail properties, with an average age of 33.83 years, are generally newer than the apartment propert ies, but older than the office properties. In the sample a relatively high proportion of the retail prope rties is in above average condition (22 percent), 70 percent are in av erage condition and 8 percent ar e in fair or poor condition. The average retail property has 1.26 floors and 57 parking spaces. The fifteen metropolitan markets in the retail sample include: Chicago, Dallas/Forth Worth, Denver, Fort Lauderdal e, Houston, Las Vegas, Los Angeles, 108

PAGE 120

109 Table 18. Summary Statistics of Retail Data by Markets Table 18. Summary Statistics of Retail Data by Markets o s/ W v t Ft. Lauderdale All Chicag DallaFortorth Dener Detroi Re O 1 20 s O 9 9 Obs 516 Var Mea D Dev. MeanStDev. Dev. Std. Dev. Mean Std. Dev. tail iable bs1205 Obs 69Ob 566 bs69 Obs 63 nStd.ev. Mean Std. d. Mean Std.Mean PR 335 4,51, 195 3,1,67 1 ,5,217 90,,662,719 7,159,872 AG 7. 4 6. 1 24. 3 9 25.06 14.77 SQFT 7 28,19 5 0 732 32,303 61,435 LAFT 776,1 6150,1 874,5 0524 135,260 232,726 FL . 024 114 1.17 0.48 PAG 1. 7 6. 165 0. 10 79.42 157.83 X 2. 5 26.15 0.10 Y 1 4. -4 6. 410 0. 8 80.18 0.07 Binary Variales ICE E NDSQ OORS RKIN b 2,46,5,89086 450,804 3,32,5,41707 6711,72,30,104 4,82539 195,02,1695 3 33.83 2388.20 33.18 1014.85 28.50 292.41 23.5 19,327 39,732 14,46 30,561 6 50,00 20,350 42,93 14,70 ,683 ,171 1561 45,60 1,115 1467 210,166 ,184 1619 81,32 1,76 1.26 0671.58 0.95 106. 1.15 048. 0.46 57.19 1299 34.39 89.79182.6 71.85 1420 57.90 4.94 37.04 5.3641.92 0.25 387 0.1 39.76 0.1742.43 0.26 -06.74 14287.8 0.20 -996 0.2 -05.013 -83.2 0.26 EX 0. 03 0.04 REPL 0.13 0.06 12 0.17 0. EX 02 2 REEPL 00 1 EX 05 7 COIO 03 COIO 88 COIO 09 BUU 10 SALESK 02 POLE 05 YR 08 6 YR 19 6 YR 19 9 YR 19 3 YR 20 4 YR 13 9 YR 0. 02 0.02 RELQ LQ_R CH NDITN_BA NDITN_A NDITN_AA YEROT LEAEBAC RTSA 1999 2000 2001 2002 2003 2004 2005 0.06 0.06 002 0.08 0. 0.0 0.03 0.01 001 0.05 0. 0.0 0.22 0.13 015 0.30 0. 0.0 0.08 0.09 007 0.09 0. 0.0 0.70 0.77 044 0.59 0. 0.7 0.22 0.15 050 0.32 0. 0.1 0.15 0.06 029 0.17 0. 0.1 0.03 0.03 005 0.04 0. 0.0 0.03 0.01 008 0.05 0. 0.0 0.05 0.05 005 0.04 0. 0.0 0.18 0.19 018 0.17 0. 0.2 0.17 0.27 011 0.21 0. 0.1 0.18 0.18 017 0.21 0. 0.1 0.20 0.18 028 0.14 0. 0.1 0.17 0.11 016 0.15 0. 0.1 0.05 0.02 05 0.07 0. 7 7 6 8 2 4

PAGE 121

110 Houston Las Vegas Los Angeles Oakland Phoenix Retail Obs 479 Obs 426 Obs 2150 Obs 508 Obs 1208 Variable Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. PRICE 3,386,795 6,043,097 3,003,188 4,419,804 2,497,678 4,973,999 1,764,776 3,895,942 3,325,548 5,465,216 AGE 17.14 15.54 13.83 15.07 40.45 23.85 47.05 24.71 16.44 15.12 SQFT 36,343 61,370 17,496 32,437 16,074 34,442 11,441 21,573 24,755 42,493 LANDSQFT 148,025 217,303 77,962 120,109 44,441 102,719 36,064 84,461 123,656 190,970 FLOORS 1.09 0.51 1.03 0.16 1.23 0.56 1.29 0.55 1.04 0.32 PARKING 104.27 169.06 91.97 135.59 41.20 95.61 31.66 87.70 69.43 135.55 X 29.80 0.16 36.13 0.10 34.05 0.15 37.81 0.13 33.49 0.14 Y -95.44 0.17 -115.14 0.14 -118.25 0.19 -122.13 0.18 -112.00 0.19 Binary Variables Table 18. Continued EXREPL 0.16 0.30 0.14 0.13 0.05 EXRELQ 0.03 0.06 0.08 0.09 0.02 RELQ_REPL 0.02 0.03 0.04 0.03 0.01 EXCH 0.21 0.39 0.26 0.25 0.08 CONDITION_BA 0.09 0.10 0.08 0.12 0.05 CONDITION_A 0.61 0.33 0.80 0.81 0.61 CONDITION_AA 0.30 0.57 0.12 0.07 0.35 BUYEROUT 0.27 0.52 0.03 0.03 0.47 SALELEASEBACK 0.04 0.04 0.01 0.02 0.04 PORTSALE 0.03 0.02 0.02 0.01 0.02 YR1999 0.08 0.04 0.04 0.04 0.03 YR2000 0.13 0.17 0.16 0.22 0.18 YR2001 0.11 0.13 0.13 0.12 0.17 YR2002 0.17 0.17 0.23 0.17 0.16 YR2003 0.25 0.18 0.20 0.28 0.21 YR2004 0.24 0.25 0.17 0.13 0.19 YR2005 0.02 0.07 0.06 0.03 0.07

PAGE 122

111 Table 18. Continued Riverside /San Bernardino San Diego San Francisco Seattle Tucson Retail Obs 429 Obs 638 Obs 339 Obs 943 Obs 406 Variable Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. PRICE 3,443,945 6,400,585 3,104,985 5,680,510 1,720,856 3,461,331 2,269,765 4,717,861 1,981,372 3,567,048 AGE 20.09 17.56 31.03 20.72 73.74 27.28 37.34 27.28 23.62 17.43 SQFT 28,994 53,552 21,303 43,002 8,079 19,008 16,935 36,226 17,724 36,612 LANDSQFT 122,360 206,666 78,581 151,654 9,960 62,657 66,683 139,353 94,710 176,835 FLOORS 1.08 0.28 1.23 0.66 2.27 1.08 1.23 0.56 1.06 0.48 PARKING 83.40 147.94 62.54 132.70 5.89 42.48 55.14 122.77 70.46 131.64 X 34.01 0.25 32.87 0.20 37.73 0.09 47.56 0.29 32.23 0.09 Y -117.20 0.47 -117.12 0.12 -122.41 0.06 -122.29 0.13 -110.93 0.07 Binary Variables EXREPL 0.28 0.21 0.12 0.21 0.19 EXRELQ 0.05 0.09 0.18 0.08 0.03 RELQ_REPL 0.07 0.06 0.07 0.04 0.02 EXCH 0.40 0.36 0.37 0.33 0.24 CONDITION_BA 0.01 0.05 0.09 0.13 0.15 CONDITION_A 0.75 0.83 0.84 0.61 0.67 CONDITION_AA 0.24 0.12 0.07 0.26 0.18 BUYEROUT 0.05 0.04 0.02 0.10 0.30 SALELEASEBACK 0.02 0.01 0.01 0.03 0.05 PORTSALE 0.01 0.02 0.02 0.04 0.10 YR1999 0.04 0.04 0.10 0.05 0.02 YR2000 0.10 0.14 0.40 0.14 0.14 YR2001 0.10 0.18 0.10 0.20 0.11 YR2002 0.22 0.18 0.11 0.18 0.14 YR2003 0.24 0.24 0.15 0.21 0.20 YR2004 0.21 0.17 0.10 0.19 0.26 YR2005 0.09 0.06 0.05 0.03 0.12

