Citation
Bankruptcy studies

Material Information

Title:
Bankruptcy studies empirical works on prediction and financial markets
Added title page title:
Empirical works on prediction and financial markets
Creator:
Bi, Keqian, 1941-
Publication Date:
Language:
English
Physical Description:
v, 158 leaves : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Assets ( jstor )
Bankruptcy ( jstor )
Bond rating ( jstor )
Business structures ( jstor )
Finance ( jstor )
Financial risk ( jstor )
Investment risks ( jstor )
Modeling ( jstor )
Predictive modeling ( jstor )
T tests ( jstor )
Bankruptcy -- Forecasting ( lcsh )
Money market -- Mathematical models ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1989.
Bibliography:
Includes bibliographical references (leaves 155-157).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Keqian Bi.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute 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.
Resource Identifier:
001500652 ( ALEPH )
AHB3389 ( NOTIS )
21387281 ( OCLC )

Downloads

This item has the following downloads:


Full Text



















BANKRUPTCY STUDIES: EMPIRICAL WORKS ON
PREDICTION AND FINANCIAL MARKETS







By

KEQIAN BI


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


1989










ACKNOWLEDGEMENTS

I am greatly indebted to the members of my dissertation

supervisory committee: Dr. Roy L. Crum and Dr. Haim Levy,

committee cochairmen, and Dr. David Denslow. Their support,

advice, and guidance have made this dissertation a reality and

have greatly enhanced my academic experience at the University

of Florida.

I would also like to thank my family for their unwavering

support, without which this dissertation would have been

impossible. Even as I write now, I think of my kind, loving

parents, Professors Zhongjie Bi and Shaoxiang Huang, who,

though half a world away, have wished so much for my success

in this Ph.D. program. I want to share it today with many

people, but with them most of all.

Finally, I want to dedicate my dissertation to a new

future for my homeland, China. The lessons and memories of

May and June 1989 will be remembered forever.












TABLE OF CONTENTS


ACKNOWLEDGEMENTS .

ABSTRACTS. ..... .

CHAPTERS


1 OVERVIEW AND OUTLINE. ... 1
Topic Overview. 1
Dissertation Outline. .. .. 5
Note. 7

2 LITERATURE REVIEW ... .. 8
Summary of Current Literature 8
Significance of this Dissertation .. 14

3 METHODOLOGY TO STUDY FINANCIAL MARKETS
AND BANKRUPTCY .. 17
Note 22

4 RESULTS ON FINANCIAL MARKETS AND BANKRUPTCY 27
Notes .. 35

5 EMPIRICAL MODEL OF BANKRUPTCY .. 48
Variable Definitions. ... 51
Hypotheses. ..... .. 57

6 METHODOLOGY TO TEST EMPIRICAL MODEL .. 62
The Logist Analysis Methodology .. 62
Logist versus Discriminant Analysis .. .64
Sample Design .. .. 67


7 RESULTS OF EMPIRICAL MODELS OF BANKRUPTCY

8 SUMMARY AND CONCLUSIONS ..
Research Summary. .
Future Research .

APPENDIX .. ..

REFERENCES .. .

BIOGRAPHICAL SKETCH. .


. 74

. 99
. 99
. 102

. 104

. 154

. 157


iii


o o







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

BANKRUPTCY STUDIES: EMPIRICAL WORKS ON
PREDICTION AND FINANCIAL MARKETS

By

Keqian Bi

August 1989


Chairman: Roy L. Crum
Cochairman: Haim Levy
Major Department: Finance, Insurance, and Real Estate

This dissertation studied two areas not covered by

current bankruptcy research. First, the question whether

financial markets could assess bankruptcy risk beforehand was

addressed. Second, the problems of long-term bankruptcy

prediction models were discussed, and a new model with

reasonably accurate long- as well as short-term predictive

powers was developed.

On the subject of financial markets, this dissertation

looked at whether the market could assess bankruptcy risk

ahead of time. The risk premia of corporate bonds of bankrupt

companies were used to investigate whether there was any rise

in risk premia as companies approached bankruptcy. Then, an

event study was done to determine whether bond downgrades

affected the performance of a later-bankrupt company's stock's

daily rates of return.





It was found that risk premia did increase as companies

approached bankruptcy, and that the risk premia of companies

which were later liquidated rose more than the risk premia of

companies which later re-organized. The event-study showed

that the daily rates of returns of stocks of companies which

later went bankrupt did fall significantly during the first

downgrade, unlike those of companies which did not.

On the subject of bankruptcy prediction, this

dissertation looked at whether an accurate, long-term

bankruptcy model could be developed using variables which

assess a company's fundamental characteristics as well as its

current financial position. A model based on the following

eight variables was developed: return on assets, fixed charge

coverage, balance ratio, market-to-book ratio, relatedness

ratio, net rate of management stock acquisitions, relative

sales growth, and capital intensity. It had achieved high

classification success and predictive accuracy, and, unlike

those of past models, its predictive powers did not decrease

over time.

Finally, work was done on the use of logist analysis in

bankruptcy predictions. It was shown that any non-random,

state-based sample selection technique would distort the

probabilities generated by logist models. A formula was

derived to remedy this problem by providing an adjustment

scheme whereby the resulting probabilities could be made to

work for a general population.













CHAPTER 1
OVERVIEW AND OUTLINE



Topic Overview



Bankruptcy is one of the most important topics in modern

finance. It plays a strong, visible role in all dimensions

of financial economics, such as the efficient market theory,

portfolio theory, capital asset pricing theory, option pricing

theory, and agency theory. Thus, understanding bankruptcy is

important to financial research in both theory and practice.

In theory, if we thoroughly understood the dynamics and causes

of bankruptcy, we should be able to make the risk of

bankruptcy a parameter in market valuations of debt and

equity. Assuming that investors are naturally risk averse,

such formulas can determine how the probability of bankruptcy

affects the average investor's utility. In practice,

researchers have tried to develop predictive models of

bankruptcy that can alert interested parties to the impending

dangers of bankruptcy before it is too late to take corrective

actions. Because practically any group involved with a

company would be interested in the risk of bankruptcy it





2

faces, potentially widespread demand for such a model has

inspired much research in this area. To date, several

approaches have been taken to develop predictive models; they

are discussed in the Literature Review section.

While substantial work has been done on various aspects

of bankruptcy, including empirical forecasting and prediction,

we believe that past works have left a few areas unexplored.

First, many successful predictive models of bankruptcy have

been developed, but most of these models base their assessment

of a company's bankruptcy risk primarily on its current

financial position. They rely heavily on such accounting data

as retained earnings and capitalization ratios. Some of these

predictive models have grown quite complex. Because they use

mostly current financial data, however, their predictive power

is limited to a relatively short period before bankruptcy,

usually one year. When they are used to predict bankruptcy

further into the future than one year, their accuracy falls

significantly.

Second, no one, to date, has studied whether the market

can somehow predict bankruptcy. Modern finance considers

financial markets to be efficient, but no research has yet

been done on whether this efficiency extends to the evaluation

of companies' bankruptcy risk. After all, if the market were

truly efficient, then it should include the risk of bankruptcy

as one of the determinants of the value of a company's

securities. Hence, the market's assessment of a company's






3

securities' risk should include as a component the company's

chances of bankruptcy.

In this dissertation, we address both of these areas that

present bankruptcy research has largely overlooked. First,

we explore the market's ability to "sense" coming bankruptcy

by looking at the risk premia of corporate bonds. We first

examine whether such risk premia rise as the event of

bankruptcy approaches. Then, we investigate what role the

bond rating agencies, Standard & Poor's and Moody's, play in

the market assessment of a company's overall risk in general

and bankruptcy risk in particular. In other words, if the

market actually does include bankruptcy risk in its overall

assessment of a company's risk, then how do S&P's and Moody's

ratings affect this market assessment? Do they provide new

information to the market and, therefore, serve as a crucial

link in the process, or does the market itself already reflect

all of the information these rating agencies provide?

To answer this question we have to study two problems.

First, assuming that the first consistent downgrade of a

corporate bond provides the most relevant information, we

determine whether S&P's and Moody's downgrades precede or

follow increases in the bond's risk premium.1 Second, with

the same assumption, we explore what impact, if any, the first

bond downgrade has on the performance of the company's stock.

Arguably, if nothing happens to the stock's return or the bond

risk premium after a bond downgrade, then we can infer that





4

the market has already absorbed all the information S&P's and

Moody's provide in the rating change. In that case, we can

say that S&P's and Moody's provided no new information in the

market's assessment of a company's bankruptcy risk. If,

however, the market reacts strongly after a bond downgrade and

the bond's risk premium rises or the stock return changes

significantly with the downgrade, then we can say that S&P's

and Moody's do provide new information to the market in its

assessment of bankruptcy risk. In that case, we can further

conclude that bond downgrades are good indicators of increased

bankruptcy risk and correspondingly higher risk premia.

To study bond risk premia and bankruptcy, we first

compile the yield-to-maturity of bonds of companies that went

bankrupt and subtract from them the yield to maturity of

government bonds. Then, to assess the impact of bond rating

downgrades, we use the methodology known as "event study,"

which is described in greater detail later in Chapter 3.

The second area that has largely been overlooked in

bankruptcy research is the "nearsightedness" problem of

current predictive models of bankruptcy. We attempt to

overcome this shortcoming by building an empirical model that

includes new sets of variables with a long-term orientation.

We believe that as one tries to predict bankruptcy farther

ahead in time, a company's current financial position becomes

less significant, while other, currently unexplored factors

play increasingly larger roles. For example, a company's






5

current return on assets (ROA) may be critical to whether it

is solvent or bankrupt within the next few years, but its

long-term financial health would probably depend less on

current profit levels than on the fundamental characteristics

of the company that predetermine future ROA. Hence, to make

accurate long-term bankruptcy forecasts, we must examine not

only a company's current financial position, but also its more

fundamental operating characteristics, such as its lines of

business, degree of diversification, management efficiency,

and growth.

We adopt this approach to develop an empirical model of

bankruptcy that can forecast the extent of bankruptcy risk.

In addition to a group of four variables designed to assess

a company's current financial position, we adopt variables

that describe a company's management ownership position, its

degree of diversification, the lines of business in which it

operates and their relationship, and its growth relative to

the rest of the industry. We expect that a model based on

this broader specification of variables will have high long-

term as well as short-term predictive powers.


Dissertation Outline


In the following chapters, we present the research design

and findings of our dissertation. Chapter 2 examines current

literature on both bankruptcy prediction and the effects of

bond rating changes, which, as stated earlier, are used to





6

study financial markets' signals about bankruptcy. Chapter

3 discusses the techniques and methodologies we employ to

study the financial market's ability to sense impending

bankruptcy. Then Chapter 4 identifies the data sources for

our work on financial markets and bankruptcy and presents the

results we obtained in the market-oriented part of the

dissertation. Chapter 5 discusses in depth the rationale for

and the variables used in our empirical model of bankruptcy.

Attention is focused on why each variable was chosen, what

values each variable should have, and what we expect our model

to tell us about the role each of our variables should play

in assessing bankruptcy risk. In Chapter 6 the methodology

we used to build our empirical model is described. This is

logist analysis, and we will discuss the advantages it has

over the more traditional discriminant analysis. Chapter 7

gives the results we obtained from tests of our empirical

model. Finally, Chapter 8 summarizes our research, presents

the conclusions we reach, and points out possibilities for

future research.












Note

1. "Consistent" means a bond downgrade after which there were
no upgrades until the company filed for Chapter 11. In other
words, this downgrade is the first of a series of downgrades
which eventually lead to bankruptcy, uninterrupted by any
upgrades of the same bond.














CHAPTER 2
LITERATURE REVIEW


Summary of Current Literature



Market Efficiency Studies of Bond Rating Changes



Katz (1974), in one of the earliest works on bond rating

changes, developed an event-oriented methodology for testing

the efficiency of the bond market. He looked for "unusual

behavior" in a bond's yield to maturity twelve months prior

to and five months after a rating change. His data consisted

of electric utilities bonds from 1966 to 1972. Katz derived

a quadratic regression equation of yield to maturity at any

given time, t, based on maturity, total float, and coupon

rate. Then, he compared his expected yields with the actual

yields and the changes in the actual yields with premium

differentials of two rating classes. He concluded that no

anticipation exists prior to a public announcement of a rating

change. After the rating change, there was a lag of six to

ten weeks before yield-to-maturity fully adjusted to the new

rating class.




9

Weinstein (1977) tried to determine if bond rating

changes contained new information by studying the bonds'

prices during the time period surrounding rating change

announcements. His sample consisted of utilities and

industrial bonds from July 1962 to July 1974. Weinstein

started with portfolios which, for every month, contained all

bonds with a given rating. He then constructed a series of

risk-adjusted returns for each bond by subtracting the return

on the appropriate rating class portfolio from the return on

the given bond. He selected the bonds that had a rating

change and looked at if those bonds had abnormal returns

during periods of rating changes. Weinstein concluded that

bond rating changes caused no significant price change during

or after the announcement, and that adjustments in the market

were made 18 to six months before the event. Hence, his study

suggested that rating changes provided no new information.

Pinches and Singleton (1978) studied the effects of bond

rating changes on the market returns of stocks during the

period from January 1950 to September 1972. For each stock,

they derived a market return based on its beta and measured

the actual return against the expected return for a period of

thirty months before to twelve months after a rating change.

Their study calculated disturbance terms (residuals) of stock

returns during the period. Pinches and Singleton concluded

that all changes attributable to companies' financial

situations were fully anticipated 15 to 18 months ahead of





10

time, while all changes attributable to company-specific

events were anticipated six months ahead of time. Thus,

although there were abnormally high and low returns

corresponding to upgrades and downgrades, respectively, before

a rating change, there were normal returns after the rating

change. Again, a study concluded that bond rating downgrades

provided no new information to the market.

Finally, Griffin and Sanvincente (1982) used three

different methodologies to study the effects of rating changes

on common stock prices. Their study contained 180 rating

changes from 1960 to 1975. First, they used a portfolio

method similar to that of Weinstein (1977). Then they

employed a one-factor and a two-factor model, basing their

expected stock prices on betas, as had Pinches and Singleton

(1978). They found that although rating upgrades had no

effect on stock prices, downgrades did have significant

effects. Because of the inconclusive nature of their results,

further research in this area was necessary.

The methodology employed in this dissertation differs

from previous works in several important ways. First,

previous authors used yield-to-maturity, an absolute value,

as their indicator of return. It is our position that

absolute yield-to-maturity, in this application, is not an

accurate measure of return. Instead, we suggest using a

relative value, the risk premium, which is defined as the

difference between a bond's yield-to-maturity and the yield-






11

to-maturity of a risk-free security. Second, we construct our

samples not by industry but by the nature of the event. In

other words, we defined the event as the filing of Chapter 11

under the Federal Bankruptcy Act. As far as we know, this is

the first study of bond rating changes to be based on data of

companies from all industries.




Theoretical Models of Bankruptcy


Wilcox (1971) is one of the earliest and most primitive

theoretical models of bankruptcy. It assumes that a company

starts with a positive amount of capital, K, which changes

randomly over time. Positive changes in K indicate positive

cash flow and increases in the company's assets, while

negative changes in K indicate financial losses which require

the company to liquidate assets. When a company's K is

sufficiently negative, it becomes bankrupt. Expressions for

the expected probability of bankruptcy, as well as time to

bankruptcy, are mathematically derived, just as they would be

for the gamblers' game.

Scott (1976) and (1977) attempted to improve on this

simple model. Scott's early models assumed that a company has

a potentially infinite life and can meet losses by selling

debt or equity in an efficient market without incurring

flotation costs. They further assumed that the secondary

market for real assets is imperfect and that a firm begins






12

with an optimal level of assets. Therefore, it would much

rather sell securities and debt than liquidate assets to cover

its losses. Scott then showed that a company would remain

solvent as long as stockholder wealth, measured by market

value, remained positive.

Scott (1981) developed a revised version of the earlier

model. In this newer model, Scott assumed that a company may

have imperfect access to external capital, so it might incur

flotation costs when it sells securities, or there may be a

tax system which favors internally-financed corporate

investments. Further, systematic imperfections in the market

valuation of securities can hinder corporate access to

external capital. This model, however, also assumed that the

company has no debt and can issue only equity. Thus,

according to this model, a company will go bankrupt when the

market value of its securities is less than the amount of

investment needed at times of negative income. Therefore,

bankruptcy is not the result of a conflict of the benefits and

costs of debt, but rather the product of investment managers'

mistakes.



Empirical Works on Bankruptcy



Beaver's 1966 paper was the first empirical work that

tried to build a predictive model of bankruptcy. He looked

at 30 accounting ratios which could be used to predict






13

bankruptcy, and for each ratio he derived a cut-off point for

bankruptcy. He concluded that three ratios were the best

predictors of financial failure: Cash Flow/Total Assets, Net

Income/Total Debt, and Cash Flow/Total Debt.

Altman, Haldeman, and Narayanan (1977), a follow-up of

Altman (1969), used the more complex multi-variate

discriminant analysis approach to build a predictive model.

Their work included all industrial failures from 1969 to 1975

with at least $20 million in assets, which made a sample of

53 bankrupt firms, and Altman et al. found a matching sample

of 58 non-bankrupt firms. The samples were matched by

industry, year of bankruptcy, and size of assets. Their model

included seven variables: return on assets (ROA), stability

of earnings, debt service (times-interest-earned or TIE),

cumulative profitability, current assets/liabilities ratio,

capitalization, and size. After using various statistical

techniques, Altman et al. derived a value ZETA as the cut-off

for bankruptcy. This model, commonly known as the ZETA model,

is highly accurate, especially when bankruptcies are near.

Today, it is the leading model for predicting bankruptcy, and

because many financial institutions use it, it has become an

industry standard.

Ohlson (1980) took another approach to bankruptcy

prediction by using logist analysis to build his model. His

sample included 105 failed firms, but he did not find a

matching sample by asset size. Hence, among his nine





14

variables, size became the most significant one. Since his

model had error rates of 17.4% for non-bankrupt (type I error)

and 12.4% for bankrupt firms (type II error) even just one

year before bankruptcy, it has remained more-or-less an

academic curiosity and has not attained the same widespread

use as Altman's ZETA model.

Zavgren (1985) extended Ohlson's work by including more

variables and extending the length of the study. Her work

looked for the important factors in the short- and long-term

predictions of bankruptcy. Zavgren found that profitability

was not significant in either the short- or long-run. Rather,

her study showed that the ability to meet obligations is

significant in the short-run, while efficiency ratios and

liquidity are important in the long-run. Zavgren's study,

then, reduces a company's bankruptcy risk to two issues, that

of short-term endurance (as measured by the ability to meet

obligations) and fundamental characteristics (as measured by

the efficiency ratio and basic liquidity.)



Significance of this Dissertation


While the existing literature is already quite advanced,

we believe a few areas have been left unexamined. First,

while work has been done on bond rating changes, there has

been no research that tries to link bankruptcy with financial

market reactions. No one to date has looked at the trend that






15

bond risk premia take as a company approaches bankruptcy, even

though bankruptcy risk, in theory, should be a primary risk

included in risk premia. Further, while researchers have

studied bond rating changes' effects on both the stock and

bond markets to see if such rating changes contained new

information, no study has linked the information these

downgrades provide with a company's risk of filing for Chapter

11 and thus declaring bankruptcy. Since bond downgrades are

meant to warn investors of possible default and bankruptcy,

whether such downgrades have any impact on financial markets

should be directly linked to the market's assessment of a

company's bankruptcy risk. So far, however, research has left

this area untouched.

Second, as we stated in the previous Chapter, current

bankruptcy prediction models have looked mostly at current

financial data. Only Zavgren (1985) has tried to examine

certain fundamental characteristics, and her study shows that

such information does indeed have a role in empirical studies

of bankruptcy, especially when we are dealing with long-term

bankruptcy prediction. Hence, we assert that fundamental

characteristics have largely been overlooked by present

bankruptcy research, and our dissertation will address this

area more systematically than Zavgren did.

Our research contributes to the financial research of

bankruptcy and market efficiency in several ways. By studying

the bond market and bankruptcy, we attempt to determine





16

whether the financial market can adequately assess bankruptcy

risk on its own and whether bond ratings play a part in this

assessment of risk. We then look at ways to augment or

reinforce market signals via prediction models with

significant early warning capabilities. In this regard, our

study covers a period longer than those of its predecessors.

Further, we introduce variables that assess the fundamental

characteristics of a company to forecast long-term bankruptcy.

On the theoretical side, this research can lead to

establishing a relationship between certain fundamental

characteristics of a company and its financial position a few

years into the future.

In the area of market efficiency, we attempt to assess

whether the markets are truly efficient in anticipating one

specific type of risk--bankruptcy risk. Further, in our event

study, we study whether the bond rating downgrades actually

do provide new information to the market. We do not, however,

do this by merely looking at whether the downgrade trailed or

led a rise in the risk premium, because we believe such

indications are in themselves not significant. After all, a

downgrade that trails a rise in risk premium might be regarded

as the leading downgrade to a subsequent rise in the risk

premium. Hence, we will instead concentrate on whether

downgrades make a significant impact on the market.















CHAPTER 3
METHODOLOGY TO STUDY FINANCIAL MARKETS AND BANKRUPTCY



Our research on financial markets and bankruptcy

encompasses two topics. First, we look at the trend of the

risk premia of corporate bonds of companies that later went

bankrupt. This trend tells us if the markets can correctly

assess increasing chances of bankruptcy and default as

bankruptcy nears. Then we examine the impact of bond

downgrades on the returns of a company's stock. We determine

if ratings play a significant role in providing the market

with new information.

To study whether bond risk premia increase as a company

approaches bankruptcy, we selected a sample of bonds based on

two criteria. First, they had to be publicly traded bonds

listed in the Standard & Poor's Bond Guide with a bond rating

from either S&P's or Moody's. The second criterion was that

the companies which issued the bonds later filed for Chapter

11 between September 1977 and October 1988. Only 50 corporate

bonds had bond ratings and other available data from S&P's or

Moody's adequate for our purposes. These 50 bonds, listed in

Table 1, are used to study risk premia and bankruptcy. Then,





18

from the Analytical Record of Bond Yields and Yield Spreads,

published by the Salomon Brothers, we obtained the monthly

yield-to-maturities for US government securities. The risk

premium for each of our companies is the excess of the yield-

to-maturity of its bond over the yield-to-maturity of an US

government bond with the same maturity.

We then studied the risk premia of our 50 companies as

they moved towards bankruptcy. We separated the companies

into two groups--those companies which later re-organized and

those that were later liquidated.1 Next, we compared the risk

premia of bonds in those two groups as the companies

approached bankruptcy to see if the market's assessment of

risk went so far as to differentiate companies which could

later re-organize from those that could not.

To study the impact of bond downgrades, we use the event

study methodology. An event study compares the impact of an

event on security holders with the predictions made by a model

that approximates what would have happened if the event had

not taken place. In effect, we try to compare what happened

with what a model tells us should have happened. Our event

is the first consistent downgrade rating change, and the size

of the impact is measured by the disturbance of the stock's

daily rates of return. The number of months between the first

downgrade bond rating change and the month of filing Chapter

11 for our sample is given in Table 2.

In our work, the Mean Adjusted Returns Model, as






19

discussed in Fama (1976) and Masulis (1980) is used as the

basis for the statistical studies. This model uses the mean

returns on an individual stock over a representative period

of time before the event period to estimate a stock's expected

mean return. This "comparison period" is then compared with

the daily rates of return over the period of rating change.

Because our research focuses on financial distress and

bankruptcy, both of which are long-term processes, we selected

a two-year time period before the event as the "comparison

period." Since there are a different number of business days

in any given year, we simplified things by defining 510

business days as "two years." Thus, our comparison period is

510 days to 11 days before the rating change. The actual

event, the bond rating change, is taken as the 21-day period

beginning ten days before and ending ten days after the

downgrade announcement.

For this study, we construct a sample of companies that

were listed on the New York or American Stock Exchanges, that

had bond downgrades, and that later filed for bankruptcy. We

then found a matching sample of companies that satisfied the

first two criteria but did not later file for bankruptcy.

Once we have used our comparison period to determine a

stock j's mean daily return, Aj, the event period disturbance

term Et, which measures the impact caused by the bond rating

change, is given by:

Ejt = Rjt ~j,





20

with t being the date in the comparison period,
t = -10..10,
where Rj is the realized daily return of the
stock j at time t, which was read from the
CRSP Daily Return Tape.

The average disturbance term for N events (firms) is:

avg E, = 1/N (E Eit)

with j = 1..N.

The null hypothesis, avg Et equals zero, means that a downward

bond rating change has no effect on shareholders' daily

returns. Since we believe that the first downgrade of a bond

gives the earliest signal and, hence, the most information to

stockholders about the risks of financial distress, we expect

the null hypothesis to be rejected for our Chapter 11 sample.

Conversely, we expect not to be able to reject the above null

hypothesis for our matching sample of firms that did not file

for Chapter 11.

The variance of avg E. is:

Var (E) = (1/499) E (avg Et X(E))

where X(E) = (1/500) Z (avg Et)

When t=0, we would be testing for the disturbance on event

date. The t-statistic used to determine whether avg E0

differs significantly from zero with 499 degree of freedom is:


t = avg E0 / J (Var (E))

The cumulative error over a particular event time interval is:

CE(a,b) = Z (avg Et)

where -10 5 a < b < +10






21

We hypothesize that the cumulative errors in the bankrupt and

matching groups are statistically different, implying

different effects of bond rating changes on stock rates of

return.

Finally, we separate the Chapter 11 group into two sub-

groups: one sub-group of companies that filed for Chapter 11

and later re-organized and another of companies that filed but

were later liquidated. We compute an average Et for bond

downgrades of companies in the two sub-groups. Then, we

employ the t-test and F-test to analyze the differences for

statistical significance between the two sub-groups. The null

hypothesis is that the means for the two sub-groups of

companies should become statistically equivalent. By

reasoning explained previously, we expect that the null

hypothesis will be rejected. Thus, we expect the means for

the two sub-groups of companies to become statistically

significantly different as we approach the filing of Chapter

11.