PAGE 123

112 Oakland, Phoenix, Riverside/San Bernardino, San Diego, San Francisco, Seattle and Tucson. Los Angeles is once again the largest market represented in the sample, based on number of observations (2,150). The second la rgest represented market is Chicago, with 2,069 retail sales. The smallest sample represents the San Francisco metropolitan area with 339 observations. Based on inspection of the summary statis tics by market reported in Table 18, I notice that property characteristics display s ubstantial variation by market. For example, properties in some markets tend to be much older than in ot hers. Average property age is 74 years in San Francisco, 48 years in Chica go, 47 years in Oakland and 40 years in Los Angeles. In contrast, properties in Las Vegas are only 14 years old on average, while properties in Phoenix and Dalla s are only 16 years old on average. The data indicate that retail properties are the oldest in San Franci sco, but also the smallest, having on average only 8,079 sq. ft. of improvements and 9,960 s q. ft. of land. Naturally, the older San Francisco properties are associated with lim ited parking only 5.89 spaces, compared to an average for the entire retail sample of 57.19. The vint age effect discussed with apartments and offices is strongly expre ssed in the San Francisco market, where on average the price per square footage of improvements is $213. In contrast, across all markets the average price per s quare foot is $126, while the lo west mean price/sq. ft. of $88 is observed in Detroit. Houston contains the largest retail propert ies with an average of 36,343 sq. ft. of improvements and 148,025 sq. ft. of land. Houston, however, is the second cheapest market in the sample, after Detroit, with an average price of $93 per sq.ft. The market with the newest retail pr operties sold once again Las Vegas, which is

PAGE 124

113 also the second most expensive market after San Francisco, with an average price per sq. an ex change; and three percent represent a sale of relinquished prope out-of-state include Fort Lauderdale with 18 percent out-of-state buyers; Denver with 17 ft. of $172. Approximately 13 percent of the retail tr ansactions involve the purchase of a replacement property to complete an excha nge; six percent involve the sale of the relinquished property in rty that is also a replacement property in a separate exchange. This distribution of exchanges is similar to the one observed for office properties and suggests that there are multiple properties used by taxpayers in the upleg portion of the exchange. On average, the data suggests for one relinquished propert y there are two repla cement properties. In addition, in several markets the number of replacement exchanges far outnumbers the relinquished property sales. For example, in Tucson the ratio of replacement to relinquished sales is 6.3; in the Dallas / Fort h Worth area this ratio is 6. In Houston and Las Vegas the ratios are each 5.3. In Sa n Francisco, however, relinquished sales outnumber replacement property purchases by a factor of 1.5. These ratios also can be used as indicators of which markets have b een preferred by investor s for investment once they dispose of their relinquished properties. Approximately 15 percent of all buyers reside out-of-state ; out of 12,015 retail sales 1,850 involved an out-of-state buyer. This percentage is very similar to the 16 percent share of out-of-state buyers observed in the office sample. The markets with the highest share of buyers that are out of st ate are: Las Vegas, 52 percent; Phoenix, 47 percent; Tucson, 30 percent; Dallas/Forth Wo rth, 29 percent; and Houston, 27 percent Other markets of potential interest when quan tifying the effect of having a buyer that is

PAGE 125

114 percent; Detroit, 10 percent; Seattle, 10 per cent; Chicago, 6 percent; and Los Angeles, 3 percent. Approximately three percent of all retail sales represent sale-leaseback transac This yields 340 sale-leaseback transactions in the retail sample, which is similar to number of such transactions observed with office sales (342). Because of the small number of sale-leaseback transactions a nd their relatively equal distribution across markets there are only two markets where I can expect to observe results that are statistically meaningful. Thes e markets are Chicago, where a total of 56 sale-leasebacks are observed and Phoenix, which has 50 such transactions. Finally, portfolio sales represen t on average 3 percent of all retail sales. This yiel a sample of 347 portfolio transa ctions. In contrast there were only 222 portfolio the office sample. The market with the larges t percentage of portfolio sales is Tucson, where these transactions represent 10 percent of all observations. Another market relatively high percentage of bulk purchases is Dallas / Forth Worth, where portfolio sales represent 8 percent of all sales. Finally, there are four additional markets with larger numbers of portfolio sales: Denver (33 obs erva tions. the ds sales in with a tions), Detroit (30 observations), Los Ange rties t les (40 observations) and Seattle (37 observations). Table 19 presents t-tests for differences be tween the mean price of retail prope in a comparison group composed of all proper ties not associated w ith any conditions or atypical motivation, and the groups of propertie s that are the subject of interest. The EXREPL group represents properties that are part of replaceme nt exchange only and tha are not associated with othe r conditions of sale. The EXRELQ group represents relinquished exchanges only; RELQ_REPL is a group of properties that were part of two

PAGE 126

115 separate exchanges; BUYEROUT contains sales to out-of-state buyers only; the SALELEA SEBACK group includes sale-leaseba ck transactions only and PORTSALE repres Table 19. Differences in Mean Pr ices of Control Sample and Identified Groups of Interest Mean Value Price ue ents portfolio sales. for Retail Properties Retail Observations of Sales Standard Error T-test Val Control Group 7,932 1,655,211 38,984 EXREPL 1,232 2,919,869 131,304 -11.33 EXRELQ 691 2,045,297 12 6,705 -2.84 RELQ_REPL 47 3,639,787 630,065 -3.78 BUYEROUT 1,166 5,773,483 263,312 -28.75 SALELEASEBACK 146 1,649,350 173,243 0.02 PORTSALE 177 3,323,456 582,685 -6.07 The table illustrates that the null hypothe sis that on average replacement exchanges have the same price as propert ies that are not associated with any sale conditions is rejected at less than the one percent level. The average price of replacement properties is almost twice the average price in the contro l group. Relinquished prope rties and sales that are part of two separate exchanges have likewise significantly hi gher average prices, although the t-test values associated with the corresponding mean di fferences tests are considerably lower. The BUYEROUT group and PORTSALE group also have significantly higher prices than the properties of the contro l group. With sa le-leasebacks the null hypothesis is not rejected. Next, regressions are performed based on estima ting equation (18) for each of the 15 markets. A backward stepwise method of estimation is applied, in which the only

PAGE 127

116 subjects to the selection pr ocedure are the submarket dummy variables. The proc starts with a full model which includes all submarket dummies for the specified mark and then selectively remove edure et, s dum my variables from the model that are not significant at specified level (0.1) or do not influence th e fitness of the model. All other dependent variables are not subject to the procedure. Standard errors are adjusted to account for potentiaskedast ports coefficient estimates and p-values foriable by markets. Rsquares are on average 10 plower t uarese model with office commercial real estate. R-sq varies fcent focago ma 84 percent for the Denver and Phoenix markets. c o diffelthough tially estimat c the office regns. The most notable difference is with resp ect to the coefficients on year dummies, which illustrate a different trend of property appreciation in the retail real estate market, when compared to the office real estate market. The reported results in Table 20 show th at the estimated coefficients on the structural attributes are of the predicted si gn and statistically signi ficant in all of the market regressions (with the ex ception of the coefficient on sq uare footage of land in San Francisco). The coefficient on AGE is negative and significant in all of the models. The coefficient on AGE2 is positive, as predicted, and also significant in all of the markets. The coefficients on SQFT, SQFT2, LANDSQFT, and LANDSQFT2 are all positive and highly significant, in all mode ls but one (San Francisco). a l hetero icity. Table 20 re r each va ercent han th e R-sq of the sam ua red rom 71 per r the Chi rket to The reported coefficients for stru ctural haracteristics are als rent (a not substan ) from the ed oefficients in ressio

PAGE 128

117 These results represent the strongest results with respect to age, square footage an square footage of land in the regressions ba sed on the three differe nt property types: apartment, office and retail. San Francisco, as illustrated by the summary statistics is a very different market from the other 14 retail markets studied. It has older properties tha also tend to be very small. Properties built on larger lot sizes tend to be newer and do n carry the premium of vintag d t ot e properties. Al so older properties ar e most likely located in or e four of the markets. The coefficient on the va e lowest price per square footage. The coeffi is to be similar in magnitude to the coefficients estimated in the office regr essions and much larger than the estimated close to the business center, while ne wer properties tend to be built further away from the business center. The coefficient estimates in San Francisco indicate a positive concave relationship with square footage, a nd a negative convex relationship with squar footage of land. The coefficient on the variable representing the number of parking spaces, PARKING is positive, but it is significant in only riable representing the number of floors, FLOORS is positive and significant in four of the markets, and negative and significan t in Houston. Houston is the market with the largest average retail prope rty size and also th cient on the variable represen ting below average condition, CONDITION_BA generally negative and significant in 9 out of the 15 markets. The coefficient on the variable representing above average condition, CONDITION_AA, is positive and significant in all but one of the models. The estimated coefficient on the variable of interest, indicating whether the sale was part of a replacement property exchange ( EXREPL ) is positive and significant in 13 out of the 15 regressions. The estimated coeffi cients tend