22




Note

1. A company, by our definition, is "liquidated" if it
satisfies one or more of these conditions:

1. acquired by another company or liquidated
2. listed on COMPUSTAT as "bankrupt"
3. no longer on the Wall Street Journal Index
4. no longer on the Predicasts F&S Index of Corporate
Change
5. no longer on Q-file
6. no longer on the Directory of Corporate Affiliations
7. no longer listed on the C-D System
8. no longer described in the Chapter 11 Report















) id a 4 tiBi ( .- .0 Q d(ISo U 4)U 0) ?30 A 4l4 m4 4 to
vH m oiWw 0%0m0oo if o fn oo0fo 0rf



00 D0H000H0OHD000OO00C Ol00 COD000000OD0


0
('V 0

o0

o I


0 (
HMH
0 0


U) .Q M
QF)Q

Cc Q)
VI[11


o 0

-N -
0 0
*O OOLA cc
cN mcO in o( co O
0o N- o- o- -ro -d
O0r c HH -%D c- O -
I \OCN I ON *0 ri P HoPOO dC O 14 a CA 'c *P \
r-4 N m- Un r- ODOD 0 d\ CD O% oiPOP ODCO mv CODM
Hi'r, Hl 0> \ r- N 0n s- CO'ON- 00- cp 0 \- I
H W r-\ C nL d m O *NC 'o O in o r0Co
-'t I dpa O oC I 4 M I mdP O U i- I N H I0
(A 4) wt 0 Ns O ) Od- sk v r l a) de 0 1 N NIm% I q O I -- I 0
i- a-r L H c inLO cccHmEDODLO HHLOMinopLL\H I H A CM
I \* I \\H I I ainn o 0 -
n nl Hr- I N H H ( I I ) l ) aQ r ) 0
H m -4QQ-cc Q Qo QQ~0) Q0A40
QA I .Q Q QA M
4i cPcWc) 5c 0 #)00 4 Wc c r4 0) IQ o ) 4) wc IQ0
aM 0 1 <)-I p 4 a w 0 EaQ : (U)Q l Qw m copa

s r5 s s s ^ as s a s s s s A


U
*4


0


p- HQ)
HM .) H-H
H .) -,-
.0 4 0 G 3
HH MH 0 0

rH 0 P H ca





HU
O 4 OHOO
:3 -rq Ea 0 (o ca
0 *- C *P-
*. OW *Hc-

0 Q 0 0d 04 04
0C OH M CO --
H4-4 1 P r- HH 4 d
4M 10 0 4-(0 4
4 -1H (04)V 0 U 4.) W 4-
0 t4V (4 O 0 mi 0
0 34 34 0 14 E -H 0
90 M M U U U U Pq 4


4H -r-
U C a U]


C4 H *
UC0 4 0 *C

HOHUI HO
C) m <)> C 14

Sr* Q u I r- 0 0
u O C0 c)
0n u0OgCO H
0 H -O O 4) 0 0 -H E-0
ooaHH3F3lI4s


(0

QM
OCO

Nu
SHo

H*O
tp 0


(U *r *>















4434 4 44 0 490'


rT00
0 C0 CO


000r
co r-. % a
w wcoco~


0000
co co Cco


co Nei Nq Im149
0 ri-0 0 00
co0CO C coco


0
0
0
C c-
0 e
Scoo o--
- cMu co *e
de- I 000-VO N CODONsON\m
*- ioco i'o O C
vI C'j 0-,0 %, d ri -.* G 0
\d( rl -Ieq r('NH 40v oP 1 -
co I I (ANI 0 1 d dp *-
Ir oA q -0 -M d 4r o0 v'Wr4 de O
N *) de rr de N l- rmirHHr-4 VO

HE I PN1 ,i ,0 .I 0I d Q0 I 0
4 O 1 0* L o4 N ZSQ r- ,Q ,
SE4 cm ,Q In m a a

040 04 Q 0 0 QQow
k& 4XA M( AP w
t W :3 :1Pl-) X 0) ED U k N P\hP4 N U
WM~mmm mQXumNWWQ


4 o

H >

0u 0

0 O 0


C 00a 0440J-U


O --.-I UQ I 04. 0 U
S0 0 V E.- ,U ,


40 H4 4 0
0 4004 ODN t J)3

EZ13 9mmmmp


00
*-4
n14 A
a a
*H A r
Uim 00





EHC E-4 0U E2
O OM oU
Um E # #0) 0
o0 o a ^urii
U) )0 ) 50
E-wofS SS


0)
0
*p



r4









14
..4












O"
-I
u40
qM 0

P4
*a




ta



,0
O0






--



04a



0
too




4.M
,p


0 -H
m
S-r4
r-4

.aV


srF
r r















N
coo (n M i 0Doo NW Hn H- Q 0 p-
NyHNHMHHNNCAMqNHOMOHIHO 0


N coV4
HN0
CCO-<

col co -co

00C CO


co On N
OOO
000
Scn co



OC100
H-IOO
C(n co


CD 00 00
IAO4Il.


000
n o r-i
co o co
V)Or
cOOcO


00CD%
00-00
00 0 0


COCN

- CCOD


OOH
C 00 O
,coco
coco tr-


M 00H
CDOCO I
NOcO*


H-10 0i-
o cococo
COOOo


o0 r- oi
,.000
co co
OOO c


O C-i co
HOO

r- co co


HOO
-I O 0
r- 1 o 0
0\4o
tbtb#


+ I
U UI U I I U U+
U U QU OI IQI I U+ UU
UQUamuOQQQ Ua4DU

r, CV r
00 0
co cCD


000
oo co
NCN O
COOCO


H 0 0
coo00o

r-. ,- M
000
D H r-
coD c0


U +
+ + U m
QQOQ(QQMQUU(Q


+ U O
I 4-+ I1 U

O
co 0
(A NI


0
f OO
CO (O
C.- OP>0
- 00
I d O 0


NCV-


\I dp

I

co i r
0Q
r 4 3
n 1


SCO0 CO


m
m+ m


U I I U
+ U Q40 C I U
CQ

0 0)p *h
- OOf00 00 M
Co No M% ON
N\ -- r -*
C> co H fD c
I A dp*o HO 0 co C co'4o\ O p
LtoN co co oko C)r o--a0iM c o- ONi Cor MCOM
H1 C(A ---. ri N ON -W 1- ( o ON. o N. I

N OD I p M n cl c I coo- N0H 4 -
Qco Hn ONH Hr-H.EOcOtn H IW Ln d0'%% H r 3
'p IHI H iu n \dp 0
(dlo AAmme oa o loi N-Q
E0 I N dp ar ade 4) ( 1 13 A 0) A I H M B %D 1 )
OH M 4MQQ MO 0aW r Q ON Q N
A IA U Q0 QA 4 A MO
A 4 04 N1 N A A 44343 4 MA 43 N4 A
Q 0 E-r4 0 0 x n Z U2 i a t
k AA 4a34 A 443 434o
0) 0n 3 h 3 N ( k p k P 03 0 N 0 N w


0
U rH
0
-3 ) )


- 0 .-4 O l O H OH a- U- H
4J 0U01 m M U *4 *r4- H(. 0
r- ~r uO~m r uoa o> CH c-0

H- n H*O > 0 O H ) 0 4-) 0 4( ) r-I 4


SHU40 -0 0 4 -i (0 0 r-4q 3 $4$4 H > CO 'Hr- aC
> CH r 4 )4d4I 004 U o 0 ) o *0 H
C HIOO I OC 1 OH r-(D-l04 Z MId UH04H C
QU 0U CUH<44 M r-4 HH m m0 M UOXOE 0 C
4: I -H 04 4-i M 4 u 0 ( O o Ur -O- aO U O I r-o CI-H ~O
?1U0 0 >X .Md0J OO P MU ) Ui-l <00U0 '-1-4
MH (D 0) HM 0 X54- a)-4 ) >.O M 0 4J Q -OU .O (d m ) CO>-H -~ U
r-Hlr-4E C B403 k04r r- 0 k 0 O MId rH d 9CV40 0-HE-4 U O4--
4 e < ccD~ou u ~ &&< u rH-i3>i-sa s


N%
00
00CO


rn O
COIH

+ co
+m
smM


In
0O4


0 -N


0
0
CV


00


0






0) 4


t H

mc Q)

Q --l



PP 023


Pt
Ca






4u)o 26
4J--4 0
$43








o oooooooooo oooo





Sa+ 3 .
0-00000p000000H0000

+mC 1r ro
mMNw n i t I u a T 0


1+ I H1
oa
000 4

N V
Sa
C0 CCNC CD tv
S o

o at co m- 0 *

I H N O c4m do do- )
H CNNo' de( oO O(* I r M 4.


0 o I 0I \- 4- -


Q I NAn U M WpHU MS -4
*Q raHe .QmO I AA0 Q AA l I i. -
)Oo(S MCh Q 0LO -P -, Z Z t- M % A o
0)0 a I0 o o o .4 0 U
ME-4a 0 0- m Ca QQ o.4 0

V I Q W iC Q I AU CQ A g C V


a (P 0 0
04 M 4 P ) .
P rl 010




*0 -4 0 0
HM 9H

M rr- C) .H 4 *4 U U.i l 4 4
u 0 -- H A4 V+ 4 $4 14 C
a)-4 U>M 0Q) -4U 4k-H ( *M4 4 0 d
O 4 O W- 0)0 U40V>000 k 0 00 V
X: 0) 0 M 0 P4J VE-4U 0Cp 9 aUV VOO C0o 0 U
0CWm H0UOR H 00 .0 r O 0
0 AC U W B (0 U 0M0 0 0 0 4 0 *3 tC 0
+VH -H k V N(UA0 ( 4V4VPV4J+ W 44VJ 9
k 00 B-H0 # O X X XX 0UO -r Ur-4 0 0 3
0 0 ( MA04 VJ W ) W0 (D )A-i-H CA 0 !0 0
ZZ04.E4E z-r+ U)W















CHAPTER 4
RESULTS ON FINANCIAL MARKETS AND BANKRUPTCY



For our research on risk premia and bankruptcy, we

computed and drew risk premium curves for each of our 50

companies' bonds. (These curves are presented in the

Appendix.) A solid line indicates a bond rating change by

S&P, a dotted line one by Moody's. The arrows on the graphs

indicate a bond upgrade.

In 80% of our sample, or 40 of the risk premium curves,

there is a definite upward trend: risk premia rise steadily

as the company approaches bankruptcy. Nine (18%) have

ambiguous, fluctuating trends, and only one, or 2% of our

sample, shows a downward trend where the risk premium on its

bond actually fell as it moved towards bankruptcy. Hence, our

data suggest that the market generally incorporates the

likelihood of bankruptcy adequately in its overall assessment

of risk and includes this particular risk in the valuation of

securities.

When we compare the relationship between bond rating

changes and risk premia curves, we find that for most cases

where there was an obviously increasing risk premium, the





28

rating downgrade came after the risk premium had begun to

increase. Hence, in the short-term, a lag of one to seven

months does exist. This conclusion confirms that of Weinstein

(1977), who found that markets had anticipated bond rating

changes 18 to six months before the change. It, however,

conflicts with Katz (1974), who concluded that yield-to-

maturity adjusted to the rating change six-to-ten weeks after

it happened. We believe the reason for this discrepancy is

that for measuring market-assessed risk, our risk premium, a

relative value, is better than the yield-to-maturity, an

absolute value, used by Katz. This is because while factors

such as added risk and inflation would eat away the nominal

returns of yield-to-maturity and bias the results of any study

based on it, risk premium would account for such factors and

hence provide an accurate indication of the returns on the

bond.

Our data, then, seem to suggest that bond ratings, though

popular since the early 1900's, cannot predict risk premium

changes or any upcoming default. We believe, however, that

the lag phenomenon is not particularly meaningful. For

example, a bond downgrade that trails an increase in the risk

premium may lead a later increase. Since bond rating changes

should warn investors of coming bankruptcy and default, what

really matters in the long-run is whether a bond rating

provides new information to the market on companies that later

went bankrupt. That can be determined only by seeing if bond






29

rating changes have any impact on the market. Assuming that

the first bond downgrade provides the market with the greatest

amount of new information about changes in risk, we must

further study the impact of the first bond downgrade on the

market.

Our data can be used to compare the two rating agencies,

S&P and Moody's, to see which warns investors earlier. Table

3 gives the relevant information. On average, S&P gave the

first bond downgrade 24.54 months before a company filed for

Chapter 11, while Moody's gave the first consistent bond

downgrade 21.58 months, or 2.96 months after S&P, before a

company filed for Chapter 11. Out of a sample of 50

companies, there were five bonds that received no downgrades

whatsoever from S&P's and six that received no downgrades

whatsoever from Moody's before the companies which issued them

filed for Chapter 11 bankruptcy. Although there was no rating

change, the markets did foresee the coming bankruptcy, and the

risk premia on these bonds began increasing an average of

three to five months before Chapter 11. Finally, six

companies which S&P's rated were not rated at all by Moody's.

Hence, it would seem that in this specific instance S&P

provided more timely and more complete bond ratings and

changes than Moody's in the sample and for the period we

studied. We emphasize that this is a one-sided test only, and

that it by no means can be thought of as a definitive

conclusion.





30

As described earlier in Chapter 3, we divided the sample

of companies which filed Chapter 11 into those that later re-

organized and those that were liquidated. Nine out of the 50

companies in our sample were "liquidated," while the other 41

re-organized. The average risk premium curves for these two

groups are shown in Figure 1. From this graph, we can roughly

conclude some interesting patterns. Long before Chapter 11,

the companies that would eventually be liquidated had lower

risk premia than those which would later re-organize. The two

groups had about the same risk premia 41 to 27 months before

Chapter 11. Then, as the companies approached bankruptcy,

risk premia of companies which were later liquidated rose

steadily above those of later re-organized companies.

After performing t-tests with Montgomery's formula for

the three periods (65-to-42 months before bankruptcy, 41-to-

27 months before bankruptcy, and 26-to-i months before

bankruptcy), however, we found that this was not so. We had

tested for the null hypothesis that the means between the two

groups during each month were equal. During the first period

of 65-to-42 months before bankruptcy, 20 out of 24 (83%) t-

tests were significant, and all t values were negative. By

our definition for t, this means that the mean values of risk

premia for the later liquidated firms are lower than those for

later reorganized firms. During the second period, however,

12 out of 15 (80%) t-tests were significant, and 12 of those

15 t-tests were positive. This means, contrary to what might






31

be discerned from the graph, that the mean values of risk

premia for later liquidated and later reorganized firms were

different and that the mean risk premia of the later

liquidated firms were higher. Hence, although from just

looking at a graph we had thought that the differences, which

were clearly visible, would not be statistically significant,

the t-tests show that this is not so. Finally, during the

third period, 19 out of 26 (73%) t-tests are significant, and

only one of the 26 was negative. All other t-tests during

this final period were positive. This means that during the

final period, the mean risk premia of later liquidated firms

were also significantly greater than those of the later

reorganized firms, as we had expected from looking at the

graph. Table 4 gives the means and the t-tests.

This means that the market can sense the coming of

liquidation and therefore place an added risk premium to such

risk much earlier than the 27 months we had earlier expected.

For our event study of the effects of the bond downgrade

on the daily rates of return, we first had to find the exact

press release dates of the bond rating changes. We selected

only those companies that were listed on the New York or the

American Stock Exchanges.1 Because we could not find relevant

data for two continuous years from the COMPUSTAT Daily Return

tape for all our companies, some of the observations in our

sample had to be dropped. This left us a samples of 22 pair

of bankrupt and matching companies. The matching samples have






32

equivalent downgrades during the same time period, but did not

later file Chapter 11. The companies in the two samples are

listed in Tables 5 and 6.

Tables 7 through 10 give the statistical results of our

event study. In Table 7, the t-statistic for testing the null

hypothesis that the average disturbances at event day, E,(0),

equals zero for the Chapter 11 group is -21.81, with 499

degree of freedom. This t-statistic is significant at the

0.5% level. Thus, as expected in the Research Design and

Methodology section, the null hypothesis that the average

disturbance at the event day for the Chapter 11 group is zero

can be strongly rejected. For the matching sample, however,

the t-statistic for testing E2(0) equals zero is -1.45, which

is significant only at a marginal level. The negative signs

on both the t-statistics mean that bond downgrades negatively

affect shareholder wealth.

To see if the means in the Chapter 11 and matching

samples are equal, we employ a special t-test (see

Montgomery(1984)), which is suitable for cases when we cannot

assume equal variances and when the number of observations is

less than 30. The t-statistic for testing the equality of two

means is -15.545, with 42 degree of freedom. It is

significant at the 0.5% level, so the null hypothesis that the

two means are equal is rejected. This means that the daily

rates of return of companies in the Chapter 11 sample suffered

more with bond downgrades than those in the matching sample.






33

The variances of the two groups, as given by the F-test, were

not significantly different.2

Table 8 shows the statistical analysis of the two sub-

groups in the Chapter 11 sample, the companies which re-

organized and those that were liquidated. The t-statistics

are -4.98 for companies which later re-organized and -26.94

for companies which did not. Since both are significant, the

null hypotheses are rejected. Using Montgomery's formula, we

find that the t-test is -0.32 with 9 degree of freedom, which

is not significant. Hence, the two means are not

statistically significantly different. Finally, the variance

of the re-organized companies is lower than that of the

companies which were later liquidated.3

Tables 9 and 10 give the daily prediction error and

cumulative daily prediction errors in the 21 days surrounding

the downgrade announcement. Figure 2 shows the cumulative

error curves for both the Chapter 11 and the matching samples.

These curves show that the rating downgrades had a definite

negative impact on the daily rates of return. Most

importantly, the relationship between these two curves tells

us that the impact of the first bond downgrade is much more

severe on companies which would later file for Chapter 11 than

those that would not. Hence, we may conclude that the market

does react to downward rating changes and that the rating

agencies therefore do provide significant new information to

the market with rating changes. These results conflict with






34

those of previous works mentioned in Chapter 2. We believe,

however, that the issue is still very much unresolved, as can

be seen by the seemingly contradictory results of Griffin and

Sanvincente (1982). Therefore, our work should contribute to

the ongoing debate.












Notes

1. Our two criteria for selecting companies into the Chapter
11 and matching samples for event study were:

1. The companies had to be listed either on the New York
or the American Stock Exchanges.
2. There must be two years of continuous data available
for the companies.

2. The F-test value is 1.13, with degree of freedom (21, 20).
This F-value is not significant, so the variances of the two
groups are not significantly different.

3. The F-test value is 8.13 with a positive sign, and it is
significant at the 1% level.












* In H N %u
0Q in I OIN


N(M NN H-


N 0 O N r- N -


MIN MNHIVl NN
.-oCoCrN CO0O'
cO(0N.4'.4~NNN


Io N-
PI 0HII0 0 C HO


00 OMMo0o o 0


0i
o -


O
Co C


OCOr- r-l 4o
0oo00 -4 0
rN w D\ io 0
cwcocoZcwwcow


co cn 1- M
NrH#N-()HM-4 H NN


0 O C-4

CO OD CD
cWc(X


'0 0 9-4 -V r4' CO

0000 00
0


OOHO




N



SOOO N
0 0 0 r-1


0cOc' IA 0lcr-.4UNN
N c o0 NNNIA \ O O IA N '.0
Cowowzzzwoz!r%-rlcwwO


0o0
0oZco


OeN co IA.-i 0%%DO I In
.NNOm0HmOlNHOINa-i


HOOHOOO
NN 00r-0004w
CO COV?0C On
C~eO ~0FCOCfC[


OHO
C. 0o0,
Co,.


000
o0 O N
CD00 00
OLANV
cCOCOQ


N% H M O
00000
in %O0 H-i r
o00 00 c 00


co 0 0
ON CIO
0 -N N CO
N 0 0 0 do
0 o d0 o 00 o co 4-
o N coN 04Mo aNO n M (A M
N \,- eo. -- 0 -*
Nr oo co -D i-I 0 c
IdONI e 004 COdP O mCO w' dP 4
dp V r-4 NNI- Coco o dp- -Co' 0 ON 00 cn o
oo\ o r-* H n \\-C (M o% \0a .o o- \mcnol
N r-. dp pdp nr \d. n do C4 n- 0 or'~0
*- I A oQ o IJ OD o I omde- 1d I (I r- 0
0 CO 00o \01- O ,-'N 04)dd oe I -\-dPCO N OdP- N 0
Hl dp ri- C- rH t-I in 0 CoD n 0 HH tU 0 0 uODo M r-l %r Ln 'd--,OrV -,HN HA m
In I \d N.\N-II HI Ain ln M d o -
(D 0 Hl On Q r <-r t I Mdp in <4-) CA I )IM 0 4) \0 4l
Q A IA IA 0 QW QA .Q A
44 A0NCND) A oH()O oc N44 AI 4 O A4 0 (i)IQ 0)NA(P )44 A
Q 0 0 E-QU 0 0 WZ CW
m 4M EQ w a np E U om oIn& lc M En to V) ril ')m aU)>
c~acc~ccucuQW UociiccucCCQ 2aoc camQC 2


0
4.

o0
S 0 H
H 00

0H 0 4 -. l
4J r.U CH 0 10 tm
C H ( 0 4 0 0 .
OU ) U0 aQCr U H 44
H. I N M w 1-4I 4M M


r#i*i e e CHi(
iHi<<;

M *
C* *0 0
O~0 lH S H40
S- H -4 r 10
04 OHOO P M 1 S rL-H 1
k *U *V. r a- 0
00 *HC2 'CO Ht* C D *
U0 A *-rM -4 o -4 T 0 C o0
o V'-4 VIAU HC *- 3 4HW
0 lOH .4 H 4or-I
Q5 HHrC0 04 UIOds p) 0 H C


M -,>, ou o a= o # 30u
0OH $3o-r( Z (0 UO04H C
m4 r-4 H 9 0 HU 000 X H 4 C4 0

O n noOO-( C O C (D OO H O
k Hmmo mi5O4 04 0.0 > g 00 )4
.40 0 4 r 0 l H r-l 0 9 4 0 0 4 E- 0 0 -H l-4
UUUw44h u ZHH h 0 zSzzz













O
0
C-I
10


N0 C
H-Io 0 r-
H-CoH I I Nr- 1 o


0 NN m in OHNOW r
HNHVmcOHW~Hl


(mN
00
COCO


co OI CO
000
N t- 0
C0 CO CO


0 r.-I 0(
0 H0 0
CO 00 CO


NM Oc
in in N
co co00O


N c NM 1 LO in M N
ON linON(NMH


rNI
00
coc


000
NNCO O


000


000
C0 O


0
0
0
oN <
0C
%0 do CN 0
oo co
- N %O- o
d o1 o N -
q NN 0 No OmdH
0i rH c'. LA H N m) c 0 *pP
n- -I o \c0
I I d Nde-
N d0 ) 0l H d N I N
H, 0 o-1 ~ed I0
H \NaacmcHHto
u *Ea-Hi *u) im
Uoo2 I ##QQcninin.i\
Q Qc I 0 n
*4 3 4 o H I N
CfEl -CQ -
ki C'u) ( )O 0 r44)rPk X 0) :
V) X .-IoM to twNOm w
NHWWWE MQWW


54
0
U
$4 >1
cd Z
U 0

c O-H



r C a u M C
0O H a
S-4H U 0 .C
4. 4+ r-4 4 P 4) &-4
MC000
4 -P A 0 -H 0
0 0) C 0 439 4
0Pa4 mA im


U)
0
-,-4
t H
,-4


SCd --I
U (0 9 4-
SH r
0d00

wOO

3 0) Q) (1


tI >- -.4



0 v
o Mo O M M
cv trN C O









S00l ~~






0,1-40ON.000 4
4) AP 0 "j
0ooOCD00 0 Mt
MOcvH Or 0
S4JU
no o O M-oM


-- a

,00^
'.r i '-ooo N ai
Or-100 0 0Aa) A








4)O C C









u %D 4 --% r 0 0





0>-, >
*H- 4 0 >1
dd V Vd ()
w% 0 0C C





4 04f Z *> 0 0r
(0IP P O V








o >O.Q44.
a rv (a











34 00 k
W 0 O) C0
>0 0 i 0 c

au 3 :O O +3 c 1 u0
* e O .Cd C
0000 0 M m
## # a, a A ,
nOd -rOH o 4 (
00J00M-* 0 000
MU^ ic g^=




















17

16

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1
65636159575553


o Liquidated Firms


514947454341393735333129272523211917151311 9 7 5 3 1


Months before Bankruptcy (Chapter 11)
+ Reorganized Firms


Figure 1. Risk Premia Curves Comparison

Risk premia trends for companies that filed for Chapter 11
and later re-organized versus those that filed and were later
liquidated.









Table 4. T-Test of Risk Premia Trends


Time MeanI Mean2


16.70
11.46
10.38
6.08
8.31
7.12
6.80
6.80
6.86
5.98
7.42
7.33
6.73
6.19
6.14
6.42
4.73
5.10
5.04
4.88
5.74
4.49
4.23
4.32
4.32
4.46
4.54
3.99
4.04
3.95
3.33
3.50
3.80
3.80
3.83
3.82
3.44
3.51
3.70
3.50
3.56
3.32
2.56
1.87
1.72
1.89
1.71
2.08


11.18
8.08
7.43
7.11
6.22
5.68
5.55
4.89
4.78
4.65
4.47
4.41
4.45
4.42
4.23
4.00
4.04
3.81
3.74
3.66
3.49
3.73
3.49
3.58
3.58
3.57
3.77
3.75
3.49
3.48
3.53
3.34
3.34
3.30
3.29
3.25
3.72
3.53
3.39
3.49
3.46
3.45
3.49
3.20
3.26
3.12
2.89
3.08


Diff.


5.51858
3.37600
2.94731
-1.02914
2.08476
1.43333
1.25468
1.90249
2.08516
1.32662
2.94818
2.92675
2.28420
1.77084
1.90900
2.42500
0.68789
1.28795
1.30813
1.22106
2.25051
0.76234
0.74654
0.74654
0.74759
0.88553
0.77331
0.24753
0.55032
0.46810
-0.19686
0.16256
0.45914
0.50408
0.53727
0.56812
-0.28775
-0.01832
0.30900
0.01545
0.10248
-0.12621
-0.93164
-1.32963
-1.53593
-1.22093
-1.17630
-0.99972


t-test


5.969257
3.651703
3.188007
-1.113180
2.255015
1.550387
1.357147
2.059800
2.255444
1.434954
3.188950
3.165763
2.470741
1.915457
2.064899
2.623039
0.744070
1.393130
1.414961
1.320781
2.434302
0.824601
0.807508
0.807508
0.808646
0.957843
7.588362
2.428966
5.400186
4.593351
-1.931730
1.595207
4.505502
4.946421
5.272179
5.574928
-2.823640
-0.179790
3.032172
0.151591
1.005648
-0.897810
-6.627130
-9.458170
-10.925600
-8.684910
-8.367450
-7.111410


s.l.

0.5%
0.5%
0.5%

2.5%
10%
10%
5%
2.5%
10%
0.5%
0.5%
2.5%
5%
5%
2.5%

10%
10%
10%
2.5%





0.5%
2.5%
0.5%
0.5%
5%
10%
0.5%
0.5%
0.5%
0.5%
1%

1%



0.5%
0.5%
0.5%
0.5%
0.5%
0.5%








Table 4--continued.