PAGE 129

118 Table 20. Re Dummies by Retail Markets M Dallas/ es gression Statistics for OLS Model with Structural Characteristics and Submarket arket Chicago Fort Worth Denver Detroit Ft. Lauderdale Houston Las Vegas Los Angel Observations 2069 566 699 639 516 479 426 2150 EXREP L 0.373 0.255 0.259 0.209 0.242 0.245 0.140 0.202 0.00 0.00 0.00 0.13 0.04 0.00 0.01 0.00 EXREL RELQ_ AGE -0.021 -0.029 -0.024 -0.015 -0.034 -0.032 -0.027 -0.025 AGE2 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.00 0.00 0.00 0.01 0.00 0.00 0.00 SQFT 0.031 0.00 SQFT2 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.00 LANDS 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 LANDS PARKI FLOORS 0.044 0.005 0.111 0.054 0.094 -0.158 0.080 0.005 CONDITION_BA -0.200 -0.192 -0.130 -0.034 -0.142 -0.082 0.113 -0.216 0.00 COND 337 0.313 0.275 0.253 0.139 00 0.00 0.00 0.00 0.00 BUYEROUT 0.00 0.00 0.00 0.01 0.37 0.00 0.04 0.27 SALEL 0.01 0.53 0.69 0.04 0.17 0.89 0.03 0.28 PORTS YR2000 0.60 0.19 0.21 1.00 0.96 0.32 0.29 0.31 Q 0.200 0.034 0.085 0.141 -0.133 0.244 0.163 0.158 0.00 0.81 0.21 0.32 0.50 0.06 0.17 0.00 REPL 0.395 0.423 0.363 0.217 0.060 0.310 0.185 0.2040.00 0.01 0.00 0.04 0.84 0.10 0.07 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.027 0.015 0.020 0.022 0.017 0.019 0.019 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 QFT 0.003 0.004 0.003 0.003 0.004 0.002 0.005 0.003 QFT2 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 NG 0.000 0.000 0.000 0.001 0.000 0.000 0.001 0.000 0.80 0.28 0.26 0.07 0.51 0.79 0.08 0.74 0.04 0.96 0.03 0.39 0.29 0.03 0.65 0.85 0.00 0.02 0.04 0.79 0.14 0.31 0.21 ITION_AA 0.300 0.164 0.275 0.0.00 0.00 0.00 0. 0.238 0.403 0.220 0.204 0.069 0.358 0.098 -0.096 EASEBACK 0.165 -0.065 -0.030 0.313 -0.160 0.024 0.216 0.101 ALE -0.037 0.247 -0.090 0.159 -0.139 0.408 -0.034 -0.0660.75 0.00 0.40 0.02 0.46 0.06 0.81 0.45 0.026 -0.132 0.150 0.000 -0.005 0.127 -0.130 -0.058

PAGE 130

119 Table 20. Continued Market Chicago Dallas/ W Denver Detroit Ft. Lauale Houston Las V Los As Fort orth derd egas ngele YR2001 0.090 -0.146 0.175 0.176 0.077 0.285 -0.057 0.128 0.0. 0.0.41 0.67 0. red 07 16 0. 14 03 04 0. 04 YR2002 0.182 -0.013 0.202 0.091 0.177 0.345 -0.081 0.233 0.00 0.89 0.09 0.28 0.10 0.01 0.54 0.00 YR2003 0.326 -0.001 0.371 0.288 0.478 0.419 0.106 0.454 0.00 0.99 0.00 0.00 0.00 0.00 0.43 0.00 YR2004 0.551 0.057 0.466 0.390 0.605 0.584 0.185 0.646 0.00 0.59 0.00 0.00 0.00 0.00 0.15 0.00 YR2005 0.549 0.029 0.547 0.467 0.811 0.943 0.145 0.829 0.00 0.83 0.00 0.00 0.00 0.00 0.36 0.00 CONST 14.289 13.467 13.890 12.646 13.519 13.095 13.186 15.023 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 R-squa 0.71 0.83 0.84 0.73 0.81 0.81 0.78 0.78 All standard errors are adfor poheteroskedasticit y.es arted under ieates an n balics. Depende ble is LNPRICE sell; E Binary set one if t represents salee opertLQy v ual transeprese of a relinquished property; RELQ_REPLry varet equne ifc tion repreoth salinquis p chase of ement y. Age oildiyear s; AGge sqQFT Square f mprovemFT2 re footage of total improvements square d; lanquare of la landsqft2 Square footagnd squARKING Park ing, definedber of paraces; Numb f Ni Pcondpase pece categorde be a The oegerageut vat equal touyert of st seaseback Binary vaset eqne if tion wt of sseback; po Binare set to o on was partfolio Yearly time periods from 1999 through 2005. E eded b except 19h is ed; S Biiabyin g the st in which the pro lated, as defined by Cookers. justed tential P-valu re repo the coeffic nt estim d ar e i old and it ransaction nt Vari a of a replac Log of y. EXRE ing price Binar XREPL ariable set eq variable to one if equal to action r nts sale men t pr Bina iable s al to o transa sents b le of re hed roperty and pur ootage of total i a replac ents; SQ t proper S qua GE A f the bu ng(s) in E2 A dsqft S uared; S e footag nd; e of la ared; P as num king sp floors er of loors; CONDITIO nd above average. hysical mitted cat ition of the ory is av roperty b ; buyero d on ins Binary tion. Th riable se ies inclu one if b low average, lives ou average ate; alel riable ual to o ransact as par ale-lea rtsale y variabl equa l ne if transacti inary variable t of por 99, whic sale; YRn suppr ess DUMi ach y ubmarke ar is inclu as a perty is nary var le signif oc Star br

PAGE 131

120 TMark x Riverside a San San F Seattle Tucson able 20. Continued et Oakl and Phoeni /Sa Bern n rdino Diego rancisco Observations 508 1208 429 638 339 943 406 5 0 0. EXREPL 0.320 0.086 0.24 .254 192 0.122 0.216 0.00 .00 0 0. 2 0 0. .46 0 0.EPL 2 0 0. .00 0 0.4 0 030-0 .00 0 0. 0 0 0. .00 0 0.2 0 0. .00 0 0.0 0 .000 0 -0 0 0 0.14 0.17 0.24 0.17 0.75 0.04 CONDITION_BA -0.176 -0.310 -0.274 -0.094 -0.280 -0.197 -0.020 0.01 0.00 0.10 0.18 0.00 0.00 0.77 CONDITION_AA 0.161 0.258 0.148 0.205 0.141 0.326 0.294 0.11 0.00 0.04 0.00 0.10 0.00 0.00 BUYEROUT -0.006 0.217 0.136 0.013 -0.121 0.313 0.155 0.96 0.00 0.29 0.91 0.37 0.00 0.01 SALELEASEBACK 0.177 0.161 -0.001 0.002 0.314 0.111 0.031 0.23 0.00 0.99 0.99 0.18 0.23 0.79 PORTSALE 0.003 -0.056 0.088 -0.062 -0.134 -0.212 0.195 0.99 0.57 0.59 0.67 0.09 0.01 0.03 YR2000 0.181 0.008 0.043 -0.056 0.155 0.182 -0.069 0.10 0.93 0.70 0.17 0.04 0.02 0.63 0.17 0 .00 02 0.00 0.00 EXRELQ 0.223 -0.027 0.09 .177 118 0.173 0.070 0.00 0.78 0 .01 03 0.01 0.61 RELQ_R 0.328 0.134 0.30 .362 141 0.290 0.206 0.01 0.47 0 .00 12 0.00 0.13 AGE -0.02 -0.03 -0. 0.028 .011 -0.023 -0.038 0.00 0.00 0 .00 02 0.00 0.00 AGE2 0.000 0.000 0.00 .000 000 0.000 0.000 0.00 0.00 0 .00 03 0.00 0.00 SQFT 0.042 0.016 0.01 .027 142 0.026 0.013 0.00 0.00 0 .00 00 0.00 0.00 SQFT2 0.00 0.00 0 .000 .002 0.000 0.000 0.01 0.00 0.01 0.00 0.00 0.00 0.09 LANDSQFT 0.004 0.005 0.005 0.004 -0.002 0.003 0.006 0.02 0.00 0.00 0.00 0.84 0.00 0.00 LANDSQFT2 0.000 0.000 0.000 0.000 0.000 0.000 0.00 0.10 0.00 0.00 0.00 0.00 0.00 0.00 PARKING -0.002 0.000 0.001 0.000 0.001 0.000 0.00 0.12 0.90 0.08 0.30 0.82 0.00 0.83 FLOORS 0.094 0.081 0.135 0.045 -0.035 0.012 0.055 0.05

PAGE 132

121 Table 20. Continued Market Oakland Phoenix Riverside /San Bernardino San Diego San Francisco Seattle Tucson YR2001 0.30.00.260.030.0.0.1 72 60 1 8 367 179 26 0.00 0.47 0.04 0.73 0.00 0.02 0.43 Y R2002 R2003 R2004 R2005 ONST 4 3 9 2 2 0 -squared 0.428 0.029 0.318 0.233 0.454 0.305 0.129 0.00 0.73 0.00 0.04 0.00 0.00 0.39 Y 0.640 0.214 0.439 0.303 0.453 0.299 0.194 0.00 0.01 0.00 0.01 0.00 0.00 0.18 Y 0.750 0.399 0.646 0.600 0.524 0.375 0.447 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Y 0.606 0.589 0.839 0.802 0.674 0.694 0.521 0.00 0.00 0.00 0.00 0.00 0.00 0.00 C 13.87 13.27 12.95 12.58 66.31 13.33 13.257 0.00 0.00 0.00 0.00 0.00 0.00 0.00 R 0.81 0.84 0.83 0.83 0.82 0.80 0.81 Aerrors are adjupotenrosked -valuesrted und ieates an ll standard sted for tial hete asticit y. P are repo er the coeffic nt estim d ar e in bold andDepe ndent Variable is L Log ofprice; EXL Binar e if transaction represents sale emperty. EX Binary set equaif trans represe of a operty; REPL ablel to one tion repth sanquish p chase of a ent AGE the building(s) in years; A ge sq QFT footage of total improveme FT2 Square footage l improvements squared; LANDSQFT Sfootage ; L uare fooand s PARKI rking, defumber g spac ORS Number o TIONi Pcondhe proped on ins The catncludeverage and above average. The omategoryrage; BUUT Binaable set eqone if bues out o ; S CK Binar se one if s ple-leaseb RTSA y variable set e transaction wf poale; YR ly time pom 1999 through 2005.ar is included as a binary variable except 1999, which is suppressed; SDU inary varignifying the submarket in which the pris l fined by CoSrs. italics. NPRICE ent pro selling RELQ RE variable P y vari l to one able set equ al to on action of a replac nts sale relinquished pr roperty and pur LQ_RE replacem Binary vari property set equa Age of if transa c resents bo GE2 A le of reli uared; S ed Square nts; SQ of tota quare of land ANDSQFT2 Sq f floors ; CONDI tage of l hysical quared; ition of t NG Pa erty bas ined as n pectio n. of parkin egories i es; FLO below a e, averag itted c is ave YERO ry vari ual to yer liv f state ALELEASEBA qual to one if y variable as part o t equal to rtfolio s tr ansaction wa art of sa eriods fr ack; PO LE Binar Each ye nM Year i B able si operty ocated, as de tar broke

PAGE 133

122 coefficients with the apartment models. The lo west e stimated significant coefficient is in as Vegas leest coefficient for replacemenn observed, just as with the o ffice regressions, in the Chicago model 0.37. This represents additional evidence that buyers of replacem ent properties are paying statistically significant prremin the m in the reions. The coefficient es enting a s linquished nge, LQ itive iodels a nific7 m. owever, theffictimates generally nificantly smalleroef stimates on the variable, indicating a sale that f a re EPL. These results are similar to the finding sions. The coefficient on the variable represen ting a purchase by ou t-of-state buyer, BUYEROUT is significant in all five of the markets with highest percentage of out-ofstate buyers: Las Vegas, 52 percent; Phoenix, 47 percent; Tucson, 30 percent; Dallas/Forth Worth, 29 percent; and Houston, 27 percent. This coefficient is also significant in Denver, Detroit, Seattle and Chicago. Thus, the coefficient on the variable representing a purchase by an out -of-state buyer is si gnificant in 9 of the 11 markets with large number of sales to out-of-state buyers. These findings present strong evidence that out-of-state buyers are associat ed with a price premiums. L 0.14, whi the high estimated t excha ges is ice p ums i ark ets exam ined gress timates of the variable repres ale as part of a re excha EXRE are pos n 13 m nd sig ant in arkets H se coe ient es are sig than c ficient e is part o placement exchange, EXR s with office regres The coefficients on the variable representing a sale part of the sellers relinquished exchange and the buyers replacement exchange, RELQ_REPL are positive and generally of higher magnitude than the co efficients on the variable representing a replacement exchange only ( EXREPL ). The coefficients are positive and significant in 11 out of the 15 markets.

PAGE 134

123 The coefficient of the variable representing that the sale is part of a sale-leaseback transaction, SALELEASEBACK is generally positive and significant in four markets: Chicago, Detroit, Las Vegas, and Phoenix. The preliminary analysis indicated that there were only two markets with larger numb er of sale-leaseback transactions. These m are Chicago with 56 sale-leasebacks and Phoenix with 50 such transactio arkets ns. In both of the m tion is ositive d negatively signif umber ly ble in Seattle. In conclusion, the regression results present mixed evide e iation arkets the coefficient of the variable representing a sale-leaseback transac positive and significant. These results pr esent additional evidence supporting our hypothesis that sale-leasebacks are associ ated with higher transaction prices. The coefficient of the variable representing a portfolio sale, PORTSALE, is p and statistically significant in three markets: Dallas, Detroit, and Tucson; an icant in Seattle. Tucson and Dallas re present the two markets with the largest percentage of portfolio sales observed. In th e four additional mark ets with large n of portfolio sales (Denver, Detroit, Los Angeles and Seattle), I obs erve a statistical significant and positive coefficient on PORTSALE in Detroit, and a negatively significant coefficient on this varia nce that portfolio sales are associated with price premiums. The estimated year dummies are generally positive and significant. The magnitud and significance of the year dummy coefficien ts reveal substantia l price apprec after 2001. Oakland, San Francisco and San Diego are the markets with the most dramatic price appreciation, as illustrated by the dummies. The coefficient of 0.67 for the dummy variable signifying sales in year 2005 in San Francisco, translates to 96 percent higher prices, all else equal, than in 1999. The same coefficient for the markets of Seattle

PAGE 135

124 San D mple. r expan ute presenting a replac s miums The coefficient on the variable represen ting a purchase by an out-of-state buyer, iego and Riverside, indicates a price appreciation sinc e 1999, all else equal, of 100 percent, 123 percent, and 131 percent, respectively. Submarket dummy variables are not reporte d in the regression outputs, but they serve for achieving best possi ble control of relative location in the metropolitan market and their inclusion improves significa ntly the fitness of all models. Next, I present a second m odel for the retail properties, OLS Model II, which is identical in its specification to the second model used with the office properties sa The model adds to the model specifica tion, given by equation (18) a third orde sion of the latitude and longitude coordi nates, which has the objective of effectively controlling for the absolute lo cation of the property. The th ird order expansion, rather than a simple linear form, is entered in the equation to draw a price surface based on location. The model specification is given by equation (23). Table 21 presents regression statistics from OLS Model II. Contro lling for absol location improves the R-squares of market regr essions by 0.5 to 1 per cent. No significant changes in the estimated coefficients are obs erved. With respect to the key variables of interest results are re-confirmed. The co efficient on the variable re ement exchange, EXREPL is positive and significant in 13 of the 15 regressions. The magnitude of the estimated coefficients remains largely unchanged. This reconfirm the evidence that buyers of replacement prope rties are paying significant price pre in the majority of the markets. Results with respect to relinquished exchanges, EXRELQ and a combination of relinquished and replacement exchanges, RELQ_REPL remain the same.