1.68
1.97
1.74
1.80
1.84
1.84
1.80
1.74
1.88
2.08
2.49
2.00
2.45
1.83
2.45
2.49
2.72


2.79
2.52
2.79
2.68
2.89
2.73
2.93
2.86
2.76
2.77
2.50
2.57
2.62
2.73
3.04
2.91
2.87


"MeanI and Mean." are the risk premia means for the later
liquidated and later reorganized firms, respectively. "Diff."
is their difference. "t-test" is the value of the t-test, and
"s.l." is the significance level.


-1.10180
-0.55000
-1.04960
-0.87340
-1.04792
-0.88565
-10.14058
-1.11917
-0.87986
-0.69457
-0.10409
-0.57091
-0.17650
-0.89842
-0.58684
-0.42263
-0.14889


-7.837530
-3.912360
-7.466210
-6.212830
-7.454230
-6.299980
-7.996630
-7.961060
-6.258740
-4.940710
-0.100230
-4.061090
-1.255510
-6.390810
-4.174430
-3.006340
-1.059100


0.5%
0.5%
0.5%
0.5%
0.5%
0.5%
0.5%
0.5%
0.5%
0.5%

0.5%

0.5%
0.5%
1%






a)0*

0 IA a) 41
(I
.r4 $4 I


90 44 rO44 Cd V I A 0 ) 0
ro >o

ooowo ooomo N ooN0 W
P N 0 w o0 w H o0 O0 N N N H N N N 0O N O r4 -4 N
4J oooooooooooooooo 4

> omon In o i N iiio r m ."u




C (U rq 0 0)
m+ 0+ m mOl+UO CM .M 0-



H HmUUM UAUUMa 04 0
0 Y-
. c +r- c O Lo IV o T3 I + % c. m u


o UH 9H 4
NI + 4
mo o mmmm o so r- o o o+++






r.), 0 0om E-
0 0 COP 03 0 0 o o

a 0 CI m emem aCoD mCm%3 o



S0 r- N
v de %Do -;r1, iN cO 0 oko CA %0 oiCho I do 0 0
Si r> 0 0 MC o iia N r r ( 0 H O


N r \ mM


O0- oOm- -o O- N -- NH :=
NH- NN-ro He (- 'gd0
SN -CD O % ~N P~ N o W.
00 0HAv AdH JAHn D1-4 A A 0 HN 'q -4CM (AoA o u I(


r- oNI QM -wQ in H ON zz0 in g 4) to

4 .1. .AAO ,4 .Q. OQ 0 4.



0 4 *

C U U H0
SH. k H* V
4>1- 0 I1 C O .4 0
H <04 U 4 #) Va)*
S-a) 4 p r I*to-
o >4 -OHH *0o 00 0'dO


4 CO CO -O4 I4) 0 -(l l4 (

4) 4-a H-4OCU40) 01 V ( 0 rO.)
0 IC 4- HIUC*IO. La)OU U OH p 01.-i 0 H
0 or H.a) .4 4 4 r-4 4) (d 00I 000 g
W (O M01 0 )4 4 A P4 iZ .0 $4














> 44 4) OO U 9 r. 0)opr.A uV4J 0 V
9 a1 r. 44 4 U U o ^ 1 1 o a "





.01 i Q .-r a) r. rQ U >1 00 r- M 44
0 4 P r4 4) r- V 0P4 E-4 4) 0 rn r- 0 ) tP'o





V 0) 4-)




(0 V W C D 0 V %0 r, -0C% 0 i t to.to
r- H H H C N N N N N N m m m m m m 0 -H4P=
c CM^ uca-^fwitin om- oh Io





U1^'--~-tIM ~ M N M ~ n Q -HNSd o








w 00 0a 0 00 O 0 0
ewo00NoA Con0C v-4o


W) Uw Cj
000
N CN r-I
o CO co


0% r. cn
000
ocoo
o\ m CM


0NN
Ln r-i
CO00WD
HlPo
el a1m


I 4+ +
+ CQOUU00 mIU
4mAm~uommUU~mMEU


m co r DLn o
000000
L r- Ln o r- Ln
00 0O 00 00 00 CO



U00 II
UU CQCp


+ I I
Ir I CQ ll+++ C
i

Co


4 O NI 0 in
0 0 Nr-i o n N o

co Lo n On o oC


in i co n N oH *



DA A C 1 AO A a)
p *N n H4 N A n in
4n \) n Q l n in In In a


W0 P l l W) rw p 0 0

O n r-i 0
AJQK-4J4Ar W z4
mawmsQ~a0
(U Q U QUQUQ )Z
NCA J^ -


ON N CD r-l
M ooCN
Or I C-l
scOOH


m a
m 0
V L C
r DO -4
4n H rU u


E 0 (a c to

Q)U O MU C U
in ci m m M< 0 o 0 ti o 0

HP $4 r4 )W )
Q 0 0 r. 9 V N

00) U n4 t u C A 0

O mO 0m-HQuO
CQMSfflKHhUP


00
00
01 CM


CO -
n H-l
r- I I dp

1-I


0 0
U) 0)


0 p p 3
zUM W E


coo C
0n o o

0A r-i


an
- C


-
0 dp
M NM co CM n dp



r- -4J

(U il 4) a) t 4)


S) U3
QU) U} JM

r 4 k 3 0 p
www)wwwU)U)UO


ONO
00co

co Lo rN
O~crlA.


to


14


H O

OH O
H 0O

as r

m 00
*0 a) m


i0 Ln
rlA
NNN2C
a\I


r-I



O C0
0 0
a mr-
VCOH
I o 0


0utU
0000


o o
04 01
00




0 0o
H4)

CO


rHNmo0NmVoowrmoH -I
M N-iHHr- NNNNNNMMMM mmv


mow
O0O
cococ


EP


0
*4
U


ii-'
41


4)




0
*H
























m
*M










































E-1
-14
o
*H








a
0
(0



















0E






)U



-4
UC

0
a-












4


:0
0
Cu


r-c C

0'0


zNH o
NMM Nn

Ln r q'r










Table 7. T-test for Chapter 11 (bankrupt)
and Matching Samples: Event Study Results


Chapter 11 Sample


Matching Sample


E1(0)
Std.


-0.03199700
0.00687700


H0: E,(0) =,0,
tI = -21.98
(d.f. = 499,
significant at 0.5%)


E2(0)
Std.


-0.00203900
0.00645500


Ho: E2(0) = 0,
tg = -1.45
(d.f. = 499,
significant at 10%)


HO: E1(0) = E2(0)
t-test = -15.545 (d.f.= 42)
(significant at 0.5%)









Table 8. T-test for Sub-Groups of Chapter 11
Companies: Those which were Liquidated and those which
Reorganized


Liquidated Group

E,(0) -0.03293800
Std. 0.01980400

Ho: E,(0) = 0,
t, = -4.98
(d.f.: 499,
significant at 0.5%)


Reorganized Group

E2(0) -0.02922600
Std. 0.00694700

H0: E2(0) = 0,
t2 = -26.94
(d.f.: 499,
significant at 0.5%)


Ho: E,(0) = E2(0)
t-test = -0.32 (d.f.: 9)









Table 9. Disturbance Resulting from Bond Downgrade on
Chapter 11 (Bankrupt Sample) and Statistical Analysis


Event
Day


CE
m%


avg E(t)
-Ml}

-0.006464
-0.007603
-0.001908
-0.001019
-0.016982
0.001227
-0.002338
-0.005651
-0.007840
-0.009647
-0.031997
0.018805
0.022623
0.002346
0.002065
-0.008776
-0.034234
0.014345
-0.029843
0.021802
-0.008666


Variable

E(-l)
E(0)
CE(-10,0)
CE(-5,0)
CE(-3,0)
CE(-1,0)
CE(+1,0)
CE(+3,0)
CE(+5,0)
CE(+10,0)
CE(-10,10)


Value(%)

-0.009674
-0.031997
-0.090222
-0.056246
-0.055135
-0.041644
-0.013192
0.011777
0.005066
-0.031530
-0.089755


t-statistic


-1.41
-4.65
-3.38
-2.11
-2.06
-1.56
-0.49
0.44
0.19
-1.18
-3.36


10.0%
1.0%
0.5%
2.5%
2.5%
10.0%


* 0.5%


The t-test for CE is:
t = CE*(N/2)/(Std(CE)* (T12))


-0.006464
-0.014067
-0.015975
-0.016994
-0.033976
-0.032749
-0.035087
-0.040738
-0.048578
-0.058225
-0.090222
-0.071417
-0.048794
-0.046448
-0.044383
-0.053159
-0.087393
-0.073048
-0.102891
-0.081089
-0.089755


-10
- 9
- 8
- 7
- 6
- 5
- 4
- 3
- 2
- 1
0
1
2
3
4
5
6
7
8
9
10








Table 10. Disturbance Resulting from Bond Downgrade on
Matching Sample and Statistical Analysis


Event
Day


CE
IAm


avg E(t)
Xll

-0.012418
-0.022204
0.018645
0.003892
0.010662
0.004340
0.013230
0.005412
-0.010702
0.002165
-0.002039
0.007029
-0.000817
-0.019285
0.001402
-0.003376
-0.000191
0.001111
0.005467
0.004831
-0.004599


Variable

E(-1)
E(0)
CE(-10,0)
CE(-5,0)
CE(-3,0)
CE(-1,0)
CE(+1, 0)
CE(+3, 0)
CE(+5, 0)
CE(+10,0)
CE(-10,10)


Value (%)


0.002165
-0.002039
0.010983
0.012406
-0.005164
0.000126
0.004990
-0.015112
-0.017086
-0.010467
0.002555


t-statistic


0.34
-0.32
1.13
1.28
-0.53
0.01
0.51
-1.56
-1.76
-1.08
0.26


* 10.0%
* 5.0%


The t-test for CE is:
t = CE*(N"1)/(Std(CE)* (T12))


-0.012418
-0.034622
-0.015977
-0.012085
-0.001423
0.002917
0.016147
0.021559
0.010857
0.013022
0.010983
0.018012
0.017195
-0.002090
-0.000688
-0.004064
-0.004255
-0.003144
0.002323
0.007154
0.002555


-10
- 9
- 8
- 7
- 6
- 5
- 4
- 3
- 2
- 1
0
1
2
3
4
5
6
7
8
9
10






















0.03

0.02

0.01

0

-0.01

-0.02

-0.03

-0.04

-0.05

-0.06 -

-0.07 -

-0.08 -

-0.09

-0.1

-0.11 I
-10 -9 -8 -7 -6


0 Chapter 11 Sample


-5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10


Day Surrounding Event
+ Matching Somple


Figure 2. Disturbance Due to Bond Downgrade

This graph compares the disturbance caused by a bond downgrade
on companies in the sample which filed for Chapter 11 and in
the sample which did not file for bankruptcy.














CHAPTER 5
EMPIRICAL MODEL OF BANKRUPTCY


The evidence that bond rating downgrades provide new

information to the market suggests that proper analysis of

fundamental corporate factors, such as the ones presumably

used by S&P's and Moody's, can give an early warning of

impending financial difficulties. Even so, there are at least

four problems with using bond downgrades as the sole or

primary indicator of bankruptcy risk. First, many firms whose

bonds are downgraded never experience significant financial

stress. In fact, a company may simply be trying to modify its

risk-return balance and better reposition itself in the market

place to capitalize on future opportunities. Second, although

our event study identified a differential effect between the

companies that eventually filed for protection under Chapter

11 of the bankruptcy code and those which did not have to seek

protection, that does not mean we can easily tell between the

two before the fact. In other words, how much of a difference

is "significant?" Third, S&P's and Moody's sometimes give

conflicting signals, and there have been cases where the

downgrade either does not predate the bankruptcy filing by






49

very much time or there is no downgrade at all. Finally, it

appears that consistent signals start appearing only

approximately two years ahead of the Chapter 11 filing,

assuming, of course, that they can be correctly interpreted.

Thus, bond downgrades give us about as much forewarning as the

ZETA model of Altman et al., and the latter model is probably

more a practical predictor of bankruptcy than bond downgrades.

Hence, we believe that proper analysis of fundamental

corporate characteristics would provide us with a more

effective empirical model which could determine bankruptcy

risk. What we would need is an indicator which can alert

investors and managers of developing circumstances that would

normally lead to increased probabilities of bankruptcy long

before it actually becomes necessary to file for protection

under Chapter 11. Such an indicator would allow investors to

make more informed portfolio allocation decisions, and it

would signal to corporate managers that corrective actions

need to be taken before financial difficulties become serious.

We believe that if such an indicator can be found and used to

predict financial distress and possible bankruptcy up to five

years ahead of potential bankruptcy, then the scarce resources

in the economy can be better allocated.

Current bankruptcy prediction models, however, have not

shown that they can reliably forecast bankruptcy risk more

than two years ahead of filing of Chapter 11. In fact, there

is very little evidence to date that it is even possible to





50

detect the roots of financial distress up to five years ahead

of an actual crisis. Our main purpose in this part of the

dissertation, then, is to explore this issue: can a bankruptcy

prediction model be designed that maintains significant

forecasting powers up to five years ahead of the date of

Chapter 11 filing? While we would also like to develop a

model that can surpass the predictive powers of the ZETA model

in the short-run while also preserving that power over five

years, we do not, in this dissertation, intend to search

specifically for such a "best" model. That task will be left

for future research after we have shown that it is possible

to maintain high levels of accuracy over the long-run.

Present-day finance literature does not provide us with

many works that have looked at the relationship between

various types of risk, such as bankruptcy risk, and companies'

fundamental characteristics. Zavgren (1985) provides some

evidence that asset efficiency is an important indicator which

may be of use to us, but we believe that other variables would

also be needed. Derkinderen and Crum (1988) developed a

framework known as the Potential and Resilience Evaluation

(PARE) model on the issues of long-term risk-return balance.

This framework suggests some possible variables which can be

used to assess the fundamental and strategically crucial

characteristics of a company. Basing our views from such a

fundamental perspective, as suggested by Derkinderen and Crum

(1988) and Zavgren (1985), we have arrived at a list of eight






51

variables which can assess both the short- and long-run

dimensions of the bankruptcy problem. We now turn to a

discussion of these variables and the roles they would play

in our study of long-term bankruptcy risk.


Variable Definitions


We selected our eight variables out of many potential

candidates. In selecting them, we aimed to find the minimum

number of factors that together provide signals about both

short- and long-run aspects of the companies, in isolation as

well as relative to other companies in their core industry.

Our reasons for selecting these variables were also partially

based on the availability of data, although these variables

are considered to be good indicators of differences between

the more successful companies and those with lackluster

performance records. These variables can roughly be divided

into two groups: 1). those which assess company operating and

profitability characteristics (which will be referred to as

"the Group I variables"), and 2). those which assess

fundamental company characteristics (which will be referred

to as "the Group II variables.") Existing bankruptcy

prediction models focus mostly on relationships that would be

included in the first group. Although these factors are

obviously important to bankruptcy, particularly in the short-

run, we contend that the second group must also be considered

if the model is to have adequate long-term predictive powers.





52

Four variables are included in Group I to measure the

company's financial and operating positions. The first is

return on assets, or ROA, defined as EBIT/TA. This is the

basic earning power ratio and is a strong signal of

profitability. We consider that it is a particularly good

representation of the company's performance in implementing

growth in the past. The second variable in this category is

the fixed charge coverage ratio, FCC. This ratio measures

simultaneously a company's level of debt and how well its cash

flow covers the servicing requirements of debt. It is one of

the most important indicators of the ability of the firm to

survive adversity in the short-run. The third variable is the

market-to-book ratio. The market-to-book ratio measures

investors' confidence in the firm and, as a direct result, how

well the company can tap into the equity market for capital.

In other words, this ratio indicates the extent to which

investors believe that the firm has good growth opportunities

for the future. A ratio greater than one indicates that the

return from reinvested earnings is expected to exceed the

required rate of return. Finally, the balance ratio of

income and sales, defined as the difference between the growth

rates of income and sales, focuses on the profitability of

incremental sales. It measures whether the company is

boosting sales at the expense of profits and, therefore, may

be heading into financial problems even as it continues to

expand market share.






53

To assess the second category of difference indicators,

the Group II variables which measure company fundamental

characteristics, four additional variables are used. First,

ownership assesses the agency effects of the company's

management. According to the agency theory literature, such

as Jensen and Meckling (1976), when the shareholders contract

with the management for the latter to serve the former as

agents, some cost is inevitably involved, including a dead-

weight loss. This cost occurs because management and

shareholders in most corporations (particularly the larger

ones) are usually distinctly separate groups, and they likely

have differing, even conflicting, interests. For example,

management of America's top 200 firms own only 0.1 percent of

their companies' stock, and nearly one out of ten Fortune 500

Chief Executive Officers own no stock at all in their

companies. Because of such low ownership stakes, many

scholars have questioned whether such managements really serve

the shareholders' interests. As a result of questions such

as this, management ownership has become a topic of interest

and discussion in recent years. We agree that this topic is

important and relevant to our needs in this dissertation. We

believe, however, that absolute ownership measures are not as

significant as the net transfer of ownership, which measures

whether management, as a whole, increased or decreased its

ownership share of the corporation during a period of time.

Assuming that management has information the average





54

shareholder does not, then whether a management is a net

purchaser or a net seller of the company's stock shows

management's confidence in its own projects. Further, it also

shows whether management is committed to the success of the

company or whether it is simply "grabbing for parachutes."

To incorporate these ideas into the model, we include the net

rate of management stock acquisitions as a variable, defined

as the difference between management purchases and sales of

the company's stock, measured in percentages of total equity.

The second variable in this category is the capital

intensity ratio, defined as total assets divided by sales.

This ratio measures how much a company must invest in assets

to expand its sales. Thus, it indirectly tells us both the

structure of the industry and how much the company will have

to rely on fresh capital to fuel growth. In this

dissertation, though, we do not follow the traditional

definitions of industry as given by COMPUSTAT or Dun &

Bradstreet. Rather, developing from the ideas of Tse (1987),

we regroup the COMPUSTAT SIC codes so as to define 24 lines

of industry by the nature of the business. These definitions

of industry, along with the growth rate and bankruptcy rate

of each industry, are presented in Table 11. As Table 11 and

Figure 3 show, a company's line of business does impact

significantly on bankruptcy. While the specialty

manufacturing industry (code #19) had a 13% cumulative

bankruptcy rate during the 1968-1987 period, the chemicals






55

(code #4) and utilities (code #20) industries suffered only

about a 1% cumulative bankruptcy rate during the same period.

Expanding on the information about competitiveness

contained in the industry characteristics, and to assess a

company's position within its industry, we employ the relative

sales growth ratio. This ratio is the difference between the

sales growth rate for the primary or core industry in which

the company competes and the rate of growth of the company's

sales. It indicates the competitiveness of the company within

its major industry in terms of its ability to gain market

share. It also complements the balance ratio of income and

sales, the Group I variable that measures whether the company

has gained market share at the expense of profit margins.

The final variable in the Group II variables set is the

relatedness ratio, as described by Rumelt (1982). This

variable assesses a company's diversification program in that

it indicates the extent to which the company focuses its

efforts on a portfolio of related businesses that could be

expected to have synergistic interdependencies. It is

calculated as:

(% of assets in related segments)
-----------------------------x (number of segments 1)
(number of segments)

and determines whether a company is a "single business" (a

value of zero), "unrelated diversified" (a low positive

value), or "related diversified" (a high positive value). As

Rumelt (1980) points out, different strategies of





56

diversification can strongly affect a company's long-term

prospects, and the degree of relatedness should assess the

benefits of those diversifications. According to Rumelt, the

higher the relatedness ratio the greater the chances of good

performance.

The eight variables described above cover the two broad

categories and also collectively address the various

dimensions of the PARE framework, which are: 1) the extent to

which the firm has good growth opportunities available; 2)

whether or not the market perceives that the company can

exploit the growth opportunities successfully; 3) the degree

to which the fortunes of the firm are subject to foreseeable

adversities in the future; 4) whether or not the firm has the

ability to survive such adversities.

Role of Variables in the Model

The model of bankruptcy risk developed in this

dissertation includes the eight variables discussed above and

uses them to derive a summed probability of bankruptcy. Using

Logist Analysis, which will be discussed in Chapter 6, a

probabilistic function of a cumulative score of z is derived

which, in turn, is composed of these eight factors:

P = f(z), where f' > 0.

z = ao + biX, + b2X2 + ... + bSX8,

where bl.. are coefficients,
X1 is ROA (EBIT/TA),
X2 is FCC,
X3 is Balance Ratio,
X4 is Market/Book Ratio,
X5 is Relatedness Ratio,








X6 is Net Rate of Management Stock Acquisitions,
X7 is Relative Sales Growth Rate,
X8 is Capital Intensity.


Hypotheses


Because the methodology used in this dissertation, Logist

Analysis (discussed in Chapter 6), develops a model in which

the coefficients can reveal the role each variable plays, we

can hypothesize about how each factor in the model affects

overall assessment of bankruptcy risk. Specifically,

according to our model, a negative beta means that the larger

the variable, the less the chance of bankruptcy. Similarly,

a positive beta means that the larger the variable, the

greater the chance of bankruptcy. If any of the variables

are able to take on negative values, the rules given above

about the sign of the coefficient should be reversed. We

hypothesize that all coefficients bi through b8 should be

negative.

The coefficients for return on assets, fixed charge

coverage, and the market-to-book ratio should be obvious. In

principle, the sales-income balance ratio's coefficient is

negative because if income is growing more briskly than sales,

then the company is experiencing widening profit margins,

which would inevitably lead to higher profitability and even

more growth in the future. Since long-term considerations

sometimes make it necessary to sacrifice short-run

profitability to drive out the competition and gain market





58

share, however, even good, solid companies may have negative

values for this ratio. Hence, our confidence about the sign

of this particular coefficient is less than for those of the

first three variables. More important than the sign, though,

is the idea that the balance ratio for the bankrupt companies

should be significantly different than the ratio for non-

bankrupt companies.

The relatedness ratio's coefficient is negative because

a larger relatedness ratio indicates that a company is

diversified into related lines of business, which means, as

Rumelt (1982) shows, that the company will be able to achieve

real product synergy and counter-cyclicality. The net rate

of stock acquisition is also expected to be negative because

that signals net purchase of stock by management. This is a

signal that management expects the firm to be profitable in

the long-run.

The coefficient for by, relative sales growth, is

negative because we expect that companies with good relative

sales growth rate will have a negative value for the variable,

and a negative coefficient is needed to reverse the impact on

the chances of bankruptcy. Finally, the sign of the

coefficient for capital intensity is negative because, as

Ohlson (1980) points out, larger capital intensity is

associated with larger company size, and larger industrial

companies do not go bankrupt as easily as smaller ones. On

the other hand, though, to the extent that the reciprocal of






59

the capital intensity ratio is an indicator of asset

efficiency, the negative sign would be counter to the findings

of Zavgren (1985).








Table 11. Industry Classification,
Growth and Bankruptcy Rates


Industry
Number Industry


Food
Clothing & Textiles
Paper & Publishing
Chemicals

Drugs
Petroleum Refining
Rubber & Leather
Glass & Cement
Metals


COMPUSTAT Growth
SIC Rate

2000-2199 8.94%
2200-2399 5.41%
2600-2799 9.51%
2800-2839,
2840-2899 10.84%
2830-2839 11.42%
2900-2999 12.82%
3000-3199 5.33%
3200-3299 7.45%
3300-3499 6.28%


10 Industrial Machinery 3500-3569,
3580-3599 6.33%
11 Office Machinery & 3570-3579,
Electronic Equipment 3650-3679 8.71%
12 Electrical Equipment 3680-3699 14.38%
13 Motor Vehicles 3700-3799 7.96%
14 Scientific &
Surveying Equipment 3800-3899 8.58%
15 Transportation 4000-4599,


Agriculture
Extractive
Construction

Specialty Manuf.
Utilities
Wholesale
Consumer Products
Services
Financial


4700-4799 11.20%
0100-0999 10.42%
1000-1499 10.48%
1500-1799,
2400-2499 9.93%
3900-3999 7.02%
4800-4899 14.37%
5000-5199 10.51%
5200-5999 8.78%
7000-8999 12.39%
6000-6799 19.00%


Bankruptcy
Rate

2.3622%
8.1633%
3.9130%

1.1062%
N/A
1.6393%
0.9788%
2.5000%
3.3898%

2.3490%

6.8452%
8.7805%
6.2201%

3.9474%

11.2245%
2.6316%
5.4217%

5.2239%
13.0435%
1.3245%
6.5421%
4.4983%
3.7516%
2.5097%


Growth Rate is the average annual growth rate of the industry
between 1968 and 1987. Bankruptcy Rate is the percentage of
businesses that failed during that period.





















































I I I


I I II III I II I I III II


1 3 5 7 9 11 13 15 17 19 21 23
2 4 6 8 10 12 14 16 18 20 22 24
Industry Code


Industry and Bankruptcy Rate


14,



12


10


8


Ku
i6


ps________^ _____ &____
\ ^ \

^ ^ $

G n------ ^ -- --


Cum. %


Figure 3.














CHAPTER 6
METHODOLOGY TO TEST EMPIRICAL MODEL


The Logist Analysis Methodology


The methodology used in this dissertation to build the

empirical bankruptcy model is logist analysis.

Logist analysis is a statistical method that computes the

conditional probability that a given observation belongs to

a particular class of observations if certain variables about

the observation are known. Based on a cumulative probability

function, this model does not require that independent

variables be multivariate normals or that the classes have

equal covariance matrices. Instead, the model is solved using

the maximum likelihood method. Thus, logist analysis reduces

the fundamental bankruptcy estimation problem to the

following: given that a company belongs to some pre-specified

population, what is the probability that this company will

fail within some pre-specified period of time?

Ohlson (1980), which was discussed earlier in the

Literature Review section, was probably the first work on

bankruptcy to use logist analysis. Although the research did






63

not produce a significantly viable model, the work

nevertheless provided some interesting insights into the use

of logist analysis for empirical studies of bankruptcy. In

Ohlson's model, Xi denoted a vector of predictors for company

i, p denoted a vector of unknown parameters, and P(Xi, 8),

where P is a probability function (0 P < 1), denoted the

probability of bankruptcy for a given set of vectors Xi and P.

The logarithm of the likelihood of any specific outcome, as

reflected by the binary sample space of bankruptcy versus non-

bankruptcy, is given by:

L(3) = Z log P(Xi, P) + E log (1 P(Xj, P)),

where i, j are elements of the S1 index set of bankrupt

companies and S2 index set of non-bankrupt companies,

respectively. For any specified function P, the maximum

likelihood estimates of P1, P21, ** Pn are obtained by solving:

max, L(P).