PAGE 136

125 Table 2 Dummies and Longitude, Latitude Coordinates by Retail Markets Ft. Las 1. Regression Statistics for OLS Model w ith Structural Characteristics, Submarket Dallas/ Chicago Fort Worth Denver Detroit Lauderdale Houston Vegas ObsEXREPL 0.369 0.244 0.259 0.211 0.212 0.248 0.131 ervations 2069 566 699 639 516 479 426 0.00 0.01 EXRELQ 0.204 0.031 0.085 0.142 -0.080 0.228 0.166 RELQ_RE AGE .023 0.00 0.00 0.00 0.00 0.00 0.00 0.00 AGE 0.00 0.00 0.00 0.00 0.00 0.00 0.00 SQF SQF 0.00 0.02 LAND 4 0.002 0.005 LANDSQFT2 0.000 0.000 0.000 0.000 0.000 0.000 0.000 PARKING FLO 140 0.05 0.92 0.03 0.38 0.38 0.03 0.40 CON 0.00 0.03 0.05 0.78 0.32 0.26 CON BUY SALE PORTSALE -0.031 0.255 -0.089 0.157 -0.143 0.406 -0.119 0.79 0.00 0.41 0.02 0.43 0.06 0.40 YR2000 0.00 0.00 0.00 0.13 0.07 0.00 0.82 0.20 0.33 0.68 0.07 0.15 PL 0.391 0.401 0.360 0.225 0.039 0.299 0.162 0.00 0.01 0.00 0.04 0.89 0.13 0.11 -0.021 -0.029 -0.024 -0.015 -0.036 -0.031 -0 2 0.000 0.000 0.000 0.000 0.000 0.000 0.000 T 0.027 0.015 0.020 0.022 0.017 0.019 0.019 0.00 0.00 0.00 0.00 0.00 0.00 0.00 T2 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.00 0.00 0.00 0.00 SQFT 0.003 0.004 0.003 0.003 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.000 0.000 0.000 0.001 0.000 0.000 0.001 0.79 0.23 0.27 0.08 0.57 0.80 0.09 ORS 0.042 -0.011 0.111 0.055 0.075 -0.163 0. DITION_BA -0.200 -0.188 -0.127 -0.036 -0.094 -0.079 0.102 DITION_AA 0.301 0.152 0.275 0.337 0.307 0.281 0.213 0.00 0.01 0.00 0.00 0.00 0.00 0.00 EROUT 0.239 0.414 0.222 0.198 0.079 0.363 0.094 0.00 0.00 0.00 0.01 0.31 0.00 0.04 LEASEBACK 0.160 -0.054 -0.033 0.320 -0.193 0.022 0.217 0.02 0.62 0.66 0.03 0.12 0.90 0.02 0.027 -0.132 0.155 0.001 -0.040 0.134 -0.126 0.58 0.21 0.20 0.99 0.63 0.29 0.30

PAGE 137

126 Table 21. Continued Dallas/ Ft. Chicago W Denver Detroit LauderHouston Las V Fort orth dale egas YR2001 0.090 -0.141 0.178 0.172 0.036 0.285 -0.037 0.07 0.18 0.13 0.04 0.70 0.04 0.78 YR2002 0.181 -0.016 0.206 0.088 0.124 0.354 -0.085 0.00 0.87 0.08 0.30 0.23 0.01 0.51 YR2003 0.329 -0.001 0.373 0.286 0.441 0.427 0.119 0.00 1.00 0.00 0.00 0.00 0.00 0.37 YR2004 0.557 0.045 0.470 0.386 0.576 0.592 0.204 0.00 0.68 0.00 0.00 0.00 0.00 0.11 YR2005 0.547 0.024 0.551 0.474 0.822 0.928 0.173 0.00 0.86 0.00 0.00 0.00 0.00 0.26 X3 0.000 0.025 -0.003 -0.006 0.075 0.05 0.13 0.66 0.52 0.14 Y3 0.000 -0.002 0.000 0.000 -0.005 0.001 -0.002 0.02 0.14 0.66 0.52 0.16 0.81 0.14 XY2 0.000 -0.008 0.001 -0.024 0.004 -0.014 0.06 0.13 0.66 0 .15 0.80 0.15 YX2 -0.005 0.007 -0.023 0.52 0.80 0.15 CONST 25.823 28.413 4.718 12.772 69.724 -1.313 -93.981 0.00 0.17 0.74 0.08 0.00 0.95 0.00 R-squared 0.72 0.83 0.84 0.73 0.82 0.81 0.79 All standard errors are adjustotentiaskedas -valueported coeftimatar e in ed for p l hetero ticit y. P s are re under the ficient es es and bold ndent V LNP og o price L Briableto on transaction represents sale of cementty. EXR Binaryle set e one if rep sale of a relinquished property; RELL Binriable sl to onesa ction th sale of relin prop ret pro E Ae building(s) in yea Age squared; SQ uare footage of total improvement T2 Sqotage o improv square DSQFT land ; LANDSQFT2 Square footand squa ARKIN ing, d numarkingaces; FLOO umber of fl Phnditioprope on in The es incow average and abov ity is BUY Binable sto oner lives tate ; SALEEASEBACK Binary set eqne if tr s sale-lek; POR Binarable set equa ction wa portfolio sale; YRn time m 1999 through 2005. Each yeaded as a bin 1999, which is suppressed; SDUM ary vargnifyibmarkhich thrty is loca, as defined by CoStars; X Le of pro Longitude of prop and italics. Depe ariable is RICE L f selling ; EXREP inary va set equal e if a repla Q_REP proper ary va ELQ et equa variab if tran qual to represents bo transaction resents quished erty and purchase of a placemen perty. AG ge of th rs; AGE2 FT Sq s; SQF ge of la uare fo red; P f tota l G Park ements efined as d; LAN ber of p Square footage of sp RS N oors ; CONDITION i e average. The om ysical co n of the e; rty based T spectio n categori lude bel erage, av s ted categor variable averag ual to o EROU ansaction wa ary vari part of et equal asebac if buye TSALE out of y vari L l to one if transa y variable except s part of Yearly n periods fro i r is inclu e ar ted i Bi perty; Y iable s ng the s erty. u et in w e prop broker atitud

PAGE 138

127 Table 21. Continued Los Angeles Oaklnix ive /San ardi San Diego San Francisco Seattle Tucson and Phoe R rside Bern no Obsers 21 1208 429 8 9 40EXREPL 00.3085247 254 7 0. vation 50 50 8 0. 63 338 943 6 .204 10 0. 0. 0.1 0.117 201 0.00 0.17 0.00 00 0.EXRELQ 0.21.031090 177 9 0.06 0.00 0.75 0.47 01 0.RELQ_RE 00.20.136306 360 7 0. 0.02 0.46 0.00 00 0.AGE -00.024-0.0290.029 .027 -0. 0.00 0.00 00 00 0.AGE 00.00.000000 0 0 000 0.0 0.00 0.00 0.00 00 0.SQ 0.04016 013 027 3 0.01 0.00 0.00 0.00 0.00 0.SQ 00.0000000 0.000 2 0. 0.03 0.00 0.00 00 00 0.LANDFT 0.3 0.005 0.005 0.005 4 0.005 0.00 0.01 00 0.00 00 0.LAN 00.0000000 0 0. 02 0.00 0.00 0.00 00 0. 0.32 0.81 0.00 0.36 0.91 0.53 0.00 0.01 SALELEASEBACK 0.101 0.216 0.147 -0.012 -0.010 0.341 0.098 0.010 0.29 0.13 0.01 0.94 0.95 0.10 0.30 0.93 PORTSALE -0.074 0.008 -0.049 0.046 -0.053 -0.200 -0.224 0.167 0.40 0.97 0.61 0.77 0.72 0.03 0.01 0.06 YR2000 -0.054 0.175 0.010 0.017 -0.056 0.132 0.189 -0.064 0.33 0.11 0.91 0.88 0.63 0.06 0.02 0.67 00 0. 0.0. 0.02 0.01 2 01 164 0. 4 -0 0. 0.13 0.17 2 00 0. 0. 0.01 3 0.01 65 PL .204 87 0. 0. 0.1 0.282 171 00 0. 0.0 0.14 008 0.00 21 .025 -0. -0.0230.00 039 00 0. 0.0. 0.00 0.07 0 00 2 .000 00 0. 0.0 0. 00 00 0. 0.0. 0.11 0.00 6 00 FT 031 0. 3 0. 0. 0.13 0.02 3 00 0. 0.00 0 0.00 00 FT2 .000 00 0. 00 0. 0. -0.00. 0.000 000 0.0.00 0.00 3 08 SQ 00 0.00 6 00 0. 0.0. 0.00 0.65 00 0.00 00 DSQFT2 .000 00 0. 0. 0.0 0.000 000 0. 0. 0.00 00 PARKING 0.000 -0.002 0.000 0.000 0.000 0.000 0.000 0.000 0.68 0.04 0.88 0.08 0.29 1.00 0.93 0.82 FLOORS 0.002 0.092 0.079 0.132 0.048 -0.020 0.013 0.058 0.96 0.06 0.14 0.16 0.21 0.41 0.73 0.04 CONDITION_BA -0.224 -0.171 -0.304 -0.295 -0.105 -0.243 -0.195 -0.014 0.00 0.01 0.00 0.14 0.14 0.00 0.00 0.83 CONDITION_AA 0.139 0.144 0.258 0.139 0.208 0.144 0.322 0.290 0.00 0.15 0.00 0.06 0.00 0.10 0.00 0.00 BUYEROUT -0.089 0.023 0.211 0.119 0.012 -0.081 0.310 0.150

PAGE 139

128 Table 21. Continued Los Angeles Oakland Phoenix Riverside /San Bernardino San Diego San Francisco Seattle Tucson YR2001 0.132 0.336 0.056 00000 .240 .039 .332 .177 .124 0.03 0.00 0.51 0.05 0.73 0.00 0.02 0.46 Y R2002 R2003 R2004 R2005 3 3 Y2 X2 ONST -17.75 1 --111 33 6 -squared 0.232 0.419 0.028 0.307 0.233 0.427 0.310 0.120 0.00 0.00 0.74 0.00 0.03 0.00 0.00 0.45 Y 0.457 0.645 0.212 0.420 0.303 0.406 0.302 0.179 0.00 0.00 0.01 0.00 0.01 0.00 0.00 0.24 Y 0.644 0.719 0.399 0.624 0.599 0.494 0.379 0.432 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01 Y 0.826 0.619 0.587 0.813 0.800 0.659 0.689 0.503 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 X 0.036 0.195 0.007 0.008 0.008 0.022 0.00 0.00 0.34 0.45 0.69 0.06 0.22 Y 0.002 0.012 0.000 0.001 0.000 0.000 0.001 0.001 0.00 0.00 0.33 0.57 0.68 0.00 0.06 0.21 X 0.009 0.057 0.002 0.003 0.002 0.001 0.004 0.005 0.00 0.00 0.35 0.09 0.69 0.00 0.06 0.23 Y 0.004 0.10 C 0 39.120 10.912 77.57 4.638 93.551 0.259 9.135 0.09 0.00 0.26 0.46 0.22 0.05 0.03 0.01 R 0.78 0.82 0.84 0.83 0.83 0.84 0.80 0.81 Arrors are adjupoteteroskedasticit y. are unefftim ar e in ll standard e sted for ntial he P-values reported der the co icient es ates and bold and italics. Depe ndent Variable is CE sellin; EXRE ble sl to transac of emer Q variabaif trans rep sale of a REL vaquaif tra reotred property and purchase of a rentty. A ge of tding(sars; A Age squared; SQFT Square footage T2 Square footage of tota l improvements squared; LANDSQFT land ; LA footnd ; PARKING Pafinedberg s L Number of floors ; CONDITIONi Ph condition of the property basespecti e categories include below a, average and above. The omegorrage; ROUT ary va eque if livate ; SA Binar seto o nsactpart oase RT Biable set equal to one if transaction w of p sale Yearlyperiod1999h 2005. Each ycluded as a bin 1999, which is suppressed; Biiableng tarkichperty is lo Stars; X udeerty; itudert LNPRI Log of g price PL Binar y varia et equa one if tion represents sale relinquished property; a replac LQ_REP ent prop Binary ty. EXREL riable set e Binary l to one le set equ nsa ction l to one presents b action h sale of resents linquishe eplacem SQF proper GE A he buil ) in ye GE2 of total improvements; NDSQFT2 Square Square footage of paces; F age of la squared rking, de as num of parkin OORS ysical d on in o n. Th verage e averag LELEASEBACK itted cat y variable y is ave t equal BUYE Bin ion was riable set f sale-le al to on back; PO buyer SALE es out of st inary var ne if tr a as part ortfolio ; YR n SDUM time ary va s from signify throug he subm ear is in the pro ary variable except cated, as defined by Co i of prop nr Y Long i e of prop e t in wh r broke Latit y.

PAGE 140

129 BUYEROUT remains siificant and positive in 9 out of the iied 11 markets of interest. These findings reconfirm the str ong evidence that retail property out-of-state buyers pay p The coefficient of the variable representing a sale-leas eback transaction, SALELEASEBACK res po andfican he mts ocagetroit, Las Vegas, and Phoenix. It is also now si ant i Fr co a10 nt el. These r re tden porr hesi saebacks e associated with higher tran pri Finallyts r uned sp ee ieheble representing a portfolio sale, PORTSALE. Based on OLS Model II, the percentage sa les price changes corresponding to the estimated and statistically significant co efficients folacet exes EPL relinquished enge REL linquished and replacement exchanges RELQ_REPL, purchasesy out-of-state buyers, BUYEROUT ale-leasbacks, LELEASE a es L p i et. d to complete a replacement exchange is 26.06 percent. This premium is similar to the premiu m paid with office gn dentif rice pre miums. main sitive signi t in t arke f Chi o, D gnific n San ancis t the perce lev esults confirm he evi ce s up ting ou ypoth s that le-leas ar saction ces. resul emain chang with re ct to th coeffic nt on t varia r rep men chang EXR xcha s, EX Q, re b s e SA BACK nd portf olio sal PORTSA E, are resented n Table 22 by mark Table 22 indicates that the percentage price effect of replacement property exchange is economically si gnificant and varies from 12.46 percent in Seattle to 44.61 percent in Chicago. Note that the market s with the lowest a nd highest premiums associated with replacement exchanges are the same for retail and office properties. The average price premium paid when a property is use

PAGE 141

130 replacement property exchanges and is larger than predicted net benefits from a taxdeferred exchange, based on the simulation analysis in Chapter 5. In contrast, based on th e simulation analys is with a holding period of less than 10 years m g and holds the repl acement property for anoth anges that is larger than any predicted benefit, espec ant are ium nges 5 percent in Los Angele s to 49.39 percent in Dallas. ranging h The coeffi purchases by out-of-state buyers. both for the relinquished and the replacement property, the net benefit fro exchange is 8.46 percent of pric e, at most. This value is achieved when assuming annual appreciation of 20 percent, and 20 percent cap ital gain tax rate. In creasing both holdin periods to 20 years yields a maximum price be nefit of less than 12 percent. Finally, if an investor exchanges a fully depreciated prope rty er 39 years, the price bene fit of this strategy is only about 14 percent. This clearly shows that investors pay a premium in exch ially when they have a short-term investment horizon. The implied percentage effects of reli nquished exchanges on selling price are on average 19.57 percent. The implie d price effects, when this variable is si gnific smaller than the effects associated with replacement exchanges. The price prem associated with a sale being part of both replacement and relinquished exchanges ra from 22.6 Retail purchases by out-of-state buyers are associated with a price premium from 9.82 percent in Las Vegas to 51.21 percent in Dallas. Dalla s is also the market wit the highest price premium associated with out -of-state buyers of office real estate average price premium associated with this type of transactio n is 28.29 percent. The cient is significant in 9 out of the identified 11 mark ets of interest. These results present strong evidence that there is a si gnificant price premium associated with

PAGE 142

131 Table 22. Marginal Effects for Significant Coefficients for Variables of Interest in ReRELQ_ BUYER tail Regressions MarketObs EXREPLEXRELQ SALE LEASE PORT Dallas/ Fort Ft. Lauderdale51623.60% Houston47928.19% 43.76% Las Vegas42614.01% 9.82% Los Angeles215022.69%17.86%22.65% Oakland50836.36%23.91% Phoenix1208 23.55%15.81% Bernardino 42927.99% 35.79% % -20.04% Tucso REPL OUT BACK SALE Chicago206944.61%22.60%47.79%27.00%17.31% Worth 56627.65% 49.39%51.21% 29.04% Denver 69929.54% 43.32%24.84% Detroit 639 25.19%21.89%37.76%17.01% n 40622.27% 16.20% 18.22% Riverside /San San Diego 63828.90%19.37%43.39% San Francisco33920.55%14.88% Seattle 94312.46%18.78%32.64%36.34

PAGE 143

132 Sale-leasebacks are associated with an average price premium of 23.63 percent. Price effects from sale-leaseback transacti ons range from 15.82 percent in Phoenix to 37.76 percent in Detroit. These results are of statistical and ec onomic significance and support the evidence that there is a price pr emium associated with sale-and-leaseback transactions. Portfolio sales are associated with a pr ice premium of 17.01 percent in Detroit and 29.04 percent in Dallas, and a price discount of 20.04 percent in Sea ttle. These results present mixed evidence of a price premiu m (discount) associated with portfolio transactions. The next chapter summarizes the results and offers concluding remarks.