Because we do not as yet have a full theory of

bankruptcy, however, we cannot easily find an appropriate

class of functions P. As a practical matter, therefore, we

can only select a function for the sake of computational and

interpretative simplicity. One such function is the logistic

function:

P = (1 + exp{-yi))',

where y, = EZ jXij = P'Xi.

This formula has two implications. First, P is increasing in

y. Second, y is equal to log (P / (1 P)).





64

Like discriminant analysis, logist analysis weights the

independent variables and creates a score for each company.

The Z score obtained may be used to determine the probability

of membership in a group where:

Probability of bankruptcy
= 1 / (1 + exp(-z))
= 1 / (1 + exp(-a + blX + ... + bpXp)).

The b coefficients are weighted so as to maximize the joint

probability of bankruptcy for the known bankrupt companies and

the probability of non-bankruptcy for those companies that did

not go bankrupt. Unlike the coefficients derived from

discriminant analysis, these coefficients tell us the role

that each individual variable plays in the overall empirical

model. Therefore, we can use them to analyze which factors

are the most significant in long-term bankruptcy forecasts.


Loqist versus Discriminant Analysis



Much previous bankruptcy work, most significantly Altman

et al. (1977), has employed the traditional linear

discriminant analysis, although both the linear and the

quadratic forms have been used. For our dissertation, though,

we consider that logist analysis can yield a superior model

because logist analysis does not share many of the problems

faced by discriminant analysis.

First, linear discrimination is basically a multivariate

technique that assigns a score to each element in a sample






65

using a linear combination of independent variables. The

multivariate approach is very appealing because it reduces

several financial dimensions of a problem to a single score.

In general, such reductions have been quite successful. The

bankruptcy models derived from discriminant analysis tend to

have high classification accuracies, at least in the short-

to medium-terms. Serious questions, however, have been raised

about whether so many factors and dimensions of a complex

financial problem like bankruptcy can validly be reduced to

a single score, or whether crucial information would be lost

during the process of such a reduction.

Second, discriminant analysis has several statistical

requirements that are difficult to meet for most samples.

For discriminant analysis to work, the independent variables

must be multivariate normals, and the covariance matrices of

the original and hold-out groups must be equivalent. In

practice, satisfying both assumptions is difficult. The

requirement that the independent variables have multivariate

normal distributions, for example, is frequently violated.

It will be violated whenever a dummy independent variable,

such as the time variable t, is used. Although some remedial

measures, such as log transformations, square root

transformations, and elimination of outliers can be used, such

methods have unclear economic implications which are often too

easily ignored. Further, in many cases the requirement that

covariance matrices be equal is also violated. This means





66

that the group covariances are not statistically equivalent,

as indicated by Box's F statistic.

A way to avoid the latter problem of unequal covariances

is to use quadratic discriminant analysis. Unlike linear

discrimination, the quadratic form does not require that

covariances must be equal. Instead, quadratic discriminant

analysis assesses the covariance of each group independently

as it builds a model. The problem, however, is that quadratic

discriminant analysis is not nearly as widely used as linear

discriminant analysis, and there are also questions about its

model-building powers. Altman et al. (1977) and Marks and

Dunn (1974) both reported that linear discriminant analysis

could achieve greater classification success than quadratic

discriminant analysis. The Marks and Dunn paper reached this

conclusion for samples where the group variances are similar,

the group means are far apart, the sample sizes are small, and

the number of variables is small. Although these two papers

do not conclusively show that linear discriminant analysis is

superior to quadratic discriminant analysis, they suggest that

there are significant problems with using the latter to build

an empirical model of bankruptcy.

We use the logist analysis method because it resolves

both major problems of discriminant analysis. First, unlike

discriminant analysis, it does not reduce all the financial

dimensions of bankruptcy to a single cut-off score. Rather,

it assesses each relevant independent variable and comes up






67

with a probability of bankruptcy, so that, given that a

company belongs to a certain sample, logist analysis provides

the probability of failure. Second, unlike linear

discriminant analysis, logist analysis does not require that

the independent variables be multivariate normals or that

groups have equal covariance matrices. Harrell and Lee (1985)

reported that even when all the assumptions of discriminant

analysis are met, logist analysis is at least as effective as

discriminant analysis. Hence, our results should be much more

significant than under discriminant analysis. Furthermore,

unlike the quadratic version of discriminant analysis, logist

analysis is a sound, proven technique that can provide good

classification accuracy.

Therefore, we believe that the logist analysis

methodology is significantly superior to discriminant analysis

for our research. For this reason, we employ it to build our

empirical model of bankruptcy.


Sample Design


Logist analysis in our research requires two groups of

companies, a bankrupt and a matching sample. Companies that

were on the COMPUSTAT Research Tape and which filed for

Chapter 11 between 1968 and 1987 are used as our bankrupt

companies sample. We select a matching sample of companies

in the same industries and with similar asset sizes but that

avoided bankruptcy.





68

Our data come from several sources. We use the COMPUSTAT

Research Tape and Industry Tape for basic data on financial

variables of both our samples. The COMPUSTAT Research Tape

provides such data for the bankrupt sample, while the Industry

Tape provides such data for our matching sample. We use the

Ownership Reporting System Tape, published by the National

Archives and Record Services, to obtain data on the management

acquisition of company stock. Next, we use the COMPUSTAT

Segments Information Tape to find the segments of our

companies and to compute their degree of diversification.

Finally, we use the COMPUSTAT Research and COMPUSTAT Industry

tapes again to calculate industry growth and bankruptcy rates.

At the start, we had 315 bankrupt companies that were

deleted from the COMPUSTAT Industry Tape and moved to the

COMPUSTAT Research Tape between 1968 and 1987 by a deletion

code of 02, which indicates bankruptcy. We could not,

however, find 5 years of continuous data for all 315 companies

because the information we needed was on several different

tapes, each of which had information on different time

periods. The COMPUSTAT Business Segments Information Tape,

for example, has data only from 1975 onward, as does the

Ownership Reporting System. The Master Current Tape of

Ownership Reporting System, however, offers data from January

1980 to August 1987, and the Master History Tape offers data

from January 1975 to April 1982. Even though we used every

available tape and even calculated several variables by hand,






69

we still could keep only 59 observations in our sample of

bankrupt companies. Most of the companies were "lost" because

we could not find ownership data about management purchase and

sale of stock or segment information about their lines of

business. After determining the composition of the bankruptcy

sample, we matched the sample by industry and asset size and

selected 63 companies that did not go bankrupt and had the

data we needed. Table 12 presents the companies in our

original bankrupt and matching samples.

Orthodox logist analysis requires that samples be

selected randomly from a population of bankrupt and non-

bankrupt companies. In almost all studies on bankruptcy which

have used the logist analysis methodology, however, the sample

has been selected using non-random, state-based criteria.

Therefore, the probability of bankruptcy derived by logist

analysis for any firm i is actually the probability in the

specific sample, not the general population. The relationship

between the probability of bankruptcy based on the population

and the probability of bankruptcy based on the sample depends

on how the sample of bankrupt companies was selected from the

population in general as well as how the sample of non-

bankrupt companies was selected. Because of this, the

probabilities from logist analysis must be adjusted for the

effects of the sample selection, or else they would become

meaningless because by selecting a sample differently, we can

derive completely different results. In the following





70

chapter, we explain how we adjusted our results and discuss

in greater detail how the selection process actually affects

the results of logist analysis.








Table 12. Original (Bankrupt) and Matching Sample
for Deriving Empirical Bankruptcy Model

Original Sample

MI CNUM Company Name To IndDNUM Asset

1 2073 ATI Inc. 84 237399 12.40
2 13900 Aldebaran Drilling Co. Inc. 87 171381 5.83
3 14419 Aldon Industries Inc. 87 22272 9.48
4 25909 American Fuel Technologies 86 42860 1.20
5 37460 Apache Energy & Minerals 84 246792 5.40
6 40150 Argonaut Energy Corp. 86 171311 15.47
7 77266 Beker Industries 85 42870 267.90
8 124187 Buttes Gas & Oil Co. 87 171311 389.52
9 140556 Capitol Air Inc. 84 154511 34.75
10 141602 Cardis Corp. 87 215013 195.20
11 159620 Chargit Inc. 87 237399 19.00
12 163742 Chemical Investors Inc. 83 32640 20.60
13 202666 Commodore Resources Corp. 82 171311 0.09
14 208106 Conner Corp. 87 182451 71.83
15221241 Cosmetic Sciences Inc. 86 238091 1.50
16 225015 Crawford Energy Inc. 85 171381 26.95
17 228885 Crutcher Resources Corp. 86 171389 99.80
18 232827 Cytox Corp. 85 215161 1.30
19 234230 Dakota Minerals Inc. 84 171311 4.70
20 236280 Danker Labs Inc. 85 143851 1.90
21238136 Datatron Inc. 85 215080 7.30
22 254674 Discovery Oil Ltd. 86 171311 15.16
23 278902 Econo Therm Energy Systems 86 93443 22.88
24 292009 Empire Oil & Gas Co. 83 171381 31.34
25 292666 Energy Exchange 85 171311 55.43
26 292935 End-Lase Inc. 86 215080 12.50
27 293799 Enterprise Technologies Inc. 85 215170 33.30
28 364652 Gamex Industries Inc. 82 193990 2.00
29 402274 Gulf Energy Corp. 84 171311 5.05
30 423276 Helionetics Inc. 86 123621 27.90
31 456704 Information Displays Inc. 84 123686 37.00
32 460380 Intl Stretch Prods 84 22200 5.88
33 460468 Intl Teldata Corp. 85 246794 0.65
34 461027 Interstate Motor Freight 84 154210 82.28
35 552813 MGF Oil Corp. 84 171311 342.92
36559150 Magic Marker Industries Inc. 86 193950 8.10
37 585163 Mego International 82 193944 46.00
38 595215 Mid-America Petroleum Inc. 86 171381 22.47
39 628300 Mutual Oil of America Inc. 86 171311 29.97
40 635080 National Business Comm. Corp. 85 225900 5.55
41 638777 NATPAC Inc. 86 225411 25.90
42 654048 Nicklos Oil & Gas Co. 85 171381 96.78
43 682121 OmniMedical 84 237600 9.00
44 712221 People's Restuarants Inc. 86 225812 41.79
45 747623 Quanta Systems Corp. 85 113664 2.40








Table 12--continued.


46 748379
47 761049
48 765361
49 771044
50 795872
51802828
52 805567
53 816068
54 817910
55 925523
56 929073
57 984010
58 984126
59 989875


QuikPrint of America Inc.
Reser's Fine Foods Inc.
Richmond Tank Car Co.
Roblin Industries
Sambo's Restaurants
Santec Corp.
Saxon Industries
Seiscom Delta, Inc.
Servamatic Systems Inc.
Viable Resources Inc.
Vuebotics Corp.
Xenerex Corp.
Xonics Inc.
Zytrex Corp.


Matching Sample


M# CNUM Company Name


1204682
2 866055
3 550819
4 524038
5255264
6870738
7 628850
8 136420
9443784
10 480827
11 205477
12 878504
13 69689
14 674098
15 872625
16 209705
17 786629
18 847660
19 192108
20 914802
21239133
22 971889
23 208258
24 739647
25 666416
26 46357
27 6351
28 250568
29 131069
30 458683
31 369032


Comptek Research Inc.
Summit Energy Inc.
Lydall Inc.
Lee Pharmaceuticals
Diversified Industries Inc.
Swift Energy Co.
NCH Corp.
Canadian Occidental Petroleum
Hudson General Corp.
Jorgensen (Earle M.) Co.
Computer Task Group Inc.
Technical Tape Inc.
Baruch-Foster Corp.
Oakwood Homes
TRC Cos Inc.
Consolidated Oil & Gas
Sage Energy Co.
Speed-O-Print Business Machines
Coeur D'Alene Mines Corp.
University Patents Inc.
Davis Water & Waste
Wilshire Oil of Texas
Conquest Exploration Co.
Prairie Oil Royalties Co. Ltd.
Northgate Exploration Ltd.
Astrex Inc.
Adams Resources & Energy Inc.
DesignCraft Industries
Callahan Mining Corp.
InterGraph Corp.
General Automation


T. IndDNUM Asset


237372
171311
22200
42844
246200
171311
42842
171311
154580
215051
237372
32640
171311
182451
238911
171311
171311
215081
171040
143851
215051
171311
9480
171311
171040
215065
215172
193911
171040
123686
123681


12.44
28.80
50.14
6.50
68.23
12.60
272.10
317.94
48.30
197.11
29.92
31.00
18.10
56.00
3.04
123.23
139.40
7.65
22.58
18.18
36.48
64.04
17.19
32.75
117.60
25.30
35.58
7.62
55.97
28.90
38.60


246794
12013
133743
93312
225812
123688
32600
171382
175900
171311
237391
171311
143861
113674


2.30
9.70
101.30
32.13
220.94
2.90
486.60
23.36
26.36
6.25
2.10
30.95
15.40
6.50








Table 12-continued.

32 732852 Pope, Evans & Robbins Inc. 84 22330 4.89
33 647072 New Mexico & Arizona Land 85 246519 36.49
34 893552 Transcon Inc.-California 84 154213 79.30
35960878 Westmoreland Coal Co. 84 171211 376.30
36 55716 BSN Corp. 86 193949 22.24
37 43147 Artra Group Inc. 82 193960 51.91
38 553748 MSR Explorations Ltd. 86 171311 26.90
39 655555 Nord Resources Corp. 86 171090 66.10
40204909 Computer Factory Inc. 85 225995 13.88
41885539 Three D Department 86 225700 21.15
42 379355 Global Natural Resources Inc. 85 171311 105.80
43 29429 American Science Engineering 84 237391 13.97
44 362232 G R I Corp 86 225961 49.34
45 362360 GTI Corp 85 113679 13.05
46 170819 Christiana Companies 83 246552 44.48
47 208093 Connelly Containers Inc. 86 12030 18.50
48 879369 TeleFlex Inc. 83 133714 109.30
49 707389 Penn Engineering & Mfg Corp. 85 93452 39.01
50476502 Jerrico Inc. 81 225812 238.50
51238085 Datametrics Corp. 85 123688 3.80
52 313693 Federal Paper Board Co. 82 32631 472.83
53 212576 Convest Energy Partners 86 171311 34.49
54 658136 North Canadian Oils Ltd. 86 171311 182.00
55 427879 Hershey Oil Corp. 84 171311 30.38
56 27258 American List Corp. 84 237331 2.39
57 208285 Conquest Exploration Co. 84 171311 123.16
58 942622 Watsco Inc. 84 143822 19.97
59 590262 Merrimac Industries Inc. 84 113663 9.34

"M#" is the matching number, which is used to match the
observations in the bankrupt and matching samples. "CNUM" and
"DNUM" are the company and industry classifications. "To" is
the year during which the company filed for Chapter 11 for
companies in the bankrupt sample, or the last year data was
collected for our research for companies in the matching
sample. "Ind." is the industry code according to our
definitions. "Asset" is the company's asset size.














CHAPTER 7
RESULTS OF EMPIRICAL MODELS OF BANKRUPTCY



Using the sample data described in Chapter 6, we were

able to derive five probabilistic models, designated P1

through P5. All of them are based on the factors described in

Chapter 5, and they assess bankruptcy risk one through five

years ahead of time corresponding to the subscription Pn. The

only difference among these distinct models is in their

coefficients: they all assume the form discussed in Chapter

5. Table 13 shows the signs of the coefficients of the

variables for each model P,.

Most of these coefficients conform to our expectations

as explained in Chapter 5, but others show significant

differences. The signs for the coefficients of the balance

ratio and the net rate of management stock acquisitions seem

to vary randomly but lean towards being positive, while we

expected both to be negative. Both of these variables can be

positive or negative, and our expectation was stated for the

"normal" case for which the expected value of the variable is

positive. Looking at the raw data, a significant number of

the values in both samples were negative, so we would have to






75

reverse the coefficient sign convention. Hence, the positive

signs give the expected signal and it is only our view of

"normal" values for the variables that could not be verified

empirically. We suspect that this result can be explained in

large part by the way the matching sample was constructed.

The match was made by industry and asset size, and it is

evident that many of the pairs came from "troubled"

industries. It is an empirical question, but we suspect that

a random sample from all industries would conform to the

original expectations.

The other sign anomalies are not as troubling. The first

four variables (Group I) are expected to be most significant

close to bankruptcy, and three of the four have the correct

signs in the first three years. Also, the last four variables

(Group II) are expected to be most significant in earlier

years, and three of the four have the correct sign in the last

four years. We believe that this pattern confirms the

validity of our expectations.

Logist analysis provides a technique that allows us to

find the most significant variables in any predictive model.

A variable is "most significant" if, by Chi-Square Q-statistic

and MLE's statistic, they meet the requirements of entry and

stay significance levels pre-specified for the model. Our

entry and stay significance levels were set at 0.05. Table

14 shows the most significant variables in each of our models.

Capital intensity is significant in all periods. This means





76

that the nature of a company's line of business always plays

a significant role. When bankruptcy is far into the future,

our data indicate that this factor plays the largest role of

the eight variables used in our model. As we approach

bankruptcy, especially one year ahead of bankruptcy, however,

ROA, FCC, and Market/Book Ratio become increasingly

significant. This confirms our view that in the short-term,

variables in the company operating and profitability

indicators group that are weighted toward a company's current

financial data would play significant roles. Other factors,

such as relative sales growth, balance ratio of income and

sales growth, and management stock acquisition, have also

played increasingly larger roles near the time of bankruptcy.

To investigate this timing phenomenon further, we built

and tested separate predictive models based on the groups

defined in Chapter 5. Group I was composed of company

operating and financial indicators and included ROA, market-

to-book ratio, FCC, and balance ratio. Group II was composed

of fundamental company characteristics indicators and included

the net rate of management stock acquisitions, the relative

sales growth rate, the relatedness ratio, and capital

intensity. We then compared the predictive power and

effectiveness of the main models built with all eight

variables with those of the Group I and Group II models.

Table 15 shows the predictive powers and effectiveness of all

three sets of models. Figure 4 shows the predictive power of






77

our main model, and Figure 5 compares the Group I and Group

II models.

We then investigated into the classification powers of

our model by examining the empirical probability density

functions for bankrupt and non-bankrupt firms. We divided the

range of probability of bankrupt from 0 to 1.0 into ten equal

intervals. The percentage of bankrupt and non-bankrupt firms

relative to the total number of firms that they present which

fall within each of these intervals for the five different

time periods ti through t5 are tabulated and shown in Tables

16 and 17. The percentages are plotted against the mid-value

of the interval to obtain the discrete approximation of the

distributions of the bankrupt and non-bankrupt probabilities

in Figures 6 to 15. The probabilities of the two groups are

shown to diverge significantly by their respective bar graphs.

The bankrupt group is clearly skewed toward the higher

probabilities of failure that our model derived, while the

non-bankrupt group is clearly skewed toward the lower

probabilities of failure. These results are similar to those

of Zavgren (1985). Her paper, however, only presented

original probabilities, while we present both the original

probabilities and those probabilities adjusted for sample

selection (as discussed below).

As mentioned in the previous chapter, because we

selected our data with non-random, state-based criteria, we

must adjust the probabilities derived from logist analysis.





78

For any given firm i in the general population with a

probability P of bankruptcy, logist analysis would give a

probability P' of bankruptcy for that company in our specific

sample. We must find ways of finding a relationship between

P and P' and also between the structures of our samples.

Assuming that there are N, bankrupt and N2 non-bankrupt firms

in the general population and n, bankrupt and n2 non-bankrupt

firms in our bankrupt and non-bankrupt samples, P', according

to Bayes' formula for conditional probability, is equal to:

P' = P (n,/N1) / {P (n,/N,)+ (l-P) (n2/N2)) (1)

Previous work, such as Palepu (1986), which tried to predict

merger targets using logist analysis models, have partially

explored this relationship, but he only gave a formula for P'

in the special case when n,=N,. We, however, derived a general

formula for P' for cases when nI is not equal to N1, and n2 is

not equal to N2:

Let al = n,/N,,
a2 = n2/N2,
Then formula (1) can rewritten as

P' = (a, P)/{(al*P) + (l-P) a2)} (2)

Substituting P = (1 + exp(-Yj))"' into (2), we derive the
formula for any value of a, and a2:

P' = (1 + exp{lg(a2/a1)-Yj}) 1. (3)

This formula implies the following relationships between the

samples and P' and P:

If aI = a2:
P' = P, Type I error will not change,
Type II error will not change.









If a, > ag:
P' > P, Type I error will increase,
Type II error will decrease.

If al < a.,
P' < P, Type I error will decrease,
Type II error will increase.

According to this formula, without adjustments, we can

derive a model with an artificially high or artificially low

type I error by settling for an artificially low or

artificially high type II error, or vice versa, simply by

selecting the right samples. After adjustments, artificial

type I and II errors are still possible because, even then,

a, and a2 are part of the model. Therefore, to have meaningful

probability models with the logist analysis methodology, we

must always first adjust for al and a2 and then report their

values along with our type I and type II errors. Table 18

presents such data for our sample.

Since a, > a2 in our samples, type I error increased after

adjustments were made, while type II error decreased. We can

observe this same phenomenon by comparing Figures 6 through

10 with Figures 11 through 12, respectively.

Our data lead to some interesting conclusions. First,

we have succeeded in building a model whose predictive power

does not fall precipitously as time before bankruptcy

increases. Whereas Altman's ZETA model, under optimal

conditions, is 96% accurate one year before bankruptcy but

only 60% accurate five years before bankruptcy, our model's





80

predictive power, based on data from one experimental sample,

is relative stable. It is most effective one year before

bankruptcy, with 96.0% accuracy (as measured by the C-Index

of logist analysis), but even at its lowest point, three years

before bankruptcy, it still has 89.0% accuracy, and five years

before bankruptcy it is 95.1% accurate.

Second, we have found that during different periods

before bankruptcy, different groups of variables become

important in predicting bankruptcy. When close to bankruptcy,

the Group I financial variables have very high predictive

power. As we move farther back in time and try to predict

bankruptcy farther ahead of time, we found that these

variables' predictive powers started to wane. The Group II

fundamental variables became increasingly more powerful as we

moved farther back in time. This demonstrates that there is

a time relationship between our fundamental variables and the

future financial position of the company.








Table 13. Signs of Coefficients

Model bI b2 b3 b4 b5 b6 b7 b_

P + + +

P2 + +

P3 + + -

P4 + + -

P5 + -

PI through P5 are models for predicting bankruptcy one to five
years ahead of time.

bI through b8 are coefficients of the following variables:
b,: ROA (EBIT/TA)
b2: FCC
b3: Balance Ratio
b4: Market/Book Ratio
b: Relatedness Ratio
b: Net Rate of Management Stock Acquisitions
b7: Relative Sales Growth Rate
b8: Capital Intensity.







Table 14. Most Significant Variables in
Forecasting Bankruptcy

Model Most Significant Variables

P, ROA, FCC, Balance Ratio, Capital Intensity,
Market/Book Ratio

P2 Relative Sales Growth, FCC, Capital Intensity

P3 ROA, Net Rate of Management Stock Acquisitions,
Relative Sales Growth, Capital Intensity

P4 Capital Intensity

P5 Capital Intensity
P through P5 are models for forecasting bankruptcy one through
five years ahead of time, respectively.









Table 15.


Predictive Power and Effectiveness
of Different Models


Years:


Eight-Variable Models:
C-Index: 0.960
X2 59.75
(SL with 8 DF) 0.50%
L.H. Ratio Index 0.5969

Group I Models:
C-Index: 0.929
X2 49.63
(SL with 4 DF) 0.50%
L.H. Ratio Index 0.4958

Group II Models:
C-Index: 0.819
X2 16.45
(SL with 4 DF) 0.50%
L.H. Ratio Index 0.1643


0.957
69.31
0.50%
0.6133


0.873
39.88
0.50%
0.3529


0.770
27.43
0.50%
0.2427


0.892
48.48
0.50%
0.4135


0.771
16.33
0.50%
0.1365


0.835
35.19
0.50%
0.3001


0.909
44.51
0.50%
0.4819


0.740
10.33
5.00%
0.1074


0.863
35.31
0.50%
0.3823


5


0.951
45.94
0.50%
0.5980


0.598
3.83
50.00%
0.0460


0.889
32.47
0.50%
0.4227


The C-Index comes from logist analysis and indicates the
classification accuracy of the model. X2 (or Chi-Square) is
the -2 log likelihood ratio chi-square statistic of the model.

























IUU




80 -




60-




40




20





1 2 3 4 5


Yeors before Bankruptcy












Figure 4. Classification Accuracy of Empirical Model of
Bankruptcy


Accuracy

























Group I


Group II


1 2 3 4 5


Years before Bankruptcy











Figure 5. Classification Accuracy of Models Based on Group I
and Group II Variables






s 86



i3

dPe doPdP dPOP O dPdP dOde 00
icoto O n c co H 00- n com w0
i00 0 0 0 N 0 0 0v 0 0 m 0 d (0



o u-I eq
000 0HO NHH < 0000 O


0 N 0,4
r.dp dodP dp ef dpdP do C -r


O cn c* -eqtn no

0 ~ n rOc Oqq> r o w o

'4-4
0 *** **

co ion toGo Ln N qeqm N
r-

oo ooe oM ON O' o
wp ap de do dp do d dP dp de
> in oo mn mo m










J .** ,
r4 dP P dpoo dP or0 p
4Q tnocr IC










A 000 oA V o



Lf *. .* ** "
*I- ,. il H






S HO do doPdP doO dEdp0 dfd 0M
rq8 ino4mcn onoc oOm>Dw om w $o 4j



w 8M 4M Al C; C









M.0 0 0q i 0 d
c 0cow ioo o0o 0I0 .oo 0




Sdpd d0 dP dPd 0 p Op dP dP 0
10- Ln in oo 1n0Ln inwO in %o
4 0%M 000 HOM 0m 0 -



H f 0MO ei OP df 0M e Ov Ov a
*o0 Io 0 own -N 0
HI 4
'M dpg dpdp edpd d dp d > )
1) 1m 1 9 0) 1 1 w


Eo r. ra rCt

-1 0 cJ 0 "q 0 f o rin 0 t go o q inq 0 to zC o



















O't


0 M
00





ON








or

* *
dP d



0 c

Octo


0
-,.4


4-P







-H
muI



'iq


.l



u-I
c0
.o





A 0


0cl







S41

(0





to


OON


rlOW
OH d

00C0 0

N r

dp dp



mn o o


Sdp










\o o
# 0




c ;oo
0ooo
H 0
dkp dp









nmoo
** *
dp dP



c10O







dP dP
In oo










dp dp
Ln in o


de

0 (M fM



cL LA





de 0\
U)
O..J0






Oi N

0p\ d
*
MM





LOO

Co rmo


d dP


0 N N

ONN








** *
0000




ocoo0




U)I r-I o
0 w 0

0 C
(n


0) 0


DH >QM >AH
I 0 I I 0) 1 1 4 )

0 o 4 -r HO *-H 0 o r-I
S Z CO E-4 Zn m E


dOe d

i\ in in
** *

OaCO

LOLN N










do tO CM
OOILA
CMM


w Ln r-
S* *
LNA






,0 L> N
LOCm


* *
O 0


ino


* *
. 0
)O




* *
No%



**ro

* *
%00

O 0


coo
de dp
r 0





H 0
O *
00 0


dP d





*
O~rl

dP dp















0c.O


(no
e 0




dPdP
\00
NO


* *

n u?
W r


Ln w 0

o tn o
*~o**
000D


SNO
Os 0




00
co

in
NO
LA


0 8


: *






4-4
4J









:0
1)4












.0






-'.4
-ato
4-)
C






















0
4.
0



l 0









r


4J V 4J 4.)4.) r-H 0
3 aaa LO to
M4 > 4a >1ai

Oi C O E -C
00 -'. -' 0 r0 S:
2;o 2 sam = I?