PAGE 144

CHAPTER 10 CONCLUSION I study t he role that buyer a nd seller motivations play in determining sales prices in comm I n in comparables sales data and also could influence sales price. In partic their effect on observed transaction prices. I also examine the pricing of properties that are purchased by out-of-state buyers, as well as sales that are part of condominium conversions, portfolio transactions or sale-leaseback transactions. This dissertation finds severa l interesting results that ha ve not been discussed in previous published work. First, this dissertation represents the fi rst work that quantifies the size of the exchange market nationwide, as well as defines conceptually and empirically the magnitude of possible effects of exchanges on transaction prices in different markets across the country. Second, I find a significant positive marginal effect related to replacement property exchanges, which is consistent with the Ta x Capitalization Hypothesis and the Imperfect Substitute Hypothesis interpretations. This e ffect is robust across the 15 markets studied in the apartment, office and retail properties samples. More importantly, the price effect differs substantially across residential and non-residential properties, as well as across ercial real estate markets. Various c onditions of sale, which can be viewed as distinct motivations, appear to be quite comm on in commercial real es tate transactions. examine several conditions of sale which represent distinct motivations that are frequently see ular, I focus attention on the use of tax-deferred exchanges nationwide and 133

PAGE 145

134 markets. I document an average price premiu m of 10.5 percent for apartment replacement exchanges, 28.3 percent for office replacemen t exchanges, and 26.1 percent for retail replacement exchanges. The results dem any investor s pay a price premium in exchanges that is significantly larg ted tax benefit of the exchange, espec with es y pay in or in whole, the gain from f the price effect associated with purchases by out-of-state buyer s. es, as t in apartment markets, 23.5 percent in office markets and 28.3 perce onstr ate that m er th an the expec ially if they have a short-term i nvestment horizon. Based on the comparison with the simulation model results and observed si gnificance in the majority of the studied markets, it is clear that replacement exchange s are associated with price premiums that are of statistical and economic significance. The observed percentage pri ce effects of replacement ex changes, combined the predictions from the theoretical model have important implications for decision making regarding 1031 exchanges and suggest that participants in ta x-delayed exchang that have a short-term investment horizon need to be careful, since the value the the form of a higher replacement property price may offset, in pa rt the deferment of taxes. Third, the analysis o s suggests that significant price prem iums are often paid by out-of-state buyer These premiums differ substantially across resi dential and non-residential properti well as across geographic markets. On averag e, the price premium paid by out-of-state buyers is 16.7 percen nt in retail markets. Fourth, I study the effect that sale-leaseback transactions have on price. I find that the coefficient on the variable, indicating a sale-leaseback is generally positive and significant in the markets of potential interest Sale-leasebacks are as sociated with a 28.3

PAGE 146

135 percent price premium in office markets a nd a 23.6 percent pri ce premium in retail markets 1 These results present evidence supporti ng the hypothesis that sale-leaseback are associated with higher transaction prices. Fifth, I study the extent to which portfolio sales are associated with higher prices Portfolio sales are associated with price di scounts in two markets: apartments in San Francisco and retail properties in Seattle. When I exclude these two results, apartment and retail portfolio sales are associated w ith price premiums of 12.5 and 21.4 percent, respectively. These results present mixed evi s den ce that portfolio sales are associated with price ferent impacts on transaction prices in commercial real estate the premiums. Sixth, I also study the potential price impact of apar tment sales, motivated by condo converters. The analysis shows that c ondo conversions are asso ciated with a price premium of 16.4 percent in San Diego. In summary, the results demonstrate that exchanges, as well as various other investor motivations, have dif across different markets and property types. The results of this dissertation are especially important for the real estate discipline. Because of the exhaustive nature of dataset and my comprehensive analysis of fi fteen of the largest metropolitan markets in the United States, this work can be used as an important reference by appraisers for the adjustment of prices when a sales comparison approach is used. 1 Sale-leasebacks in San Diego are associated with a di scount of 14 percent, but they are not included in the calculation of average impact, since the result may be driven by outliers.

PAGE 147

LIST OF REFERENCES Allen, M. T., R. C. Rutherford and T. M. Springer, 1993, The Wealth Effects of Corporate Real Estate Leasing, Journal of Real Estate Research 8:4, 567-578 Alvayay, J. R., R. C. Rutherford and W. S. Smith, 1995, Tax Rules and the Sale and Leaseback of Corporate Real Estate, Journal of the American Real Estate and Urban Economics Association 23, 207-38 Black, R., and J. Diaz, III, 1996, The Use of In formation versus Asking Price in the Real Property Negotiation Process, Journal of Property Research 13:4, 287-297 Booth, G. G., J. L. Glascock and S. K. Sa rkar, 1996, A Reexamination of Corporate Sell : 195-202 Brealey, R., and C. Young, 1980, Debt, Taxes and Leasing A Note, Journal of Camp ation and Financial Deal Structuring: Ev idence from REIT Property Portfolio House Price Patterns: A Comp arison of Four Models, os Office Market: Price 83-106 CoStar Comps Professional, 2006, Available from www.costar.com//products/comps/ Offs of Real Estate Assets, Journal of Real Esta te Finance and Economics 12 Finance, 35, 1245-1250 bell, R. D., M. Petrova and C.F. Sirman s, 2003, Wealth Effect s of Diversific Acquisitions, Real Estate Economics 31:3, 347-366 Case, B., J. Clapp, R. Dubin and M. Rodr iguez, 2004,Modeling Spatial and Temporal Journal of Real Estate Finance and Economics, 29:2, 167-191 Clapp, J., 2003, A Semiparametric Met hod for Valuing Residential Locations: Application to Automated Valuation, Journal of Real Estate Finance and Economics, 27:3, 303-320 Colwell, P. F., H. J. Munneke and J. W. Trefzger, 1998, Chicag Indices, Location and Time, Real Estate Economics, 26:1 Accessed on July 26, 2006 Diaz, J., III and J. Hansz, 1997, How Valu ers Use the Value Opinions of Others, Journal of Property Valuation and Investment 15:3, 256 Diaz, J., III and M. Wolverton, 1998, A L ongitudinal Examination of the Appraisal Smoothing Hypothesis, Real Estate Economics 26:2, 349 136

PAGE 148

137 Diaz, J., III, R. Zhao and R. Black, 1999, Does Contingent Reward Reduce Negotiation Anchoring?, Journal of Property Investment and Finance 17:4, 374 Downs, D. H., and B.A. Slade, 1999, Char acteristics of Full-Disclosure, TransactionBased Index of Commerc ial Real Estate, Journal of Real Estate Portfolio Management 5:1, 95-104 Eppli, M. J., and J. D. Shilling, 1995, Speed of Adjustment in Commercial Real Estate Ezzell, J. R., and P. P. Vora, 2001, Leasi ng versus Purchasing: Direct Evidence on a Fickes, M., January 2003, exchanges do more than save taxes, National Real Fik, T., D. Ling and G. F. Mulligan, 2003, Modeling Spatial Variation in Housing Fisher, L. M., 2004, The Wealth Effects of Sale and Leasebacks: New Evidence, Real Estate Economics, 32:4, 619-643 Frew, J., and D. G. Jud, 2003, Estimati ng the Value of Apartment Buildings, Journal of Real Estate Research 25:1, 77-86 Forge closure state Research, 9:3, 313-318 ort from l Estate ciation 19, 1-24 eal Estate and Urban Economics Association 19, 567-582 Glower, M., D. Haurin and P. H. Hendershott, 1998, Selling Time and Selling Price: The Influence of Seller Motivation, Real Estate Economics 26:4, 719-740 Halvo y Variables in Semilogarithmic Equations, American Economics Review 70:3, 474-475 Markets, Southern Economic Journal 61, 1127-45 Corporations Motivations for Leasi ng and Consequences of Leasing, Quarterly Review of Economics and Finance 41, 33-47 Estate Investor 59-62 Prices: A Variable In teraction Approach, Real Estate Economics 31:4, 623-646 y, F. A., R. C. Rutherford, and M. L. VanBuskirk, 1994, E ffects of Fore Status on Residential Selling Price, Journal of Real E Geltner, D., 1989, Estimating Real Estates Systematic Risk from Aggregate Level Appraisal-Based Returns, Journal of the American Real Estate and Urban Economics Association 17:4, 463-481 Geltner, D., B. D. Kluger and N. C. Miller, 1988, Optimal Price and Selling Eff the Perspectives of th e Broker and the Seller, Journal of the American Rea and Urban Economics Asso Glascock, J. L., W. N. Davidson, III and C. F. Sirmans, 1991, The Gains from Corporate Selloffs: The Case of Real Estate Assets, Journal of the American R rsen, R., and R. Palmquist, 1980, The Interpretation of Dumm