0d d






omo
o c4 c4
r-l

dp do
a n o




** *


0H


U)mo
Hoo


ONO
dpdP

H 0 0




LO C o

C;
* *











































0.25 0.45
0.15 0.35


0.65 0 85
0.75 0.95


Probability Interval Mid-Volues










Figure 6. Unadjusted Empirical Probability Density
Function, 1 Year Before Bankruptcy


Non-bkpt

Bankrupt


0.7-


0.6-


0.5-


0.4-


0.3-


0.2-


0.1-


0.05


mnnIJinl J


























Non-bkpt


Bankrupt


0.05 0.25 0.45 0.65 0.85
0.15 0.35 0.55 0.75 0.95


Probability Interval Mid-Values












Figure 7. Unadjusted Empirical Probability Density
Function, 2 Years Before Bankruptcy


























Non-bkpt


3arkrupt


0.05 0.25 0.45 0.65 0.85
0.15 0.35 0.55 0.75 0.95

Probability Interval Mid-Values











Figure 8. Unadjusted Empirical Probability Density
Function, 3 Years Before Bankruptcy



























Non-bkpt


Bankrupt


0.05 0.25 0.45 0.65 0.85
0.15 0.35 0.55 0.75 0.95


Probability Interval Mid-Values












Figure 9. Unadjusted Empirical Probability Density
Function, 4 Years Before Bankruptcy




























0.8


0.7Non-bkpt


Bankrupt
0.6
c

0
a 0.5-
o

0.4-
0
C
0
0.3-
U


0.2-


0.1 -


0--
0.05 0.25 0.45 0.65 0.85
0.15 0.35 0.55 0.75 0.95


Probability Interval Mid-Values














Figure 10. Unadjusted Empirical Probability Density
Function, 5 Years Before Bankruptcy























0.7


0.6


0.5


0.4-


0.3-


0.2-


0.1-


0.05 0.25
0.15


7 urJ l~a


0.65 0.85
0.55 0.75


Prob'atlitly Interval Mid-Values











Figure 11. Adjusted Empirical Probability Density
Function, 1 Year Before Bankruptcy


Non-bkpt

Bankrupt


Y

























0.8


0.7


0.6


0.5-


0.4-


0.3-


0.2-


0.1-


Probability Interval Mid-Values













Figure 12. Adjusted Empirical Probability Density
Function, 2 Years Before Bankruptcy


15 0.25 0 45 0.65 0.85
0.15 0.35 0.55 0.75 0.95


Non-bkpt


Bankrupt



























Non-bkpt


Bankrupt


0.05 0.25 0.45 0.65 0.85
0.15 0.35 0.55 0.75 0.95


Probability Interval Mid-Values












Figure 13. Adjusted Empirical Probability Density
Function, 3 Years Before Bankruptcy




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EEX7GCBA1_URHWZ0 INGEST_TIME 2017-07-12T20:52:18Z PACKAGE AA00003340_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 2

%$1.5837&< 678',(6 (03,5,&$/ :25.6 21 35(',&7,21 $1' ),1$1&,$/ 0$5.(76 %\ .(4,$1 %, $ ',66(57$7,21 35(6(17(' 72 7+( *5$'8$7( 6&+22/ 2) 7+( 81,9(56,7< 2) )/25,'$ ,1 3$57,$/ )8/),//0(17 2) 7+( 5(48,5(0(176 )25 7+( '(*5(( 2) '2&725 2) 3+,/2623+< 81,9(56,7< 2) )/25,'$

PAGE 3

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

PAGE 4

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

PAGE 5

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nV VWRFNnV GDLO\ UDWHV RI UHWXUQ ,9

PAGE 6

,W ZDV IRXQG WKDW ULVN SUHPLD GLG LQFUHDVH DV FRPSDQLHV DSSURDFKHG EDQNUXSWF\ DQG WKDW WKH ULVN SUHPLD RI FRPSDQLHV ZKLFK ZHUH ODWHU OLTXLGDWHG URVH PRUH WKDQ WKH ULVN SUHPLD RI FRPSDQLHV ZKLFK ODWHU UHRUJDQL]HG 7KH HYHQWVWXG\ VKRZHG WKDW WKH GDLO\ UDWHV RI UHWXUQV RI VWRFNV RI FRPSDQLHV ZKLFK ODWHU ZHQW EDQNUXSW GLG IDOO VLJQLILFDQWO\ GXULQJ WKH ILUVW GRZQJUDGH XQOLNH WKRVH RI FRPSDQLHV ZKLFK GLG QRW 2Q WKH VXEMHFW RI EDQNUXSWF\ SUHGLFWLRQ WKLV GLVVHUWDWLRQ ORRNHG DW ZKHWKHU DQ DFFXUDWH ORQJWHUP EDQNUXSWF\ PRGHO FRXOG EH GHYHORSHG XVLQJ YDULDEOHV ZKLFK DVVHVV D FRPSDQ\n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

PAGE 7

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nV XWLOLW\ ,Q SUDFWLFH UHVHDUFKHUV KDYH WULHG WR GHYHORS SUHGLFWLYH PRGHOV RI EDQNUXSWF\ WKDW FDQ DOHUW LQWHUHVWHG SDUWLHV WR WKH LPSHQGLQJ GDQJHUV RI EDQNUXSWF\ EHIRUH LW LV WRR ODWH WR WDNH FRUUHFWLYH DFWLRQV %HFDXVH SUDFWLFDOO\ DQ\ JURXS LQYROYHG ZLWK D FRPSDQ\ ZRXOG EH LQWHUHVWHG LQ WKH ULVN RI EDQNUXSWF\ LW

PAGE 8

IDFHV SRWHQWLDOO\ ZLGHVSUHDG GHPDQG IRU VXFK D PRGHO KDV LQVSLUHG PXFK UHVHDUFK LQ WKLV DUHD 7R GDWH VHYHUDO DSSURDFKHV KDYH EHHQ WDNHQ WR GHYHORS SUHGLFWLYH PRGHOV WKH\ DUH GLVFXVVHG LQ WKH /LWHUDWXUH 5HYLHZ VHFWLRQ :KLOH VXEVWDQWLDO ZRUN KDV EHHQ GRQH RQ YDULRXV DVSHFWV RI EDQNUXSWF\ LQFOXGLQJ HPSLULFDO IRUHFDVWLQJ DQG SUHGLFWLRQ ZH EHOLHYH WKDW SDVW ZRUNV KDYH OHIW D IHZ DUHDV XQH[SORUHG )LUVW PDQ\ VXFFHVVIXO SUHGLFWLYH PRGHOV RI EDQNUXSWF\ KDYH EHHQ GHYHORSHG EXW PRVW RI WKHVH PRGHOV EDVH WKHLU DVVHVVPHQW RI D FRPSDQ\n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n EDQNUXSWF\ ULVN $IWHU DOO LI WKH PDUNHW ZHUH WUXO\ HIILFLHQW WKHQ LW VKRXOG LQFOXGH WKH ULVN RI EDQNUXSWF\ DV RQH RI WKH GHWHUPLQDQWV RI WKH YDOXH RI D FRPSDQ\nV VHFXULWLHV +HQFH WKH PDUNHWnV DVVHVVPHQW RI D FRPSDQ\nV

PAGE 9

VHFXULWLHVn ULVN VKRXOG LQFOXGH DV D FRPSRQHQW WKH FRPSDQ\nV FKDQFHV RI EDQNUXSWF\ ,Q WKLV GLVVHUWDWLRQ ZH DGGUHVV ERWK RI WKHVH DUHDV WKDW SUHVHQW EDQNUXSWF\ UHVHDUFK KDV ODUJHO\ RYHUORRNHG )LUVW ZH H[SORUH WKH PDUNHWnV DELOLW\ WR nnVHQVH FRPLQJ EDQNUXSWF\ E\ ORRNLQJ DW WKH ULVN SUHPLD RI FRUSRUDWH ERQGV :H ILUVW H[DPLQH ZKHWKHU VXFK ULVN SUHPLD ULVH DV WKH HYHQW RI EDQNUXSWF\ DSSURDFKHV 7KHQ ZH LQYHVWLJDWH ZKDW UROH WKH ERQG UDWLQJ DJHQFLHV 6WDQGDUG t 3RRUnV DQG 0RRG\nV SOD\ LQ WKH PDUNHW DVVHVVPHQW RI D FRPSDQ\nV RYHUDOO ULVN LQ JHQHUDO DQG EDQNUXSWF\ ULVN LQ SDUWLFXODU ,Q RWKHU ZRUGV LI WKH PDUNHW DFWXDOO\ GRHV LQFOXGH EDQNUXSWF\ ULVN LQ LWV RYHUDOO DVVHVVPHQW RI D FRPSDQ\nV ULVN WKHQ KRZ GR 6t3nV DQG 0RRG\nV UDWLQJV DIIHFW WKLV PDUNHW DVVHVVPHQW" 'R WKH\ SURYLGH QHZ LQIRUPDWLRQ WR WKH PDUNHW DQG WKHUHIRUH VHUYH DV D FUXFLDO OLQN LQ WKH SURFHVV RU GRHV WKH PDUNHW LWVHOI DOUHDG\ UHIOHFW DOO RI WKH LQIRUPDWLRQ WKHVH UDWLQJ DJHQFLHV SURYLGH" 7R DQVZHU WKLV TXHVWLRQ ZH KDYH WR VWXG\ WZR SUREOHPV )LUVW DVVXPLQJ WKDW WKH ILUVW FRQVLVWHQW GRZQJUDGH RI D FRUSRUDWH ERQG SURYLGHV WKH PRVW UHOHYDQW LQIRUPDWLRQ ZH GHWHUPLQH ZKHWKHU 6t3nV DQG 0RRG\nV GRZQJUDGHV SUHFHGH RU IROORZ LQFUHDVHV LQ WKH ERQGnV ULVN SUHPLXP 6HFRQG ZLWK WKH VDPH DVVXPSWLRQ ZH H[SORUH ZKDW LPSDFW LI DQ\ WKH ILUVW ERQG GRZQJUDGH KDV RQ WKH SHUIRUPDQFH RI WKH FRPSDQ\nV VWRFN $UJXDEO\ LI QRWKLQJ KDSSHQV WR WKH VWRFNnV UHWXUQ RU WKH ERQG ULVN SUHPLXP DIWHU D ERQG GRZQJUDGH WKHQ ZH FDQ LQIHU WKDW

PAGE 10

WKH PDUNHW KDV DOUHDG\ DEVRUEHG DOO WKH LQIRUPDWLRQ 6t3nV DQG 0RRG\nV SURYLGH LQ WKH UDWLQJ FKDQJH ,Q WKDW FDVH ZH FDQ VD\ WKDW 6t3nV DQG 0RRG\nV SURYLGHG QR QHZ LQIRUPDWLRQ LQ WKH PDUNHWnV DVVHVVPHQW RI D FRPSDQ\nV EDQNUXSWF\ ULVN ,I KRZHYHU WKH PDUNHW UHDFWV VWURQJO\ DIWHU D ERQG GRZQJUDGH DQG WKH ERQGnV ULVN SUHPLXP ULVHV RU WKH VWRFN UHWXUQ FKDQJHV VLJQLILFDQWO\ ZLWK WKH GRZQJUDGH WKHQ ZH FDQ VD\ WKDW 6t3nV DQG 0RRG\n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nV FXUUHQW ILQDQFLDO SRVLWLRQ EHFRPHV OHVV VLJQLILFDQW ZKLOH RWKHU FXUUHQWO\ XQH[SORUHG IDFWRUV SOD\ LQFUHDVLQJO\ ODUJHU UROHV )RU H[DPSOH D FRPSDQ\nV

PAGE 11

FXUUHQW UHWXUQ RQ DVVHWV 52$f PD\ EH FULWLFDO WR ZKHWKHU LW LV VROYHQW RU EDQNUXSW ZLWKLQ WKH QH[W IHZ \HDUV EXW LWV ORQJWHUP ILQDQFLDO KHDOWK ZRXOG SUREDEO\ GHSHQG OHVV RQ FXUUHQW SURILW OHYHOV WKDQ RQ WKH IXQGDPHQWDO FKDUDFWHULVWLFV RI WKH FRPSDQ\ WKDW SUHGHWHUPLQH IXWXUH 52$ +HQFH WR PDNH DFFXUDWH ORQJWHUP EDQNUXSWF\ IRUHFDVWV ZH PXVW H[DPLQH QRW RQO\ D FRPSDQ\nV FXUUHQW ILQDQFLDO SRVLWLRQ EXW DOVR LWV PRUH IXQGDPHQWDO RSHUDWLQJ FKDUDFWHULVWLFV VXFK DV LWV OLQHV RI EXVLQHVV GHJUHH RI GLYHUVLILFDWLRQ PDQDJHPHQW HIILFLHQF\ DQG JURZWK :H DGRSW WKLV DSSURDFK WR GHYHORS DQ HPSLULFDO PRGHO RI EDQNUXSWF\ WKDW FDQ IRUHFDVW WKH H[WHQW RI EDQNUXSWF\ ULVN ,Q DGGLWLRQ WR D JURXS RI IRXU YDULDEOHV GHVLJQHG WR DVVHVV D FRPSDQ\nV FXUUHQW ILQDQFLDO SRVLWLRQ ZH DGRSW YDULDEOHV WKDW GHVFULEH D FRPSDQ\nV PDQDJHPHQW RZQHUVKLS SRVLWLRQ LWV GHJUHH RI GLYHUVLILFDWLRQ WKH OLQHV RI EXVLQHVV LQ ZKLFK LW RSHUDWHV DQG WKHLU UHODWLRQVKLS DQG LWV JURZWK UHODWLYH WR WKH UHVW RI WKH LQGXVWU\ :H H[SHFW WKDW D PRGHO EDVHG RQ WKLV EURDGHU VSHFLILFDWLRQ RI YDULDEOHV ZLOO KDYH KLJK ORQJn WHUP DV ZHOO DV VKRUWWHUP SUHGLFWLYH SRZHUV 'LVVHUWDWLRQ 2XWOLQH ,Q WKH IROORZLQJ FKDSWHUV ZH SUHVHQW WKH UHVHDUFK GHVLJQ DQG ILQGLQJV RI RXU GLVVHUWDWLRQ &KDSWHU H[DPLQHV FXUUHQW OLWHUDWXUH RQ ERWK EDQNUXSWF\ SUHGLFWLRQ DQG WKH HIIHFWV RI ERQG UDWLQJ FKDQJHV ZKLFK DV VWDWHG HDUOLHU DUH XVHG WR

PAGE 12

VWXG\ ILQDQFLDO PDUNHWVn VLJQDOV DERXW EDQNUXSWF\ &KDSWHU GLVFXVVHV WKH WHFKQLTXHV DQG PHWKRGRORJLHV ZH HPSOR\ WR VWXG\ WKH ILQDQFLDO PDUNHWn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

PAGE 13

1RWH &RQVLVWHQW PHDQV D ERQG GRZQJUDGH DIWHU ZKLFK WKHUH ZHUH QR XSJUDGHV XQWLO WKH FRPSDQ\ ILOHG IRU &KDSWHU ,Q RWKHU ZRUGV WKLV GRZQJUDGH LV WKH ILUVW RI D VHULHV RI GRZQJUDGHV ZKLFK HYHQWXDOO\ OHDG WR EDQNUXSWF\ XQLQWHUUXSWHG E\ DQ\ XSJUDGHV RI WKH VDPH ERQG

PAGE 14

&+$37(5 /,7(5$785( 5(9,(: 6XPPDU\ RI &XUUHQW /LWHUDWXUH 0DUNHW (IILFLHQF\ 6WXGLHV RI %RQG 5DWLQJ &KDQJHV .DW] f LQ RQH RI WKH HDUOLHVW ZRUNV RQ ERQG UDWLQJ FKDQJHV GHYHORSHG DQ HYHQWRULHQWHG PHWKRGRORJ\ IRU WHVWLQJ WKH HIILFLHQF\ RI WKH ERQG PDUNHW +H ORRNHG IRU XQXVXDO EHKDYLRU LQ D ERQGn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

PAGE 15

:HLQVWHLQ f WULHG WR GHWHUPLQH LI ERQG UDWLQJ FKDQJHV FRQWDLQHG QHZ LQIRUPDWLRQ E\ VWXG\LQJ WKH ERQGVn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f VWXGLHG WKH HIIHFWV RI ERQG UDWLQJ FKDQJHV RQ WKH PDUNHW UHWXUQV RI VWRFNV GXULQJ WKH SHULRG IURP -DQXDU\ WR 6HSWHPEHU )RU HDFK VWRFN WKH\ GHULYHG D PDUNHW UHWXUQ EDVHG RQ LWV EHWD DQG PHDVXUHG WKH DFWXDO UHWXUQ DJDLQVW WKH H[SHFWHG UHWXUQ IRU D SHULRG RI WKLUW\ PRQWKV EHIRUH WR WZHOYH PRQWKV DIWHU D UDWLQJ FKDQJH 7KHLU VWXG\ FDOFXODWHG GLVWXUEDQFH WHUPV UHVLGXDOVf RI VWRFN UHWXUQV GXULQJ WKH SHULRG 3LQFKHV DQG 6LQJOHWRQ FRQFOXGHG WKDW DOO FKDQJHV DWWULEXWDEOH WR FRPSDQLHVn ILQDQFLDO VLWXDWLRQV ZHUH IXOO\ DQWLFLSDWHG WR PRQWKV DKHDG RI

PAGE 16

WLPH ZKLOH DOO FKDQJHV DWWULEXWDEOH WR FRPSDQ\VSHFLILF HYHQWV ZHUH DQWLFLSDWHG VL[ PRQWKV DKHDG RI WLPH 7KXV DOWKRXJK WKHUH ZHUH DEQRUPDOO\ KLJK DQG ORZ UHWXUQV FRUUHVSRQGLQJ WR XSJUDGHV DQG GRZQJUDGHV UHVSHFWLYHO\ EHIRUH D UDWLQJ FKDQJH WKHUH ZHUH QRUPDO UHWXUQV DIWHU WKH UDWLQJ FKDQJH $JDLQ D VWXG\ FRQFOXGHG WKDW ERQG UDWLQJ GRZQJUDGHV SURYLGHG QR QHZ LQIRUPDWLRQ WR WKH PDUNHW )LQDOO\ *ULIILQ DQG 6DQYLQFHQWH f XVHG WKUHH GLIIHUHQW PHWKRGRORJLHV WR VWXG\ WKH HIIHFWV RI UDWLQJ FKDQJHV RQ FRPPRQ VWRFN SULFHV 7KHLU VWXG\ FRQWDLQHG UDWLQJ FKDQJHV IURP WR )LUVW WKH\ XVHG D SRUWIROLR PHWKRG VLPLODU WR WKDW RI :HLQVWHLQ f 7KHQ WKH\ HPSOR\HG D RQHIDFWRU DQG D WZRIDFWRU PRGHO EDVLQJ WKHLU H[SHFWHG VWRFN SULFHV RQ EHWDV DV KDG 3LQFKHV DQG 6LQJOHWRQ f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nV \LHOGWRPDWXULW\ DQG WKH \LHOG

PAGE 17

WRPDWXULW\ RI D ULVNIUHH VHFXULW\ 6HFRQG ZH FRQVWUXFW RXU VDPSOHV QRW E\ LQGXVWU\ EXW E\ WKH QDWXUH RI WKH HYHQW ,Q RWKHU ZRUGV ZH GHILQHG WKH HYHQW DV WKH ILOLQJ RI &KDSWHU XQGHU WKH )HGHUDO %DQNUXSWF\ $FW $V IDU DV ZH NQRZ WKLV LV WKH ILUVW VWXG\ RI ERQG UDWLQJ FKDQJHV WR EH EDVHG RQ GDWD RI FRPSDQLHV IURP DOO LQGXVWULHV 7KHRUHWLFDO 0RGHOV RI %DQNUXSWF\ :LOFR[ f LV RQH RI WKH HDUOLHVW DQG PRVW SULPLWLYH WKHRUHWLFDO PRGHOV RI EDQNUXSWF\ ,W DVVXPHV WKDW D FRPSDQ\ VWDUWV ZLWK D SRVLWLYH DPRXQW RI FDSLWDO ZKLFK FKDQJHV UDQGRPO\ RYHU WLPH 3RVLWLYH FKDQJHV LQ LQGLFDWH SRVLWLYH FDVK IORZ DQG LQFUHDVHV LQ WKH FRPSDQ\nV DVVHWV ZKLOH QHJDWLYH FKDQJHV LQ LQGLFDWH ILQDQFLDO ORVVHV ZKLFK UHTXLUH WKH FRPSDQ\ WR OLTXLGDWH DVVHWV :KHQ D FRPSDQ\nV LV VXIILFLHQWO\ QHJDWLYH LW EHFRPHV EDQNUXSW ([SUHVVLRQV IRU WKH H[SHFWHG SUREDELOLW\ RI EDQNUXSWF\ DV ZHOO DV WLPH WR EDQNUXSWF\ DUH PDWKHPDWLFDOO\ GHULYHG MXVW DV WKH\ ZRXOG EH IRU WKH JDPEOHUVn JDPH 6FRWW f DQG f DWWHPSWHG WR LPSURYH RQ WKLV VLPSOH PRGHO 6FRWWnV HDUO\ PRGHOV DVVXPHG WKDW D FRPSDQ\ KDV D SRWHQWLDOO\ LQILQLWH OLIH DQG FDQ PHHW ORVVHV E\ VHOOLQJ GHEW RU HTXLW\ LQ DQ HIILFLHQW PDUNHW ZLWKRXW LQFXUULQJ IORWDWLRQ FRVWV 7KH\ IXUWKHU DVVXPHG WKDW WKH VHFRQGDU\ PDUNHW IRU UHDO DVVHWV LV LPSHUIHFW DQG WKDW D ILUP EHJLQV

PAGE 18

ZLWK DQ RSWLPDO OHYHO RI DVVHWV 7KHUHIRUH LW ZRXOG PXFK UDWKHU VHOO VHFXULWLHV DQG GHEW WKDQ OLTXLGDWH DVVHWV WR FRYHU LWV ORVVHV 6FRWW WKHQ VKRZHG WKDW D FRPSDQ\ ZRXOG UHPDLQ VROYHQW DV ORQJ DV VWRFNKROGHU ZHDOWK PHDVXUHG E\ PDUNHW YDOXH UHPDLQHG SRVLWLYH 6FRWW f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n PLVWDNHV (PSLULFDO :RUNV RQ %DQNUXSWF\ %HDYHUnV SDSHU ZDV WKH ILUVW HPSLULFDO ZRUN WKDW WULHG WR EXLOG D SUHGLFWLYH PRGHO RI EDQNUXSWF\ +H ORRNHG DW DFFRXQWLQJ UDWLRV ZKLFK FRXOG EH XVHG WR SUHGLFW

PAGE 19

EDQNUXSWF\ DQG IRU HDFK UDWLR KH GHULYHG D FXWRII SRLQW IRU EDQNUXSWF\ +H FRQFOXGHG WKDW WKUHH UDWLRV ZHUH WKH EHVW SUHGLFWRUV RI ILQDQFLDO IDLOXUH &DVK )ORZ7RWDO $VVHWV 1HW ,QFRPH7RWDO 'HEW DQG &DVK )ORZ7RWDO 'HEW $OWPDQ +DOGHPDQ DQG 1DUD\DQDQ f D IROORZXS RI $OWPDQ f XVHG WKH PRUH FRPSOH[ PXOWLYDULDWH GLVFULPLQDQW DQDO\VLV DSSURDFK WR EXLOG D SUHGLFWLYH PRGHO 7KHLU ZRUN LQFOXGHG DOO LQGXVWULDO IDLOXUHV IURP WR ZLWK DW OHDVW PLOOLRQ LQ DVVHWV ZKLFK PDGH D VDPSOH RI EDQNUXSW ILUPV DQG $OWPDQ HW DO IRXQG D PDWFKLQJ VDPSOH RI QRQEDQNUXSW ILUPV 7KH VDPSOHV ZHUH PDWFKHG E\ LQGXVWU\ \HDU RI EDQNUXSWF\ DQG VL]H RI DVVHWV 7KHLU PRGHO LQFOXGHG VHYHQ YDULDEOHV UHWXUQ RQ DVVHWV 52$f VWDELOLW\ RI HDUQLQJV GHEW VHUYLFH WLPHVLQWHUHVWHDUQHG RU 7,(f FXPXODWLYH SURILWDELOLW\ FXUUHQW DVVHWVOLDELOLWLHV UDWLR FDSLWDOL]DWLRQ DQG VL]H $IWHU XVLQJ YDULRXV VWDWLVWLFDO WHFKQLTXHV $OWPDQ HW DO GHULYHG D YDOXH =(7$ DV WKH FXWRII IRU EDQNUXSWF\ 7KLV PRGHO FRPPRQO\ NQRZQ DV WKH =(7$ PRGHO LV KLJKO\ DFFXUDWH HVSHFLDOO\ ZKHQ EDQNUXSWFLHV DUH QHDU 7RGD\ LW LV WKH OHDGLQJ PRGHO IRU SUHGLFWLQJ EDQNUXSWF\ DQG EHFDXVH PDQ\ ILQDQFLDO LQVWLWXWLRQV XVH LW LW KDV EHFRPH DQ LQGXVWU\ VWDQGDUG 2KOVRQ f WRRN DQRWKHU DSSURDFK WR EDQNUXSWF\ SUHGLFWLRQ E\ XVLQJ ORJLVW DQDO\VLV WR EXLOG KLV PRGHO +LV W VDPSOH LQFOXGHG IDLOHG ILUPV EXW KH GLG QRW ILQG D PDWFKLQJ VDPSOH E\ DVVHW VL]H +HQFH DPRQJ KLV QLQH

PAGE 20

YDULDEOHV VL]H EHFDPH WKH PRVW VLJQLILFDQW RQH 6LQFH KLV PRGHO KDG HUURU UDWHV RI b IRU QRQEDQNUXSW W\SH HUURUf DQG b IRU EDQNUXSW ILUPV W\SH ,, HUURUf HYHQ MXVW RQH \HDU EHIRUH EDQNUXSWF\ LW KDV UHPDLQHG PRUHRUOHVV DQ DFDGHPLF FXULRVLW\ DQG KDV QRW DWWDLQHG WKH VDPH ZLGHVSUHDG XVH DV $OWPDQnV =(7$ PRGHO =DYJUHQ f H[WHQGHG 2KOVRQnV ZRUN E\ LQFOXGLQJ PRUH YDULDEOHV DQG H[WHQGLQJ WKH OHQJWK RI WKH VWXG\ +HU ZRUN ORRNHG IRU WKH LPSRUWDQW IDFWRUV LQ WKH VKRUW DQG ORQJWHUP SUHGLFWLRQV RI EDQNUXSWF\ =DYJUHQ IRXQG WKDW SURILWDELOLW\ ZDV QRW VLJQLILFDQW LQ HLWKHU WKH VKRUW RU ORQJUXQ 5DWKHU KHU VWXG\ VKRZHG WKDW WKH DELOLW\ WR PHHW REOLJDWLRQV LV VLJQLILFDQW LQ WKH VKRUWUXQ ZKLOH HIILFLHQF\ UDWLRV DQG OLTXLGLW\ DUH LPSRUWDQW LQ WKH ORQJUXQ =DYJUHQnV VWXG\ WKHQ UHGXFHV D FRPSDQ\nV EDQNUXSWF\ ULVN WR WZR LVVXHV WKDW RI VKRUWWHUP HQGXUDQFH DV PHDVXUHG E\ WKH DELOLW\ WR PHHW REOLJDWLRQVf DQG IXQGDPHQWDO FKDUDFWHULVWLFV DV PHDVXUHG E\ WKH HIILFLHQF\ UDWLR DQG EDVLF OLTXLGLW\f 6LJQLILFDQFH RI WKLV 'LVVHUWDWLRQ :KLOH WKH H[LVWLQJ OLWHUDWXUH LV DOUHDG\ TXLWH DGYDQFHG ZH EHOLHYH D IHZ DUHDV KDYH EHHQ OHIW XQH[DPLQHG )LUVW ZKLOH ZRUN KDV EHHQ GRQH RQ ERQG UDWLQJ FKDQJHV WKHUH KDV EHHQ QR UHVHDUFK WKDW WULHV WR OLQN EDQNUXSWF\ ZLWK ILQDQFLDO PDUNHW UHDFWLRQV 1R RQH WR GDWH KDV ORRNHG DW WKH WUHQG WKDW

PAGE 21

ERQG ULVN SUHPLD WDNH DV D FRPSDQ\ DSSURDFKHV EDQNUXSWF\ HYHQ WKRXJK EDQNUXSWF\ ULVN LQ WKHRU\ VKRXOG EH D SULPDU\ ULVN LQFOXGHG LQ ULVN SUHPLD )XUWKHU ZKLOH UHVHDUFKHUV KDYH VWXGLHG ERQG UDWLQJ FKDQJHVn HIIHFWV RQ ERWK WKH VWRFN DQG ERQG PDUNHWV WR VHH LI VXFK UDWLQJ FKDQJHV FRQWDLQHG QHZ LQIRUPDWLRQ QR VWXG\ KDV OLQNHG WKH LQIRUPDWLRQ WKHVH GRZQJUDGHV SURYLGH ZLWK D FRPSDQ\nV ULVN RI ILOLQJ IRU &KDSWHU DQG WKXV GHFODULQJ EDQNUXSWF\ 6LQFH ERQG GRZQJUDGHV DUH PHDQW WR ZDUQ LQYHVWRUV RI SRVVLEOH GHIDXOW DQG EDQNUXSWF\ ZKHWKHU VXFK GRZQJUDGHV KDYH DQ\ LPSDFW RQ ILQDQFLDO PDUNHWV VKRXOG EH GLUHFWO\ OLQNHG WR WKH PDUNHWnV DVVHVVPHQW RI D FRPSDQ\nV EDQNUXSWF\ ULVN 6R IDU KRZHYHU UHVHDUFK KDV OHIW WKLV DUHD XQWRXFKHG 6HFRQG DV ZH VWDWHG LQ WKH SUHYLRXV &KDSWHU FXUUHQW EDQNUXSWF\ SUHGLFWLRQ PRGHOV KDYH ORRNHG PRVWO\ DW FXUUHQW ILQDQFLDO GDWD 2QO\ =DYJUHQ f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

PAGE 22

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f§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

PAGE 23

&+$37(5 0(7+2'2/2*< 72 678'< ),1$1&,$/ 0$5.(76 $1' %$1.5837&< 2XU UHVHDUFK RQ ILQDQFLDO PDUNHWV DQG EDQNUXSWF\ HQFRPSDVVHV WZR WRSLFV )LUVW ZH ORRN DW WKH WUHQG RI WKH ULVN SUHPLD RI FRUSRUDWH ERQGV RI FRPSDQLHV WKDW ODWHU ZHQW EDQNUXSW 7KLV WUHQG WHOOV XV LI WKH PDUNHWV FDQ FRUUHFWO\ DVVHVV LQFUHDVLQJ FKDQFHV RI EDQNUXSWF\ DQG GHIDXOW DV EDQNUXSWF\ QHDUV 7KHQ ZH H[DPLQH WKH LPSDFW RI ERQG GRZQJUDGHV RQ WKH UHWXUQV RI D FRPSDQ\nV VWRFN :H GHWHUPLQH LI UDWLQJV SOD\ D VLJQLILFDQW UROH LQ SURYLGLQJ WKH PDUNHW ZLWK QHZ LQIRUPDWLRQ 7R VWXG\ ZKHWKHU ERQG ULVN SUHPLD LQFUHDVH DV D FRPSDQ\ DSSURDFKHV EDQNUXSWF\ ZH VHOHFWHG D VDPSOH RI ERQGV EDVHG RQ WZR FULWHULD )LUVW WKH\ KDG WR EH SXEOLFO\ WUDGHG ERQGV OLVWHG LQ WKH 6WDQGDUG t 3RRUnV %RQG *XLGH ZLWK D ERQG UDWLQJ IURP HLWKHU 6t3nV RU 0RRG\nV 7KH VHFRQG FULWHULRQ ZDV WKDW WKH FRPSDQLHV ZKLFK LVVXHG WKH ERQGV ODWHU ILOHG IRU &KDSWHU EHWZHHQ 6HSWHPEHU DQG 2FWREHU 2QO\ FRUSRUDWH ERQGV KDG ERQG UDWLQJV DQG RWKHU DYDLODEOH GDWD IURP 6t3nV RU 0RRG\nV DGHTXDWH IRU RXU SXUSRVHV 7KHVH ERQGV OLVWHG LQ 7DEOH DUH XVHG WR VWXG\ ULVN SUHPLD DQG EDQNUXSWF\ 7KHQ

PAGE 24

IURP WKH $QDO\WLFDO 5HFRUG RI %RQG
PAGE 25

GLVFXVVHG LQ )DPD f DQG 0DVXOLV f LV XVHG DV WKH EDVLV IRU WKH VWDWLVWLFDO VWXGLHV 7KLV PRGHO XVHV WKH PHDQ UHWXUQV RQ DQ LQGLYLGXDO VWRFN RYHU D UHSUHVHQWDWLYH SHULRG RI WLPH EHIRUH WKH HYHQW SHULRG WR HVWLPDWH D VWRFNn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
PAGE 26

ZLWK W EHLQJ WKH GDWH LQ WKH FRPSDULVRQ SHULRG W ZKHUH 5-W LV WKH UHDOL]HG GDLO\ UHWXUQ RI WKH VWRFN M DW WLPH W ZKLFK ZDV UHDG IURP WKH &563 'DLO\ 5HWXUQ 7DSH 7KH DYHUDJH GLVWXUEDQFH WHUP IRU 1 HYHQWV ILUPVf LV DYJ (W 1 r e (MWf ZLWK M 1 7KH QXOO K\SRWKHVLV DYJ (W HTXDOV ]HUR PHDQV WKDW D GRZQZDUG ERQG UDWLQJ FKDQJH KDV QR HIIHFW RQ VKDUHKROGHUVn GDLO\ UHWXUQV 6LQFH ZH EHOLHYH WKDW WKH ILUVW GRZQJUDGH RI D ERQG JLYHV WKH HDUOLHVW VLJQDO DQG KHQFH WKH PRVW LQIRUPDWLRQ WR VWRFNKROGHUV DERXW WKH ULVNV RI ILQDQFLDO GLVWUHVV ZH H[SHFW WKH QXOO K\SRWKHVLV WR EH UHMHFWHG IRU RXU &KDSWHU VDPSOH &RQYHUVHO\ ZH H[SHFW QRW WR EH DEOH WR UHMHFW WKH DERYH QXOO K\SRWKHVLV IRU RXU PDWFKLQJ VDPSOH RI ILUPV WKDW GLG QRW ILOH IRU &KDSWHU 7KH YDULDQFH RI DYJ (W LV 9DU (f f r e DYJ (W ;(ff ZKHUH ;(f f r ( DYJ (Wf :KHQ W ZH ZRXOG EH WHVWLQJ IRU WKH GLVWXUEDQFH RQ HYHQW GDWH 7KH WVWDWLVWLF XVHG WR GHWHUPLQH ZKHWKHU DYJ ( GLIIHUV VLJQLILFDQWO\ IURP ]HUR ZLWK GHJUHH RI IUHHGRP LV W DYJ ( 9DU (f f 7KH FXPXODWLYH HUURU RYHU D SDUWLFXODU HYHQW WLPH LQWHUYDO LV &(DEf ( DYJ (Wf ZKHUH D E

PAGE 27

:H K\SRWKHVL]H WKDW WKH FXPXODWLYH HUURUV LQ WKH EDQNUXSW DQG PDWFKLQJ JURXSV DUH VWDWLVWLFDOO\ GLIIHUHQW LPSO\LQJ GLIIHUHQW HIIHFWV RI ERQG UDWLQJ FKDQJHV RQ VWRFN UDWHV RI UHWXUQ )LQDOO\ ZH VHSDUDWH WKH &KDSWHU JURXS LQWR WZR VXEn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

PAGE 28

1RWH $ FRPSDQ\ E\ RXU GHILQLWLRQ LV OLTXLGDWHG LI LW VDWLVILHV RQH RU PRUH RI WKHVH FRQGLWLRQV DFTXLUHG E\ DQRWKHU FRPSDQ\ RU OLTXLGDWHG OLVWHG RQ &203867$7 DV EDQNUXSW QR ORQJHU RQ WKH :DOO 6WUHHW -RXUQDO ,QGH[ QR ORQJHU RQ WKH 3UHGLFDVWV )t6 ,QGH[ RI &RUSRUDWH &KDQJH QR ORQJHU RQ 4ILOH QR ORQJHU RQ WKH 'LUHFWRU\ RI &RUSRUDWH $IILOLDWLRQV QR ORQJHU OLVWHG RQ WKH &' 6\VWHP QR ORQJHU GHVFULEHG LQ WKH &KDSWHU 5HSRUW

PAGE 29

7DEOH %RQG 6DPSOH IRU 5LVN 3UHPLD 6WXGLHV &RPSDQ\ 1DPH %RQG 1DPH &K ,QG $OOHJKHQ\ ,QWnO 6XE 6) 'HE b n D $OOLV&KDOPHUV 0IJf &RUS 6) 'HE b n $PDUH[ ,QF 6XE 6) 'HE b n D $PHU +HDOWK &DUH 0JPW 6XE 1W b n $QJOR &R ,QF *WG 6XE 6) 'HE b n H $UJR 3HWUROHXP 6XE 6) 'HE b n D %DOGZLQ8QLWHG &RUS 6XE 6) 'HE b n I %$6,; &RUS 6XE 'HE b n %HNHU ,QGXV 6HU 6XE 6) 'HE b n D %UDQLII $LUZD\V 6) 'HE b n %XWWHV *DV t 2LO 'HE b n D &KDUWHU &R 6XE 'HE b n G &KHPHWURQ &RUS 6) 'HE b n E &ROHFR ,QGXV 6XE 6) 'HE b n 2K &U\VWDO 2LO &R 6XE 6) 'HE b n E (PRQV ,QGXV ,QF (J 7U &WIV 6HU b n )LUVW 5HSXEOLF %DQN 6) 'HE b n D )RRG )DLU 6WRUHVf 6) 'HE b n F )6& &RUS 6U 6) 'HE b n *DPEOHV &UHGLW 6U 1RWHV b n H *OREDO 0DULQH ,QF 6U 6XE 'HE b n H +DUGZLFNH &RnV ,QF 6XE 6) 'HE b n J ,QIRUH[ ,QF 6XE 'HE b n E ,WHO &RUS ;V 6XE 'HE b n I -RKQV0DQYLOOH 6) 'HE b n K -RQHV t /DXJKOLQ 6WHHO 6XE 'HE b n D /LRQHO &RUS 6XE 6) 'HE b n 2K /79 &RUS ;V 6) 'HE b n 0FOHDQ ,QGXV 6XE 'HE b n 0HJR ,QWnO 6XE 6) 'HE b n 2K 0LFKLJDQ *HQHUDO ;V 6U 6) 'HE b n D 0LVVLRQ ,QV *URXS 6) 'HE b n

PAGE 30

7DEOH f§FRQWLQXHG 0RUWRQ 6KRH &RV ,QF 6U 6) 'HE b n E 1RUWK $PHULFDQ &DU &RUS (TXLS 7U b n E 3HWWLERQH &RUS ;V 6XE 6) 'HE b n E 3XEOLF 6HUYLFH 1HZ +DPS VW 9 b n D 5DGLFH &RUS 6XE 6) 'HE b n F 6KDURQ 6WHHO 6XE 6) 'HE b n E 6PLWK ,QWnO 6) 'HE b n H 6WRUDJH 7HFKQRORJ\ 1WV b n E 6XQEHDP &RUS 6) 'HE b n G 7H[DFR &DSLWDO ([WnG 1W b n 7H[DFR ,QF 'HE b n F 7H[DV ,QWL $LUOLQHV ;V 6XE 'HE b n 7RGG 6KLS\DUGV 6U 6XE 1WV b n :HGWHFK &RUS 6U 6XE 1WV b n E :HVWHUQ &R 1R $PHULFD [V 6XE 'HE b n H :KLWH 0RWRU &RUS 6) 'HE b n D :LFNHV &RUS 6) 'HE b n G :LOVRQ )RRG 'HE b n H &K LQGLFDWHV WKH GDWH RQ ZKLFK WKH FRPSDQ\ ILOHG IRU &KDSWHU ,QG LQGLFDWHV WKH LQGXVWU\ FODVVLILFDWLRQ RI WKH FRPSDQ\ E\ 6WDQGDUG t 3RRUnV %RQG *XLGH 6RXUFH 6WDQGDUG t 3RRUnV %RQG *XLGH WR

PAGE 31

7DEOH 'DWH RI )LUVW %RQG 'RZQJUDGH DQG RI )LOLQJ IRU &KDSWHU &RPSDQ\ 1DPH %RQG 1DPH %HI $IWHU'DWH &K $KHDG $OOHJKHQ\ ,QWnO 6XE 6) 'HE b n %% % $OOLV&KDOPHUV 0IJf &RUS 6) 'HE b n $ %%% $PDUH[ ,QF 6XE 6) 'HE b n % &&& $PHU +HDOWK &DUH 0JPW 6XE 1W b n % $QJOR &R ,QF *WG 6XE 6) 'HE b n % % $UJR 3HWUROHXP 6XE 6) 'HE b n % &&& %DOGZLQ8QLWHG &RUS 6XE 6) 'HE b n %% %% %$6,; &RUS 6XE 'HE b n % &&& %HNHU ,QGXV 6HU 6XE 6) 'HE b n&&& %UDQLII $LUZD\V 6) 'HE b n %%% %% %XWWHV *DV t 2LO 'HE b n % % &KDUWHU &R 'XE 'HE b n % % &KHPHWURQ &RUS 6) 'HE b n %%% %%% &ROHFR ,QGXV 6XE 6) 'HE b n % % &U\VWDO 2LO &R 6XE 6) 'HE b n % % (PRQV ,QGXV ,QF (T 7U &WIV 6HU b n % &&& )LUVW 5HSXEOLF %DQN 6) 'HE b n $$ $ )RRG )DLU 6WRUHVf 6) 'HE b n %% )6& &RUS 6U 6) 'HE b n % &&& *DPEOHV &UHGLW 6U 1RWHV b n $ %%% *OREDO 0DULQH ,QF 6U 6XE 'HE b n % % +DUGZLFNH &RnV ,QF 6XE 6I'HE b n FFF ,QIRUH[ ,QF 6XE 'HE b n % ,WHO &RUS 6XE 'HE b n %% % -RKQV0DQYLOOH 6) 'HE b n $$ $ -RQHV t /DXJKOLQ 6WHHO 6XE 'HE b n %% % /LRQHO &RUS 6XE 6) 'HE b n % /79 &RUS 6) 'HE b n %% % 0FOHDQ ,QGXV 6XE 'HE b n % 0HJR ,QWnO 6XE 6) 'HE b n % &&& 0LFKLJDQ *HQHUDO ;V 6U 6) 'HE b n &&& & 0LVVLRQ ,QV *URXS 6) GHE b n $ %%

PAGE 32

7DEOH f§FRQWLQXHG 0RUWRQ 6KRH &RV ,QF 6U 6) 'HE b n % 1RUWK $PHULFDQ &DU &RUS (T 7U b n %%% % 3HWWLERQH &RUS ;V 6XE 6) 'HE b n %% % 3XEOLF 6HUYLFH 1HZ +DPS VW 9 b n %%% %%% 5DGLFH &RUS 6XE 6) 'HE b n % 6KDURQ 6WHHO 6XE 6) 'HE b n %% % 6PLWK ,QWnO 6) 'HE b n $ %%% 6WRUDJH 7HFKQRORJ\ 1WV b n %%% %% 6XQEHDP &RUS 6) 'HE b n $ %%% 7H[DFR &DSLWDO ([WnG 1W b n $$ $ 7H[DFR ,QF 'HE b n $$ $$ 7H[DV ,QWL $LUOLQHV 6XE 'HE b n % &&& 7RGG 6KLS\DUGV 6U 6XE 1WV b n % FF :HGWHFK &RUS 6U 6XE 1WV b n % FF :HVWHUQ &R 1R $PHULFD 6XE 'HE b n %% % :KLWH 0RWRU &RUS 6) 'HE b n % FFF :LFNHV &RUS 6) 'HE b n %%% %% :LOVRQ )RRG 'HE b n %% % %HI DQG $IWHU DUH WKH ERQG UDWLQJV EHIRUH DQG DIWHU WKH ILUVW ERQG GRZQJUDGH 'DWH LV WKH \HDU DQG PRQWK RI WKH ILUVW ERQG GRZQJUDGH &K LV WKH \HDU DQG PRQWK LQ ZKLFK WKH FRPSDQ\ ILOHG IRU &KDSWHU $KHDG LV WKH QXPEHU RI PRQWKV E\ ZKLFK WKH ERQG GRZQJUDGH SUHFHGHG WKH ILOLQJ RI &KDSWHU 6RXUFH 6WDQGDUG t 3RRUnV %RQG *XLGH

PAGE 33

&+$37(5 5(68/76 21 ),1$1&,$/ 0$5.(76 $1' %$1.5837&< )RU RXU UHVHDUFK RQ ULVN SUHPLD DQG EDQNUXSWF\ ZH FRPSXWHG DQG GUHZ ULVN SUHPLXP FXUYHV IRU HDFK RI RXU FRPSDQLHVn ERQGV 7KHVH FXUYHV DUH SUHVHQWHG LQ WKH $SSHQGL[f $ VROLG OLQH LQGLFDWHV D ERQG UDWLQJ FKDQJH E\ 6t3 D GRWWHG OLQH RQH E\ 0RRG\nV 7KH DUURZV RQ WKH JUDSKV LQGLFDWH D ERQG XSJUDGH ,Q b RI RXU VDPSOH RU RI WKH ULVN SUHPLXP FXUYHV WKHUH LV D GHILQLWH XSZDUG WUHQG ULVN SUHPLD ULVH VWHDGLO\ DV WKH FRPSDQ\ DSSURDFKHV EDQNUXSWF\ 1LQH bf KDYH DPELJXRXV IOXFWXDWLQJ WUHQGV DQG RQO\ RQH RU b RI RXU VDPSOH VKRZV D GRZQZDUG WUHQG ZKHUH WKH ULVN SUHPLXP RQ LWV ERQG DFWXDOO\ IHOO DV LW PRYHG WRZDUGV EDQNUXSWF\ +HQFH RXU GDWD VXJJHVW WKDW WKH PDUNHW JHQHUDOO\ LQFRUSRUDWHV WKH OLNHOLKRRG RI EDQNUXSWF\ DGHTXDWHO\ LQ LWV RYHUDOO DVVHVVPHQW RI ULVN DQG LQFOXGHV WKLV SDUWLFXODU ULVN LQ WKH YDOXDWLRQ RI VHFXULWLHV :KHQ ZH FRPSDUH WKH UHODWLRQVKLS EHWZHHQ ERQG UDWLQJ FKDQJHV DQG ULVN SUHPLD FXUYHV ZH ILQG WKDW IRU PRVW FDVHV ZKHUH WKHUH ZDV DQ REYLRXVO\ LQFUHDVLQJ ULVN SUHPLXP WKH

PAGE 34

UDWLQJ GRZQJUDGH FDPH DIWHU WKH ULVN SUHPLXP KDG EHJXQ WR LQFUHDVH +HQFH LQ WKH VKRUWWHUP D ODJ RI RQH WR VHYHQ PRQWKV GRHV H[LVW 7KLV FRQFOXVLRQ FRQILUPV WKDW RI :HLQVWHLQ f ZKR IRXQG WKDW PDUNHWV KDG DQWLFLSDWHG ERQG UDWLQJ FKDQJHV WR VL[ PRQWKV EHIRUH WKH FKDQJH ,W KRZHYHU FRQIOLFWV ZLWK .DW] f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nV FDQQRW SUHGLFW ULVN SUHPLXP FKDQJHV RU DQ\ XSFRPLQJ GHIDXOW :H EHOLHYH KRZHYHU WKDW WKH ODJ SKHQRPHQRQ LV QRW SDUWLFXODUO\ PHDQLQJIXO )RU H[DPSOH D ERQG GRZQJUDGH WKDW WUDLOV DQ LQFUHDVH LQ WKH ULVN SUHPLXP PD\ OHDG D ODWHU LQFUHDVH 6LQFH ERQG UDWLQJ FKDQJHV VKRXOG ZDUQ LQYHVWRUV RI FRPLQJ EDQNUXSWF\ DQG GHIDXOW ZKDW UHDOO\ PDWWHUV LQ WKH ORQJUXQ LV ZKHWKHU D ERQG UDWLQJ SURYLGHV QHZ LQIRUPDWLRQ WR WKH PDUNHW RQ FRPSDQLHV WKDW ODWHU ZHQW EDQNUXSW 7KDW FDQ EH GHWHUPLQHG RQO\ E\ VHHLQJ LI ERQG

PAGE 35

UDWLQJ FKDQJHV KDYH DQ\ LPSDFW RQ WKH PDUNHW $VVXPLQJ WKDW WKH ILUVW ERQG GRZQJUDGH SURYLGHV WKH PDUNHW ZLWK WKH JUHDWHVW DPRXQW RI QHZ LQIRUPDWLRQ DERXW FKDQJHV LQ ULVN ZH PXVW IXUWKHU VWXG\ WKH LPSDFW RI WKH ILUVW ERQG GRZQJUDGH RQ WKH PDUNHW 2XU GDWD FDQ EH XVHG WR FRPSDUH WKH WZR UDWLQJ DJHQFLHV 6t3 DQG 0RRG\nV WR VHH ZKLFK ZDUQV LQYHVWRUV HDUOLHU 7DEOH JLYHV WKH UHOHYDQW LQIRUPDWLRQ 2Q DYHUDJH 6t3 JDYH WKH ILUVW ERQG GRZQJUDGH PRQWKV EHIRUH D FRPSDQ\ ILOHG IRU &KDSWHU ZKLOH 0RRG\nV JDYH WKH ILUVW FRQVLVWHQW ERQG GRZQJUDGH PRQWKV RU PRQWKV DIWHU 6t3 EHIRUH D FRPSDQ\ ILOHG IRU &KDSWHU 2XW RI D VDPSOH RI FRPSDQLHV WKHUH ZHUH ILYH ERQGV WKDW UHFHLYHG QR GRZQJUDGHV ZKDWVRHYHU IURP 6t3nV DQG VL[ WKDW UHFHLYHG QR GRZQJUDGHV ZKDWVRHYHU IURP 0RRG\nV EHIRUH WKH FRPSDQLHV ZKLFK LVVXHG WKHP ILOHG IRU &KDSWHU EDQNUXSWF\ $OWKRXJK WKHUH ZDV QR UDWLQJ FKDQJH WKH PDUNHWV GLG IRUHVHH WKH FRPLQJ EDQNUXSWF\ DQG WKH ULVN SUHPLD RQ WKHVH ERQGV EHJDQ LQFUHDVLQJ DQ DYHUDJH RI WKUHH WR ILYH PRQWKV EHIRUH &KDSWHU )LQDOO\ VL[ FRPSDQLHV ZKLFK 6t3nV UDWHG ZHUH QRW UDWHG DW DOO E\ 0RRG\nV +HQFH LW ZRXOG VHHP WKDW LQ WKLV VSHFLILF LQVWDQFH 6t3 SURYLGHG PRUH WLPHO\ DQG PRUH FRPSOHWH ERQG UDWLQJV DQG FKDQJHV WKDQ 0RRG\nV LQ WKH VDPSOH DQG IRU WKH SHULRG ZH VWXGLHG :H HPSKDVL]H WKDW WKLV LV D RQHVLGHG WHVW RQO\ DQG WKDW LW E\ QR PHDQV FDQ EH WKRXJKW RI DV D GHILQLWLYH FRQFOXVLRQ