PAGE 149

138 Hardin, III, W. G., and M. L. Wolverton, 1996; The Relationship between Foreclosure Status and Apartment Price, Journal of Real Estate Research 12:1, 101-10 9 o 113-126 Harris L., and E. Gurel, 1986, Price and Volume Effects Associated with Changes in the Haurin, 1988, The Duration of Marke ting Time of Residential Housing, Journal of the Hendershott, P. H., and D.C. Ling, 1984, Trading and the Tax Shelter Value of Holmes, A., and B. A. Slade, 2001, Do Tax-Deferred Exchanges Impact Purchase Price? l/uscode26/usc_sec_26_00001001----000Hardin, III, W. G. and M. L. Wolverton, 1999; Equity REIT Property Acquisitions: D Apartments REITs Pay a Premium?, Journal of Real Estate Research 17, S&P 500 List: New Evidence for the Existence of Price Pressures, Journal of Finance, 16:4, 815-829 Hess, A. C., and P. A. Frost, 1982, Tests fo r Price Effects of New Issues of Seasoned Securities, Journal of Finance, 37, 11-25 American Real Estate and Urban Economics Association 16:4, 396-410 Depreciable Real Estate, National Tax Journal 37, 213-223 Evidence from the Phoenix Apartment Market, Real Estate Economics, 29:4, 567 588 Internal Revenue Code, Title 26, Section 1001(c), Available from http://www.law.cornell.edu/uscode /htm .html Accessed June 26, 2006 Intern http://www.law.cornell.edu/uscode /html/uscode26/usc_sec_26_00001031----000al Revenue Code Section, T itle 26, Section 1031, Available from .html Accessed June 26, 2006 C., and A. Orr, 1999, Local Commer c Jones ial and Industrial Rental Trends and Property Market Constraints, Urban Studies, 36:2, 223-237 Kraus -588 nd A., and H. R. Stoll, 1972, Price Im pacts of Block Trading on the New York Stock Exchange, Journal of Finance 27, 569 Lambson, V., G. McQueen and B. Slade, 2004, Do Out-of-State Buyers Pay More for Real Estate? An Examination of Anchor ing-Induced Bias and Search Costs, Real Estate Economics, 32:1, 85-126. Lasfer, M. A., and M. Levis, 1998, The Dete rminants of Leasing Decisions of Small a Large Companies, European Financial Management 4, 159-184 Lewellen, W., M. Long and J. J. McC onnell, 1976, Asset Leasing in Competitive Capital Markets, Journal of Finance 31, 787-798

PAGE 150

139 Ling, D. C., and M. J. Whinihan, 1985, V aluing Depreciable R eal Estate: A New Methodology, AREUEA Journal 13:2, 181-194 Lewis, C. M., and J. S. Schallheim, 1992, Are Debt and Leases Substitutes, Journal of Financial and Quantitative Analysis 27:4, 497-511 McBurney, C. M., 2004, Section 1031 Exch anges: A Legitimate Tax Shelter for Business, Thompson FindLaw, May 2004 McBurney, C. M., and S. Boshkov, 2003, Is an Attribute of an Intangible in a 1031 onomics 10, 299-306 McLi al Real Estate Investor 45-47 Mikk dary Distributions, Journal of Financial Economics 14, 110-114 Capital Services, Moyer, R. C., and V. S. Krishnan, 1995, Sale and Leaseback Transa ctions: The Case of 34, 46-59 Munneke, H. I., and B. A. Slade, 2000, An Empi rical Study of Sample Selection Bias in Indices of Commercial Real Estate, Journal of Real Estate Finance and Munneke, H. I., and B. A. Slade, 2001, A Metropolitan Transaction-Based Commercial :1, National Residential Real Es tate Appraisal Institute, 2006, Northcraft, G. B., and M. A. Neale, 1987, Expert, Amateurs, and Real Estate: An ess 39, 228-241 Exchange a separate asset or i nherent in an Existing Asset, Journal of Taxation, 98:1, 46-53 McIntosh, W., S. H. Ott and Y. Liang, 1995, The Wealth Effects of Real Estate Transactions: The Case of REITs, Journal of Real Estate Finance and Ec nden, S., June 2004, Exchanges Test the Waters, Nation elson, W. H., and M. M. Parch, 1985, S tock Price Effects and the Costs of Secon Miller, M., and C. Upton, 1976, Leasing, Buying and the Cost of Journal of Finance, 31, 761-786 Electric Utilities, Quarterly Journal of Business and Economics, Economics, 21:1, 25-64 Price Index: A Time Vary ing Parameter Approach, Real Estate Economics 29 55-84 Myers, S. C., D. Dill and A. Bautista, 1976, Valuation of Financial Lease Contracts, Journal of Finance, 31, 799-820 http://www.nraiappraisers.com/mem bers/DEFINITION_OF_MARKET.pdf Anchoring-and-Adjustment Perspect ive on Property Pricing Decisions, Organizational Behavior and Human Decision Proc

PAGE 151

140 Quan, D., and J. Quigley, 1991, Price Form ation and the Appraisal Function in Re Estate Markets, Journal of Real Estate Finance a al nd Economics 4:2, 127-146. earch 9:2, 151-167 re and ournal of Business 45, 179-211 r Sirmans, C. F., G. K. Turnbull and J. Dombrow, 1995, Quick House Sales: Seller Slade, B., 2004, Conditions of Sale Adjustme nt: The Influence of Buyer and Seller Motivations on Sale Price, The Appraisal Journal Winter 2004, 50-56 Slovic, P., and S. Lichtenstein, 1971, C omparison of Bayesian and Regression Approaches to the Study of Information Processing in Judgment, Organizational Slovin, M., M. Sushka and J. Polonchek, 1990, Corporate Sale and Leasebacks and Smith, C. W., Jr. and L. M. Wakeman, 1984, Determinants of Corporate Leasing Starker vs. United States, 602 F. 2d 1341 (9th cir., 1979) Turnb s, ban Economics 23:4, 545-557 Valente, J., S. Wu, A. Gelfand and C. F. Sirmans 2005, Apartment Rent Prediction Wayner, S. A., 2005a, Exchanges Defer Tax Bills and Boost Broker 31-32 Wayner, S. A., 2005b, Section 1031 Exchanges: Underused Tax-Planning Tool, CPA Webb f 11:1, 62-65 Sandeion, Z., B. Smith and C. Smith, 1993, An In tegrated Approach to the Evaluation of Commercial Real Estate, Journal of Real Estate Res Scholes, M. S., 1972, The Market for Securities : Substitution versus Price Pressu the Effect of Information on Share Prices, J Shilling, J., J. Benjamin and C.F. Sirmans, 1990, Estimating Net Realizable Value fo Distressed Real Estate, Journal of Real Estate Research, 5:1, 129-140 Mistake or Luck, Journal of Housing Economics 4, 230-243 Behavior and Human Performance 6, 649-744 Shareholder Wealth, Journal of Finance 45, 289-299 Policy, Journal of Finance 40, 895-908 ull, G. K., and C. F. Sirmans, 1993, Information, Search and House Price Regional Science and Ur Tversky, A., and D. Kahneman, 1974, Judgmen t under Uncertainty: Heuristics and Biases, Science New Series 185:4157, 1124-1131 Using Spatial Model, Working Paper Commissions, Real Estate Finance, February 2005, Journal June 2005, 16-17 B., 1994, On the Reliability of Co mmercial Appraisals: An Analysis o Properties Sold from the Russell-NCREIF Index (1978), Real Estate Finance

PAGE 152

141 Weirick, W., and F. Ingram, 1990, Functional Form Choice in Applied Real Estate Analysis, Appraisal Journal, January 1990, 57 Wheaton, W. C., and R. G. Torto, 1990, An Investment Model of the Demand and Supply for Industrial Real Estate, Journal of the Americ an Real Estate an d Urban Economics Association, 18, 530-547

PAGE 153

BIOGRAPHI CAL SKETCH Milena Petrova has a Bachelor of Intern ational Business Relations from UNWE in Sofia, Bulgaria, a masters degree in busine ss administration with a major in international business from the Helsinki School of Economic s, Finland, and a Master of Science in finance from Hofstra University, NY. She has four years experience in finance and management working as an analyst for a specialized finance company and for a management consulting firm. She complete d the requirements for the Doctor of Philosophy degree (in finance) during the summer of 2006. Upon graduation from the University of Florida she will join Syracuse University, NY, as an assistant professor in finance. 142