PAGE 36

$V GHVFULEHG HDUOLHU LQ &KDSWHU ZH GLYLGHG WKH VDPSOH RI FRPSDQLHV ZKLFK ILOHG &KDSWHU LQWR WKRVH WKDW ODWHU UHn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nV IRUPXOD IRU WKH WKUHH SHULRGV WR PRQWKV EHIRUH EDQNUXSWF\ WR PRQWKV EHIRUH EDQNUXSWF\ DQG WRO PRQWKV EHIRUH EDQNUXSWF\f KRZHYHU ZH IRXQG WKDW WKLV ZDV QRW VR :H KDG WHVWHG IRU WKH QXOO K\SRWKHVLV WKDW WKH PHDQV EHWZHHQ WKH WZR JURXSV GXULQJ HDFK PRQWK ZHUH HTXDO 'XULQJ WKH ILUVW SHULRG RI WR PRQWKV EHIRUH EDQNUXSWF\ RXW RI bf W WHVWV ZHUH VLJQLILFDQW DQG DOO W YDOXHV ZHUH QHJDWLYH %\ RXU GHILQLWLRQ IRU W WKLV PHDQV WKDW WKH PHDQ YDOXHV RI ULVN SUHPLD IRU WKH ODWHU OLTXLGDWHG ILUPV DUH ORZHU WKDQ WKRVH IRU ODWHU UHRUJDQL]HG ILUPV 'XULQJ WKH VHFRQG SHULRG KRZHYHU RXW RI bf WWHVWV ZHUH VLJQLILFDQW DQG RI WKRVH WWHVWV ZHUH SRVLWLYH 7KLV PHDQV FRQWUDU\ WR ZKDW PLJKW

PAGE 37

EH GLVFHUQHG IURP WKH JUDSK WKDW WKH PHDQ YDOXHV RI ULVN SUHPLD IRU ODWHU OLTXLGDWHG DQG ODWHU UHRUJDQL]HG ILUPV ZHUH GLIIHUHQW DQG WKDW WKH PHDQ ULVN SUHPLD RI WKH ODWHU OLTXLGDWHG ILUPV ZHUH KLJKHU +HQFH DOWKRXJK IURP MXVW ORRNLQJ DW D JUDSK ZH KDG WKRXJKW WKDW WKH GLIIHUHQFHV ZKLFK ZHUH FOHDUO\ YLVLEOH ZRXOG QRW EH VWDWLVWLFDOO\ VLJQLILFDQW WKH WWHVWV VKRZ WKDW WKLV LV QRW VR )LQDOO\ GXULQJ WKH WKLUG SHULRG RXW RI bf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
PAGE 38

HTXLYDOHQW GRZQJUDGHV GXULQJ WKH VDPH WLPH SHULRG EXW GLG QRW ODWHU ILOH &KDSWHU 7KH FRPSDQLHV LQ WKH WZR VDPSOHV DUH OLVWHG LQ 7DEOHV DQG 7DEOHV WKURXJK JLYH WKH VWDWLVWLFDO UHVXOWV RI RXU HYHQW VWXG\ ,Q 7DEOH WKH WVWDWLVWLF IRU WHVWLQJ WKH QXOO K\SRWKHVLV WKDW WKH DYHUDJH GLVWXUEDQFHV DW HYHQW GD\ (f HTXDOV ]HUR IRU WKH &KDSWHU JURXS LV ZLWK GHJUHH RI IUHHGRP 7KLV WVWDWLVWLF LV VLJQLILFDQW DW WKH b OHYHO 7KXV DV H[SHFWHG LQ WKH 5HVHDUFK 'HVLJQ DQG 0HWKRGRORJ\ VHFWLRQ WKH QXOO K\SRWKHVLV WKDW WKH DYHUDJH GLVWXUEDQFH DW WKH HYHQW GD\ IRU WKH &KDSWHU JURXS LV ]HUR FDQ EH VWURQJO\ UHMHFWHG )RU WKH PDWFKLQJ VDPSOH KRZHYHU WKH WVWDWLVWLF IRU WHVWLQJ (f HTXDOV ]HUR LV ZKLFK LV VLJQLILFDQW RQO\ DW D PDUJLQDO OHYHO 7KH QHJDWLYH VLJQV RQ ERWK WKH WVWDWLVWLFV PHDQ WKDW ERQG GRZQJUDGHV QHJDWLYHO\ DIIHFW VKDUHKROGHU ZHDOWK 7R VHH LI WKH PHDQV LQ WKH &KDSWHU DQG PDWFKLQJ VDPSOHV DUH HTXDO ZH HPSOR\ D VSHFLDO WWHVW VHH 0RQWJRPHU\ff ZKLFK LV VXLWDEOH IRU FDVHV ZKHQ ZH FDQQRW DVVXPH HTXDO YDULDQFHV DQG ZKHQ WKH QXPEHU RI REVHUYDWLRQV LV OHVV WKDQ 7KH WVWDWLVWLF IRU WHVWLQJ WKH HTXDOLW\ RI WZR PHDQV LV ZLWK GHJUHH RI IUHHGRP ,W LV VLJQLILFDQW DW WKH b OHYHO VR WKH QXOO K\SRWKHVLV WKDW WKH WZR PHDQV DUH HTXDO LV UHMHFWHG 7KLV PHDQV WKDW WKH GDLO\ UDWHV RI UHWXUQ RI FRPSDQLHV LQ WKH &KDSWHU VDPSOH VXIIHUHG PRUH ZLWK ERQG GRZQJUDGHV WKDQ WKRVH LQ WKH PDWFKLQJ VDPSOH

PAGE 39

7KH YDULDQFHV RI WKH WZR JURXSV DV JLYHQ E\ WKH )WHVW ZHUH QRW VLJQLILFDQWO\ GLIIHUHQW 7DEOH VKRZV WKH VWDWLVWLFDO DQDO\VLV RI WKH WZR VXEn JURXSV LQ WKH &KDSWHU VDPSOH WKH FRPSDQLHV ZKLFK UHn RUJDQL]HG DQG WKRVH WKDW ZHUH OLTXLGDWHG 7KH WVWDWLVWLFV DUH IRU FRPSDQLHV ZKLFK ODWHU UHRUJDQL]HG DQG IRU FRPSDQLHV ZKLFK GLG QRW 6LQFH ERWK DUH VLJQLILFDQW WKH QXOO K\SRWKHVHV DUH UHMHFWHG 8VLQJ 0RQWJRPHU\n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

PAGE 40

WKRVH RI SUHYLRXV ZRUNV PHQWLRQHG LQ &KDSWHU :H EHOLHYH KRZHYHU WKDW WKH LVVXH LV VWLOO YHU\ PXFK XQUHVROYHG DV FDQ EH VHHQ E\ WKH VHHPLQJO\ FRQWUDGLFWRU\ UHVXOWV RI *ULIILQ DQG 6DQYLQFHQWH f 7KHUHIRUH RXU ZRUN VKRXOG FRQWULEXWH WR WKH RQJRLQJ GHEDWH

PAGE 41

1RWHV 2XU WZR FULWHULD IRU VHOHFWLQJ FRPSDQLHV LQWR WKH &KDSWHU DQG PDWFKLQJ VDPSOHV IRU HYHQW VWXG\ ZHUH 7KH FRPSDQLHV KDG WR EH OLVWHG HLWKHU RQ WKH 1HZ
PAGE 42

7DEOH 'DWHV RI 'RZQJUDGHV E\ 6t3 DQG 0RRG\nV &RPSDUHG &RPSDQ\ 1DPH %RQG 1DPH 6t3 $KHDG 0RRGYV $KHDG 'LII $OOHJKHQ\ ,QWnO 6XE 6) 'HE b n $OOLV&KDOPHUV 0IJf &RUS 6) 'HE b n $PDUH[ ,QF 6XE 6) 'HE b n $PHU +HDOWK &DUH 0JPW 6XE 1W b n $QJOR &R ,QF *WG 6XE 6) 'HE b n $UJR 3HWUROHXP 6XE 6) 'HE b n %DOGZLQ8QLWHG &RUS 6XE 6) 'HE b n 1$ %$6,; &RUS 6XE 'HE b n %HNHU ,QGXV 6HU 6XE 6) 'HE b n %UDQLII $LUZD\V 6) 'HE b n %XWWHV *DV t 2LO 'HE b n &KDUWHU &R 'XE 'HE b n &KHPHWURQ &RUS 6) 'HE b n &ROHFR ,QGXV 6XE 6) 'HE b n &U\VWDO 2LO &R 6XE 6) 'HE b n (PRQV ,QGXV ,QF (T 7U &WIV 6HU b n )LUVW 5HSXEOLF %DQN 6) 'HE b n )RRG )DLU 6WRUHVf 6) 'HE b n 1$ )6& &RUS 6U 6) 'HE b n 1$ *DPEOHV &UHGLW 6U 1RWHV b n 1$ *OREDO 0DULQH ,QF 6U 6XE 'HE b n +DUGZLFNH &RnV ,QF 6XE 6I'HE b n 1$ ,QIRUH[ ,QF 6XE 'HE b n ,WHO &RUS 6XE 'HE b n -RKQV0DQYLOOH 6) 'HE b n -RQHV t /DXJKOLQ 6WHHO 6XE 'HE b n /LRQHO &RUS 6XE 6) 'HE b n /79 &RUS 6) 'HE b n 0FOHDQ ,QGXV 6XE 'HE b n 0HJR ,QWn 6XE 6) 'HE b n 1$ 0LFKLJDQ *HQHUDO ;V 6U 6) 'HE b n 0LVVLRQ ,QV *URXS 6) GHE b n

PAGE 43

7DEOH f§FRQWLQXHG 0RUWRQ 6KRH &RV ,QF 1RUWK $PHULFDQ &DU &RUS 3HWWLERQH &RUS 3XEOLF 6HUYLFH 1HZ +DPS 5DGLFH &RUS 6KDURQ 6WHHO 6PLWK ,QWnO 6WRUDJH 7HFKQRORJ\ 6XQEHDP &RUS 7H[DFR &DSLWDO 7H[DFR ,QF 7H[DV ,QWL $LUOLQHV 7RGG 6KLS\DUGV :HGWHFK &RUS :HVWHUQ &R 1R $PHULFD :KLWH 0RWRU &RUS :LFNHV &RUS :LOVRQ )RRG 6U 6) 'HE b n (T 7U b n ;V 6XE 6) 'HE b n VW 9 b n 6XE 6) 'HE b n 6XE 6) 'HE b n 6) 'HE b n 1WV b n 6) 'HE b n ([WnG 1W b n 'HE b n 6XE 'HE b n 6U 6XE 1WV b n 6U 6XE 1WV b n 6XE 'HE b n 6) 'HE b n 6) 'HE b n 'HE b n 6t3 DQG 0RRG\nV LQGLFDWH WKH \HDU DQG PRQWK RI WKH ILUVW FRQVLVWHQW ERQG GRZQJUDGH E\ 6t3 DQG 0RRG\nV UHVSHFWLYHO\ $KHDG LQGLFDWHV WKH QXPEHU RI PRQWKV EHIRUH EDQNUXSWF\ WKDW GRZQJUDGH WRRN SODFH 'LI LV WKH QXPEHU RI PRQWKV EHWZHHQ 6t3nV ILUVW FRQVLVWHQW GRZQJUDGH DQG EDQNUXSWF\ PLQXV WKH QXPEHU RI PRQWKV EHWZHHQ 0RRG\nV ILUVW FRQVLVWHQW GRZQJUDGH DQG EDQNUXSWF\

PAGE 44

0RQWKV EHIRUH %DQNUXSWF\ &KDSWHU f Â’ /LTXLGDWHG )LUPV 5HRUJDQL]HG )LUPV )LJXUH 5LVN 3UHPLD &XUYHV &RPSDULVRQ 5LVN SUHPLD WUHQGV IRU FRPSDQLHV WKDW ILOHG IRU &KDSWHU DQG ODWHU UHRUJDQL]HG YHUVXV WKRVH WKDW ILOHG DQG ZHUH ODWHU OLTXLGDWHG

PAGE 45

7DEOH 77HVW RI 5LVN 3UHPLD 7UHQGV 7LPH 0HDQ 0HDQ 'LI I WWHVW V b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b

PAGE 46

7DEOH f§FRQWLQXHG b b b b b b b b b b b b b b 0HDQ DQG 0HDQ DUH WKH ULVN SUHPLD PHDQV IRU WKH OLTXLGDWHG DQG ODWHU UHRUJDQL]HG ILUPV UHVSHFWLYHO\ 'LII LV WKHLU GLIIHUHQFH WWHVW LV WKH YDOXH RI WKH WWHVW DQG VO LV WKH VLJQLILFDQFH OHYHO

PAGE 47

7DEOH 6DPSOH RI %DQNUXSW &KDSWHU f &RPSDQLHV IRU (YHQW 6WXG\ 0 &RPRDQY 1DPH %RQG 1DPH &86,3 %HI $IWHU (YHQW ,QG -RKQV0DQYLOOH 6) 'HE b n $$ $ K 6KDURQ 6WHHO 6XE 6) 'HE b n %% % E %UDQLII $LUZD\V 6) 'HE b n %%% %% $OOHJKHQ\ ,QWnO 6XE 6) 'HE b n %% % D 3XEOLF 6HUYLFH 1+ VW 9 b n %%% %%% D 6PLWK ,QWnO 6) 'HE b n $ %%% H :KLWH 0RWRU &RUS 6) 'HE b n % &&& D 0HJR ,QWnO 6XE 6) 'HE b n % &&& 2K $OOLV&KDOPHUV &RUS 6) 'HE b n $ %%% ,WHO &RUS 6XE 'HE b n %% % I 6XQEHDQ &RUS 6) 'HE b n $ %%% G %DOGZLQ8QLWHG &RUS 6XE 6) 'HE b n %% %% I $UJR 3HWUROHXP 6XE 6) 'HE b n % &&& D &KDUWHU &R 'XE 'HE b n % % G %XWWHV *DV t 2LO 'HE b n % % D /79 &RUS 6) 'HE b n %% % :HGWHFK &RUS 6U 6XE 1WV b n % && E 7RGG 6KLS\DUGV 6U 6XE 1WV b n % && *OREDO 0DULQH ,QF 6U 6XE 'HE b n % % H $QJOR &R ,QF *WG 6XE 6) 'HE b n % % D &ROHFR ,QGXV 6XE 6) 'HE b n % % K :HVWHUQ &R 1R $P 6XE 'HE b n %% % H 'DWD LQ ERWK 7DEOHV DQG DUH PDWFKHG 0 LV WKH PDWFKLQJ QXPEHU &86,3 LV WKH LGHQWLILFDWLRQ QXPEHU RI WKH FRPSDQ\ RQ &203867$7 %HI $IWHU DQG (YHQW DUH WKH UDWLQJ EHIRUH DQG DIWHU DQG WKH SUHFLVH \HDU PRQWK DQG GDWH RI WKH UDWLQJ GRZQJUDGH ,QG LV WKH LQGXVWU\ FODVVLILFDWLRQ FRGH

PAGE 48

7DEOH 0DWFKLQJ 6DPSOH IRU (YHQW 6WXG\ RI %RQG 'RZQJUDGHV 0-L &RPRDQY 1DPH &180 %RQG 1DPH %HI $IWHU (YHQW ,QG %ODFN t 'HFNHU 0IJ 1RWHV b $$ $ H .H\VWRQH &RQVRO ,QGXV 6) 'HE b %% % G 0&RUS 0 6) 'HE b %%% %% D %ODLU t &R 6XE 6I 'HE b n %% % D .DQVDV *DV t (OHF VW 0WJ b %%% %%% D +XQGVRQ %D\ 0LQLQJt6PHOWLQJ 6) 'HE b $ %%% D ,QWnO +DUYHVWHU &R 6) 'HE b % &&& D )LOPZD\V 6XE 'HE b % &&& D +HLOHPDQ *f %UHZLQJ 6) 'HE b $ %%% D &KU\VOHU &RUS 6) 'HE b %% % D %RLVH &DVFDGH 1RWHV b $ %%% G 86 +RPH 1RWHV b %% %% F &DPSDQHOOL ,QGXVWULHV 6U6XE6) 'HE b n % &&& *HQHVFR 6U 6) 1RWHV b n % % =DSDWD &RUS 6XE 'HE b n % % D 0*0 *UDQG +RWHOV 6XE 6I 'HE b %% % J 2DN ,QGXV 6) 'HE b % && &DQQRQ *URXS 6U 6XE 'HE b % FF -RVHSKVRQ ,QWnO 6XE 6) 'HE b % % 0'& +ROGLQJV 6XE 1RWHV b n % % F $QJHOHV &RUS 6U 6XE 'HE b n % % &RDVWDO &RUS 6XE([ 951W b n %% % 0 LV WKH PDWFKLQJ QXPEHU IRU WKH GDWD LQ 7DEOHV DQG &180 LV WKH LGHQWLILFDWLRQ FRGH RI WKH FRPSDQ\ RQ &203867$7 ‘W WR

PAGE 49

7DEOH 7WHVW IRU &KDSWHU EDQNUXSWf DQG 0DWFKLQJ 6DPSOHV (YHQW 6WXG\ 5HVXOWV &KDSWHU 6DPSOH ( f 6WG + (f W GI VLJQLILFDQW DW bf + (f (f WWHVW G VLJQLILFDQW DW bf 0DWFKLQJ 6DPSOH (f 6WG + (f W GI VLJQLILFDQW DW bf I f

PAGE 50

7DEOH 7WHVW IRU 6XE*URXSV RI &KDSWHU &RPSDQLHV 7KRVH ZKLFK ZHUH /LTXLGDWHG DQG WKRVH ZKLFK 5HRUJDQL]HG /LFUXLGDWHG *URXS 5HRUDDQL]HG *URXS ( f 6WG (f 6WG + ( f W GI VLJQLILFDQW DW bf + (f A W GI VLJQLILFDQW DW bf + (f (f WWHVW GI f

PAGE 51

7DEOH 'LVWXUEDQFH 5HVXOWLQJ IURP %RQG 'RZQJUDGH RQ &KDSWHU %DQNUXSW 6DPSOHf DQG 6WDWLVWLFDO $QDO\VLV (YHQW DYJ (Wf &( 'D\ ,OO bf 9DULDEOH 9DOXHbf WVWDWLVWLF ( f§f r b ( f r b &(f r b &(f r b &(f r b &(f r b &(f &(f &(f &(f &(f r b 7KH WWHVW IRU &( LV W &(r1f6WG&(f r7f f

PAGE 52

7DEOH 'LVWXUEDQFH 5HVXOWLQJ IURP %RQG 'RZQJUDGH RQ 0DWFKLQJ 6DPSOH DQG 6WDWLVWLFDO $QDO\VLV (YHQW DYJ (Wf &( 'D\ bf 9DULDEOH 9DOXH bf WVWDWLVWLF ( f ( f R 0 + 2 R &(f§f &(f &(f &(f &(f r b &(f r b &(f &(f 7KH WWHVW IRU &( LV W &(r 1f6WG&(f r7f f

PAGE 53

'D\ 6XUURXQGLQJ (YHQW Â’ &KDSWHU 6DPSOH 0DWFKLQJ 6DPSOH )LJXUH 'LVWXUEDQFH 'XH WR %RQG 'RZQJUDGH 7KLV JUDSK FRPSDUHV WKH GLVWXUEDQFH FDXVHG E\ D ERQG GRZQJUDGH RQ FRPSDQLHV LQ WKH VDPSOH ZKLFK ILOHG IRU &KDSWHU DQG LQ WKH VDPSOH ZKLFK GLG QRW ILOH IRU EDQNUXSWF\

PAGE 54

&+$37(5 (03,5,&$/ 02'(/ 2) %$1.5837&< 7KH HYLGHQFH WKDW ERQG UDWLQJ GRZQJUDGHV SURYLGH QHZ LQIRUPDWLRQ WR WKH PDUNHW VXJJHVWV WKDW SURSHU DQDO\VLV RI IXQGDPHQWDO FRUSRUDWH IDFWRUV VXFK DV WKH RQHV SUHVXPDEO\ XVHG E\ 6t3nV DQG 0RRG\n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t3nV DQG 0RRG\nV VRPHWLPHV JLYH FRQIOLFWLQJ VLJQDOV DQG WKHUH KDYH EHHQ FDVHV ZKHUH WKH GRZQJUDGH HLWKHU GRHV QRW SUHGDWH WKH EDQNUXSWF\ ILOLQJ E\

PAGE 55

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

PAGE 56

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n IXQGDPHQWDO FKDUDFWHULVWLFV =DYJUHQ f SURYLGHV VRPH HYLGHQFH WKDW DVVHW HIILFLHQF\ LV DQ LPSRUWDQW LQGLFDWRU ZKLFK PD\ EH RI XVH WR XV EXW ZH EHOLHYH WKDW RWKHU YDULDEOHV ZRXOG DOVR EH QHHGHG 'HUNLQGHUHQ DQG &UXP f GHYHORSHG D IUDPHZRUN NQRZQ DV WKH 3RWHQWLDO DQG 5HVLOLHQFH (YDOXDWLRQ 3$5(f PRGHO RQ WKH LVVXHV RI ORQJWHUP ULVNUHWXUQ EDODQFH 7KLV IUDPHZRUN VXJJHVWV VRPH SRVVLEOH YDULDEOHV ZKLFK FDQ EH XVHG WR DVVHVV WKH IXQGDPHQWDO DQG VWUDWHJLFDOO\ FUXFLDO FKDUDFWHULVWLFV RI D FRPSDQ\ %DVLQJ RXU YLHZV IURP VXFK D IXQGDPHQWDO SHUVSHFWLYH DV VXJJHVWHG E\ 'HUNLQGHUHQ DQG &UXP f DQG =DYJUHQ f ZH KDYH DUULYHG DW D OLVW RI HLJKW

PAGE 57

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f WKRVH ZKLFK DVVHVV FRPSDQ\ RSHUDWLQJ DQG SURILWDELOLW\ FKDUDFWHULVWLFV ZKLFK ZLOO EH UHIHUUHG WR DV WKH *URXS YDULDEOHVf DQG f WKRVH ZKLFK DVVHVV IXQGDPHQWDO FRPSDQ\ FKDUDFWHULVWLFV ZKLFK ZLOO EH UHIHUUHG WR DV WKH *URXS ,, YDULDEOHVf ([LVWLQJ EDQNUXSWF\ SUHGLFWLRQ PRGHOV IRFXV PRVWO\ RQ UHODWLRQVKLSV WKDW ZRXOG EH LQFOXGHG LQ WKH ILUVW JURXS $OWKRXJK WKHVH IDFWRUV DUH REYLRXVO\ LPSRUWDQW WR EDQNUXSWF\ SDUWLFXODUO\ LQ WKH VKRUW UXQ ZH FRQWHQG WKDW WKH VHFRQG JURXS PXVW DOVR EH FRQVLGHUHG LI WKH PRGHO LV WR KDYH DGHTXDWH ORQJWHUP SUHGLFWLYH SRZHUV

PAGE 58

)RXU YDULDEOHV DUH LQFOXGHG LQ *URXS WR PHDVXUH WKH FRPSDQ\nV ILQDQFLDO DQG RSHUDWLQJ SRVLWLRQV 7KH ILUVW LV UHWXUQ RQ DVVHWV RU 52$ GHILQHG DV (%,77$ 7KLV LV WKH EDVLF HDUQLQJ SRZHU UDWLR DQG LV D VWURQJ VLJQDO RI SURILWDELOLW\ :H FRQVLGHU WKDW LW LV D SDUWLFXODUO\ JRRG UHSUHVHQWDWLRQ RI WKH FRPSDQ\nV SHUIRUPDQFH LQ LPSOHPHQWLQJ JURZWK LQ WKH SDVW 7KH VHFRQG YDULDEOH LQ WKLV FDWHJRU\ LV WKH IL[HG FKDUJH FRYHUDJH UDWLR )&& 7KLV UDWLR PHDVXUHV VLPXOWDQHRXVO\ D FRPSDQ\nV OHYHO RI GHEW DQG KRZ ZHOO LWV FDVK IORZ FRYHUV WKH VHUYLFLQJ UHTXLUHPHQWV RI GHEW ,W LV RQH RI WKH PRVW LPSRUWDQW LQGLFDWRUV RI WKH DELOLW\ RI WKH ILUP WR VXUYLYH DGYHUVLW\ LQ WKH VKRUWUXQ 7KH WKLUG YDULDEOH LV WKH PDUNHWWRERRN UDWLR 7KH PDUNHWWRERRN UDWLR PHDVXUHV LQYHVWRUVn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

PAGE 59

7R DVVHVV WKH VHFRQG FDWHJRU\ RI GLIIHUHQFH LQGLFDWRUV WKH *URXS ,, YDULDEOHV ZKLFK PHDVXUH FRPSDQ\ IXQGDPHQWDO FKDUDFWHULVWLFV IRXU DGGLWLRQDO YDULDEOHV DUH XVHG )LUVW RZQHUVKLS DVVHVVHV WKH DJHQF\ HIIHFWV RI WKH FRPSDQ\nV PDQDJHPHQW $FFRUGLQJ WR WKH DJHQF\ WKHRU\ OLWHUDWXUH VXFK DV -HQVHQ DQG 0HFNOLQJ f ZKHQ WKH VKDUHKROGHUV FRQWUDFW ZLWK WKH PDQDJHPHQW IRU WKH ODWWHU WR VHUYH WKH IRUPHU DV DJHQWV VRPH FRVW LV LQHYLWDEO\ LQYROYHG LQFOXGLQJ D GHDGn ZHLJKW ORVV 7KLV FRVW RFFXUV EHFDXVH PDQDJHPHQW DQG VKDUHKROGHUV LQ PRVW FRUSRUDWLRQV SDUWLFXODUO\ WKH ODUJHU RQHVf DUH XVXDOO\ GLVWLQFWO\ VHSDUDWH JURXSV DQG WKH\ OLNHO\ KDYH GLIIHULQJ HYHQ FRQIOLFWLQJ LQWHUHVWV )RU H[DPSOH PDQDJHPHQW RI $PHULFDnV WRS ILUPV RZQ RQO\ SHUFHQW RI WKHLU FRPSDQLHVn VWRFN DQG QHDUO\ RQH RXW RI WHQ )RUWXQH &KLHI ([HFXWLYH 2IILFHUV RZQ QR VWRFN DW DOO LQ WKHLU FRPSDQLHV %HFDXVH RI VXFK ORZ RZQHUVKLS VWDNHV PDQ\ VFKRODUV KDYH TXHVWLRQHG ZKHWKHU VXFK PDQDJHPHQWV UHDOO\ VHUYH WKH VKDUHKROGHUVn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

PAGE 60

VKDUHKROGHU GRHV QRW WKHQ ZKHWKHU D PDQDJHPHQW LV D QHW SXUFKDVHU RU D QHW VHOOHU RI WKH FRPSDQ\nV VWRFN VKRZV PDQDJHPHQWnV FRQILGHQFH LQ LWV RZQ SURMHFWV )XUWKHU LW DOVR VKRZV ZKHWKHU PDQDJHPHQW LV FRPPLWWHG WR WKH VXFFHVV RI WKH FRPSDQ\ RU ZKHWKHU LW LV VLPSO\ JUDEELQJ IRU SDUDFKXWHV 7R LQFRUSRUDWH WKHVH LGHDV LQWR WKH PRGHO ZH LQFOXGH WKH QHW UDWH RI PDQDJHPHQW VWRFN DFJXLVLWLRQV DV D YDULDEOH GHILQHG DV WKH GLIIHUHQFH EHWZHHQ PDQDJHPHQW SXUFKDVHV DQG VDOHV RI WKH FRPSDQ\nV VWRFN PHDVXUHG LQ SHUFHQWDJHV RI WRWDO HTXLW\ 7KH VHFRQG YDULDEOH LQ WKLV FDWHJRU\ LV WKH FDSLWDO LQWHQVLW\ UDWLR GHILQHG DV WRWDO DVVHWV GLYLGHG E\ VDOHV 7KLV UDWLR PHDVXUHV KRZ PXFK D FRPSDQ\ PXVW LQYHVW LQ DVVHWV WR H[SDQG LWV VDOHV 7KXV LW LQGLUHFWO\ WHOOV XV ERWK WKH VWUXFWXUH RI WKH LQGXVWU\ DQG KRZ PXFK WKH FRPSDQ\ ZLOO KDYH WR UHO\ RQ IUHVK FDSLWDO WR IXHO JURZWK ,Q WKLV GLVVHUWDWLRQ WKRXJK ZH GR QRW IROORZ WKH WUDGLWLRQDO GHILQLWLRQV RI LQGXVWU\ DV JLYHQ E\ &203867$7 RU 'XQ t %UDGVWUHHW 5DWKHU GHYHORSLQJ IURP WKH LGHDV RI 7VH f ZH UHJURXS WKH &203867$7 6,& FRGHV VR DV WR GHILQH OLQHV RI LQGXVWU\ E\ WKH QDWXUH RI WKH EXVLQHVV 7KHVH GHILQLWLRQV RI LQGXVWU\ DORQJ ZLWK WKH JURZWK UDWH DQG EDQNUXSWF\ UDWH RI HDFK LQGXVWU\ DUH SUHVHQWHG LQ 7DEOH $V 7DEOH DQG )LJXUH VKRZ D FRPSDQ\nV OLQH RI EXVLQHVV GRHV LPSDFW VLJQLILFDQWO\ RQ EDQNUXSWF\ :KLOH WKH VSHFLDOW\ PDQXIDFWXULQJ LQGXVWU\ FRGH f KDG D b FXPXODWLYH EDQNUXSWF\ UDWH GXULQJ WKH SHULRG WKH FKHPLFDOV

PAGE 61

FRGH f DQG XWLOLWLHV FRGH f LQGXVWULHV VXIIHUHG RQO\ DERXW D b FXPXODWLYH EDQNUXSWF\ UDWH GXULQJ WKH VDPH SHULRG ([SDQGLQJ RQ WKH LQIRUPDWLRQ DERXW FRPSHWLWLYHQHVV FRQWDLQHG LQ WKH LQGXVWU\ FKDUDFWHULVWLFV DQG WR DVVHVV D FRPSDQ\nV SRVLWLRQ ZLWKLQ LWV LQGXVWU\ ZH HPSOR\ WKH UHODWLYH VDOHV JURZWK UDWLR 7KLV UDWLR LV WKH GLIIHUHQFH EHWZHHQ WKH VDOHV JURZWK UDWH IRU WKH SULPDU\ RU FRUH LQGXVWU\ LQ ZKLFK WKH FRPSDQ\ FRPSHWHV DQG WKH UDWH RI JURZWK RI WKH FRPSDQ\nV VDOHV ,W LQGLFDWHV WKH FRPSHWLWLYHQHVV RI WKH FRPSDQ\ ZLWKLQ LWV PDMRU LQGXVWU\ LQ WHUPV RI LWV DELOLW\ WR JDLQ PDUNHW VKDUH ,W DOVR FRPSOHPHQWV WKH EDODQFH UDWLR RI LQFRPH DQG VDOHV WKH *URXS YDULDEOH WKDW PHDVXUHV ZKHWKHU WKH FRPSDQ\ KDV JDLQHG PDUNHW VKDUH DW WKH H[SHQVH RI SURILW PDUJLQV 7KH ILQDO YDULDEOH LQ WKH *URXS ,, YDULDEOHV VHW LV WKH UHODWHGQHVV UDWLR DV GHVFULEHG E\ 5XPHOW f 7KLV YDULDEOH DVVHVVHV D FRPSDQ\nV GLYHUVLILFDWLRQ SURJUDP LQ WKDW LW LQGLFDWHV WKH H[WHQW WR ZKLFK WKH FRPSDQ\ IRFXVHV LWV HIIRUWV RQ D SRUWIROLR RI UHODWHG EXVLQHVVHV WKDW FRXOG EH H[SHFWHG WR KDYH V\QHUJLVWLF LQWHUGHSHQGHQFLHV ,W LV FDOFXODWHG DV b RI DVVHWV LQ UHODWHG VHJPHQWVf [ QXPEHU RI VHJPHQWV f QXPEHU RI VHJPHQWVf DQG GHWHUPLQHV ZKHWKHU D FRPSDQ\ LV D VLQJOH EXVLQHVV D YDOXH RI ]HURf XQUHODWHG GLYHUVLILHG D ORZ SRVLWLYH YDOXHf RU UHODWHG GLYHUVLILHG D KLJK SRVLWLYH YDOXHf $V 5XPHOW f SRLQWV RXW GLIIHUHQW VWUDWHJLHV RI

PAGE 62

GLYHUVLILFDWLRQ FDQ VWURQJO\ DIIHFW D FRPSDQ\nV ORQJWHUP SURVSHFWV DQG WKH GHJUHH RI UHODWHGQHVV VKRXOG DVVHVV WKH EHQHILWV RI WKRVH GLYHUVLILFDWLRQV $FFRUGLQJ WR 5XPHOW WKH KLJKHU WKH UHODWHGQHVV UDWLR WKH JUHDWHU WKH FKDQFHV RI JRRG SHUIRUPDQFH 7KH HLJKW YDULDEOHV GHVFULEHG DERYH FRYHU WKH WZR EURDG FDWHJRULHV DQG DOVR FROOHFWLYHO\ DGGUHVV WKH YDULRXV GLPHQVLRQV RI WKH 3$5( IUDPHZRUN ZKLFK DUH f WKH H[WHQW WR ZKLFK WKH ILUP KDV JRRG JURZWK RSSRUWXQLWLHV DYDLODEOH f ZKHWKHU RU QRW WKH PDUNHW SHUFHLYHV WKDW WKH FRPSDQ\ FDQ H[SORLW WKH JURZWK RSSRUWXQLWLHV VXFFHVVIXOO\ f WKH GHJUHH WR ZKLFK WKH IRUWXQHV RI WKH ILUP DUH VXEMHFW WR IRUHVHHDEOH DGYHUVLWLHV LQ WKH IXWXUH f ZKHWKHU RU QRW WKH ILUP KDV WKH DELOLW\ WR VXUYLYH VXFK DGYHUVLWLHV 5ROH RI 9DULDEOHV LQ WKH 0RGHO 7KH PRGHO RI EDQNUXSWF\ ULVN GHYHORSHG LQ WKLV GLVVHUWDWLRQ LQFOXGHV WKH HLJKW YDULDEOHV GLVFXVVHG DERYH DQG XVHV WKHP WR GHULYH D VXPPHG SUREDELOLW\ RI EDQNUXSWF\ 8VLQJ /RJLVW $QDO\VLV ZKLFK ZLOO EH GLVFXVVHG LQ &KDSWHU D SUREDELOLVWLF IXQFWLRQ RI D FXPXODWLYH VFRUH RI ] LV GHULYHG ZKLFK LQ WXUQ LV FRPSRVHG RI WKHVH HLJKW IDFWRUV 3 I]f ZKHUH I ] D4 EA E; E; ZKHUH E DUH FRHIILFLHQWV ; LV 52$ (%,77$f ; LV )&& ; LV %DODQFH 5DWLR ; LV 0DUNHW%RRN 5DWLR ; LV 5HODWHGQHVV 5DWLR

PAGE 63

; LV 1HW 5DWH RI 0DQDJHPHQW 6WRFN $FTXLVLWLRQV ; LV 5HODWLYH 6DOHV *URZWK 5DWH ; LV &DSLWDO ,QWHQVLW\ +\SRWKHVHV %HFDXVH WKH PHWKRGRORJ\ XVHG LQ WKLV GLVVHUWDWLRQ /RJLVW $QDO\VLV GLVFXVVHG LQ &KDSWHU f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nV FRHIILFLHQW LV QHJDWLYH EHFDXVH LI LQFRPH LV JURZLQJ PRUH EULVNO\ WKDQ VDOHV WKHQ WKH FRPSDQ\ LV H[SHULHQFLQJ ZLGHQLQJ SURILW PDUJLQV ZKLFK ZRXOG LQHYLWDEO\ OHDG WR KLJKHU SURILWDELOLW\ DQG HYHQ PRUH JURZWK LQ WKH IXWXUH 6LQFH ORQJWHUP FRQVLGHUDWLRQV VRPHWLPHV PDNH LW QHFHVVDU\ WR VDFULILFH VKRUWUXQ SURILWDELOLW\ WR GULYH RXW WKH FRPSHWLWLRQ DQG JDLQ PDUNHW

PAGE 64

VKDUH KRZHYHU HYHQ JRRG VROLG FRPSDQLHV PD\ KDYH QHJDWLYH YDOXHV IRU WKLV UDWLR +HQFH RXU FRQILGHQFH DERXW WKH VLJQ RI WKLV SDUWLFXODU FRHIILFLHQW LV OHVV WKDQ IRU WKRVH RI WKH ILUVW WKUHH YDULDEOHV 0RUH LPSRUWDQW WKDQ WKH VLJQ WKRXJK LV WKH LGHD WKDW WKH EDODQFH UDWLR IRU WKH EDQNUXSW FRPSDQLHV VKRXOG EH VLJQLILFDQWO\ GLIIHUHQW WKDQ WKH UDWLR IRU QRQn EDQNUXSW FRPSDQLHV 7KH UHODWHGQHVV UDWLRnV FRHIILFLHQW LV QHJDWLYH EHFDXVH D ODUJHU UHODWHGQHVV UDWLR LQGLFDWHV WKDW D FRPSDQ\ LV GLYHUVLILHG LQWR UHODWHG OLQHV RI EXVLQHVV ZKLFK PHDQV DV 5XPHOW f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f SRLQWV RXW ODUJHU FDSLWDO LQWHQVLW\ LV DVVRFLDWHG ZLWK ODUJHU FRPSDQ\ VL]H DQG ODUJHU LQGXVWULDO FRPSDQLHV GR QRW JR EDQNUXSW DV HDVLO\ DV VPDOOHU RQHV 2Q WKH RWKHU KDQG WKRXJK WR WKH H[WHQW WKDW WKH UHFLSURFDO RI

PAGE 65

WKH FDSLWDO LQWHQVLW\ UDWLR LV DQ LQGLFDWRU RI DVVHW HIILFLHQF\ WKH QHJDWLYH VLJQ ZRXOG EH FRXQWHU WR WKH ILQGLQJV RI =DYJUHQ f

PAGE 66

7DEOH ,QGXVWU\ &ODVVLILFDWLRQ *URZWK DQG %DQNUXSWF\ 5DWHV ,QGXVWU\ &203867$7 *URZWK %DQNUXSWF\ 1XPEHU ,QGXVWU\ 6,& 5DWH 5DWH )RRG b b &ORWKLQJ t 7H[WLOHV b b 3DSHU t 3XEOLVKLQJ b b &KHPLFDOV b b 'UXJV b 1$ 3HWUROHXP 5HILQLQJ b b 5XEEHU t /HDWKHU b b *ODVV t &HPHQW b b 0HWDOV b b ,QGXVWULDO 0DFKLQHU\ b b 2IILFH 0DFKLQHU\ t (OHFWURQLF (JXLSPHQW b b (OHFWULFDO (TXLSPHQW b b 0RWRU 9HKLFOHV b b 6FLHQWLILF t 6XUYH\LQJ (TXLSPHQW b b 7UDQVSRUWDWLRQ b b $JULFXOWXUH b b ([WUDFWLYH b b &RQVWUXFWLRQ b b 6SHFLDOW\ 0DQXI b b 8WLOLWLHV b b :KROHVDOH b b &RQVXPHU 3URGXFWV b b 6HUYLFHV b b )LQDQFLDO b b *URZWK 5DWH LV WKH DYHUDJH DQQXDO JURZWK UDWH RI WKH LQGXVWU\ EHWZHHQ DQG %DQNUXSWF\ 5DWH LV WKH SHUFHQWDJH RI EXVLQHVVHV WKDW IDLOHG GXULQJ WKDW SHULRG

PAGE 67

3HUFHQW RI %DQNUXSWF\ &XPXODWLYHf L V P P L L M t L L W ,QGXVWU\ &RGH &XP b )LJXUH ,QGXVWU\ DQG %DQNUXSWF\ 5DWH

PAGE 68

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f ZKLFK ZDV GLVFXVVHG HDUOLHU LQ WKH /LWHUDWXUH 5HYLHZ VHFWLRQ ZDV SUREDEO\ WKH ILUVW ZRUN RQ EDQNUXSWF\ WR XVH ORJLVW DQDO\VLV $OWKRXJK WKH UHVHDUFK GLG

PAGE 69

QRW SURGXFH D VLJQLILFDQWO\ YLDEOH PRGHO WKH ZRUN QHYHUWKHOHVV SURYLGHG VRPH LQWHUHVWLQJ LQVLJKWV LQWR WKH XVH RI ORJLVW DQDO\VLV IRU HPSLULFDO VWXGLHV RI EDQNUXSWF\ ,Q 2KOVRQnV PRGHO ;‘ GHQRWHG D YHFWRU RI SUHGLFWRUV IRU FRPSDQ\ L  GHQRWHG D YHFWRU RI XQNQRZQ SDUDPHWHUV DQG 3;M cf ZKHUH 3 LV D SUREDELOLW\ IXQFWLRQ 3 f GHQRWHG WKH SUREDELOLW\ RI EDQNUXSWF\ IRU D JLYHQ VHW RI YHFWRUV ;c DQG I 7KH ORJDULWKP RI WKH OLNHOLKRRG RI DQ\ VSHFLILF RXWFRPH DV UHIOHFWHG E\ WKH ELQDU\ VDPSOH VSDFH RI EDQNUXSWF\ YHUVXV QRQn EDQNUXSWF\ LV JLYHQ E\ /f ( ORJ 3;M f ORJ 3I;M f f ZKHUH L M DUH HOHPHQWV RI WKH 6 LQGH[ VHW RI EDQNUXSW FRPSDQLHV DQG 6 LQGH[ VHW RI QRQEDQNUXSW FRPSDQLHV UHVSHFWLYHO\ )RU DQ\ VSHFLILHG IXQFWLRQ 3 WKH PD[LPXP OLNHOLKRRG HVWLPDWHV RI I "Q DUH REWDLQHG E\ VROYLQJ PD[A /f %HFDXVH ZH GR QRW DV \HW KDYH D IXOO WKHRU\ RI EDQNUXSWF\ KRZHYHU ZH FDQQRW HDVLO\ ILQG DQ DSSURSULDWH FODVV RI IXQFWLRQV 3 $V D SUDFWLFDO PDWWHU WKHUHIRUH ZH FDQ RQO\ VHOHFW D IXQFWLRQ IRU WKH VDNH RI FRPSXWDWLRQDO DQG LQWHUSUHWDWLYH VLPSOLFLW\ 2QH VXFK IXQFWLRQ LV WKH ORJLVWLF IXQFWLRQ 3 H[S^\c`f ZKHUH \ ( !M;IM Sn;‘ 7KLV IRUPXOD KDV WZR LPSOLFDWLRQV )LUVW 3 LV LQFUHDVLQJ LQ \ 6HFRQG \ LV HTXDO WR ORJ 3 f§ 3ff

PAGE 70

/LNH GLVFULPLQDQW DQDO\VLV ORJLVW DQDO\VLV ZHLJKWV WKH LQGHSHQGHQW YDULDEOHV DQG FUHDWHV D VFRUH IRU HDFK FRPSDQ\ 7KH = VFRUH REWDLQHG PD\ EH XVHG WR GHWHUPLQH WKH SUREDELOLW\ RI PHPEHUVKLS LQ D JURXS ZKHUH 3UREDELOLW\ RI EDQNUXSWF\ H[S]ff H[S D E; ES;Sff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f KDV HPSOR\HG WKH WUDGLWLRQDO OLQHDU GLVFULPLQDQW DQDO\VLV DOWKRXJK ERWK WKH OLQHDU DQG WKH TXDGUDWLF IRUPV KDYH EHHQ XVHG )RU RXU GLVVHUWDWLRQ WKRXJK ZH FRQVLGHU WKDW ORJLVW DQDO\VLV FDQ \LHOG D VXSHULRU PRGHO EHFDXVH ORJLVW DQDO\VLV GRHV QRW VKDUH PDQ\ RI WKH SUREOHPV IDFHG E\ GLVFULPLQDQW DQDO\VLV )LUVW OLQHDU GLVFULPLQDWLRQ LV EDVLFDOO\ D PXOWLYDULDWH WHFKQLTXH WKDW DVVLJQV D VFRUH WR HDFK HOHPHQW LQ D VDPSOH

PAGE 71

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

PAGE 72

WKDW WKH JURXS FRYDULDQFHV DUH QRW VWDWLVWLFDOO\ HTXLYDOHQW DV LQGLFDWHG E\ %R[n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f DQG 0DUNV DQG 'XQQ f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

PAGE 73

ZLWK D SUREDELOLW\ RI EDQNUXSWF\ VR WKDW JLYHQ WKDW D FRPSDQ\ EHORQJV WR D FHUWDLQ VDPSOH ORJLVW DQDO\VLV SURYLGHV WKH SUREDELOLW\ RI IDLOXUH 6HFRQG XQOLNH OLQHDU GLVFULPLQDQW DQDO\VLV ORJLVW DQDO\VLV GRHV QRW UHTXLUH WKDW WKH LQGHSHQGHQW YDULDEOHV EH PXOWLYDULDWH QRUPDOV RU WKDW JURXSV KDYH HTXDO FRYDULDQFH PDWULFHV +DUUHOO DQG /HH f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

PAGE 74

2XU GDWD FRPH IURP VHYHUDO VRXUFHV :H XVH WKH &203867$7 5HVHDUFK 7DSH DQG ,QGXVWU\ 7DSH IRU EDVLF GDWD RQ ILQDQFLDO YDULDEOHV RI ERWK RXU VDPSOHV 7KH &203867$7 5HVHDUFK 7DSH SURYLGHV VXFK GDWD IRU WKH EDQNUXSW VDPSOH ZKLOH WKH ,QGXVWU\ 7DSH SURYLGHV VXFK GDWD IRU RXU PDWFKLQJ VDPSOH :H XVH WKH 2ZQHUVKLS 5HSRUWLQJ 6\VWHP 7DSH SXEOLVKHG E\ WKH 1DWLRQDO $UFKLYHV DQG 5HFRUG 6HUYLFHV WR REWDLQ GDWD RQ WKH PDQDJHPHQW DFTXLVLWLRQ RI FRPSDQ\ VWRFN 1H[W ZH XVH WKH &203867$7 6HJPHQWV ,QIRUPDWLRQ 7DSH WR ILQG WKH VHJPHQWV RI RXU FRPSDQLHV DQG WR FRPSXWH WKHLU GHJUHH RI GLYHUVLILFDWLRQ )LQDOO\ ZH XVH WKH &203867$7 5HVHDUFK DQG &203867$7 ,QGXVWU\ WDSHV DJDLQ WR FDOFXODWH LQGXVWU\ JURZWK DQG EDQNUXSWF\ UDWHV $W WKH VWDUW ZH KDG EDQNUXSW FRPSDQLHV WKDW ZHUH GHOHWHG IURP WKH &203867$7 ,QGXVWU\ 7DSH DQG PRYHG WR WKH &203867$7 5HVHDUFK 7DSH EHWZHHQ DQG E\ D GHOHWLRQ FRGH RI ZKLFK LQGLFDWHV EDQNUXSWF\ :H FRXOG QRW KRZHYHU ILQG \HDUV RI FRQWLQXRXV GDWD IRU DOO FRPSDQLHV EHFDXVH WKH LQIRUPDWLRQ ZH QHHGHG ZDV RQ VHYHUDO GLIIHUHQW WDSHV HDFK RI ZKLFK KDG LQIRUPDWLRQ RQ GLIIHUHQW WLPH SHULRGV 7KH &203867$7 %XVLQHVV 6HJPHQWV ,QIRUPDWLRQ 7DSH IRU H[DPSOH KDV GDWD RQO\ IURP RQZDUG DV GRHV WKH 2ZQHUVKLS 5HSRUWLQJ 6\VWHP 7KH 0DVWHU &XUUHQW 7DSH RI 2ZQHUVKLS 5HSRUWLQJ 6\VWHP KRZHYHU RIIHUV GDWD IURP -DQXDU\ WR $XJXVW DQG WKH 0DVWHU +LVWRU\ 7DSH RIIHUV GDWD IURP -DQXDU\ WR $SULO (YHQ WKRXJK ZH XVHG HYHU\ DYDLODEOH WDSH DQG HYHQ FDOFXODWHG VHYHUDO YDULDEOHV E\ KDQG

PAGE 75

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n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n EDQNUXSW FRPSDQLHV ZDV VHOHFWHG %HFDXVH RI WKLV WKH SUREDELOLWLHV IURP ORJLVW DQDO\VLV PXVW EH DGMXVWHG IRU WKH HIIHFWV RI WKH VDPSOH VHOHFWLRQ RU HOVH WKH\ ZRXOG EHFRPH PHDQLQJOHVV EHFDXVH E\ VHOHFWLQJ D VDPSOH GLIIHUHQWO\ ZH FDQ GHULYH FRPSOHWHO\ GLIIHUHQW UHVXOWV ,Q WKH IROORZLQJ

PAGE 76

FKDSWHU ZH H[SODLQ KRZ ZH DGMXVWHG RXU UHVXOWV DQG GLVFXVV LQ JUHDWHU GHWDLO KRZ WKH VHOHFWLRQ SURFHVV DFWXDOO\ DIIHFWV WKH UHVXOWV RI ORJLVW DQDO\VLV

PAGE 77

7DEOH 2ULJLQDO %DQNUXSWf DQG 0DWFKLQJ 6DPSOH IRU 'HULYLQJ (PSLULFDO %DQNUXSWF\ 0RGHO 2ULJLQDO 6DPSOH 0 &180 &RPSDQ\ 1DPH $7, ,QF $OGHEDUDQ 'ULOOLQJ &R ,QF $OGRQ ,QGXVWULHV ,QF $PHULFDQ )XHO 7HFKQRORJLHV $SDFKH (QHUJ\ t 0LQHUDOV $UJRQDXW (QHUJ\ &RUS %HNHU ,QGXVWULHV %XWWHV *DV t 2LO &R &DSLWRO $LU ,QF &DUGLV &RUS &KDUJLW ,QF &KHPLFDO ,QYHVWRUV ,QF &RPPRGRUH 5HVRXUFHV &RUS &RQQHU &RUS &RVPHWLF 6FLHQFHV ,QF &UDZIRUG (QHUJ\ ,QF &UXWFKHU 5HVRXUFHV &RUS &\WR[ &RUS 'DNRWD 0LQHUDOV ,QF 'DQNHU /DEV ,QF 'DWDWURQ ,QF 'LVFRYHU\ 2LO /WG (FRQR 7KHUP (QHUJ\ 6\VWHPV (PSLUH 2LO t *DV &R (QHUJ\ ([FKDQJH (QG/DVH ,QF (QWHUSULVH 7HFKQRORJLHV ,QF *DPH[ ,QGXVWULHV ,QF *XOI (QHUJ\ &RUS +HOLRQHWLFV ,QF ,QIRUPDWLRQ 'LVSOD\V ,QF ,QWL 6WUHWFK 3URGV ,QWL 7HOGDWD &RUS ,QWHUVWDWH 0RWRU )UHLJKW 0*) 2LO &RUS 0DJLF 0DUNHU ,QGXVWULHV ,QF 0HJR ,QWHUQDWLRQDO 0LG$PHULFD 3HWUROHXP ,QF 0XWXDO 2LO RI $PHULFD ,QF 1DWLRQDO %XVLQHVV &RPP &RUS 1$73$& ,QF 1LFNORV 2LO t *DV &R 2PQL0HGLFDO 3HRSOHnV 5HVWXDUDQWV ,QF 4XDQWD 6\VWHPV &RUS 7R ,QG '180 $VVHW

PAGE 78

7DEOH f§FRQWLQXHG 4XLN3ULQW RI $PHULFD ,QF 5HVHUnV )LQH )RRGV ,QF 5LFKPRQG 7DQN &DU &R 5REOLQ ,QGXVWULHV 6DPERnV 5HVWDXUDQWV 6DQWHH &RUS 6D[RQ ,QGXVWULHV 6HLVFRP 'HOWD ,QF 6HUYDPDWLF 6\VWHPV ,QF 9LDEOH 5HVRXUFHV ,QF 9XHERWLFV &RUS ;HQHUH[ &RUS ;RQLFV ,QF =\WUH[ &RUS 0DWFKLQFU 6DPSOH 0L &180 &RPSDQ\ 1DPH 7R ,QG '180 $VVHW &RPSWHN 5HVHDUFK ,QF 6XPPLW (QHUJ\ ,QF /\GDOO ,QF /HH 3KDUPDFHXWLFDOV 'LYHUVLILHG ,QGXVWULHV ,QF 6ZLIW (QHUJ\ &R 1&+ &RUS &DQDGLDQ 2FFLGHQWDO 3HWUROHXP +XGVRQ *HQHUDO &RUS -RUJHQVHQ (DUOH 0f &R &RPSXWHU 7DVN *URXS ,QF 7HFKQLFDO 7DSH ,QF %DUXFK)RVWHU &RUS 2DNZRRG +RPHV 75& &RV ,QF &RQVROLGDWHG 2LO t *DV 6DJH (QHUJ\ &R 6SHHG23ULQW %XVLQHVV 0DFKLQHV &RHXU 'n$OHQH 0LQHV &RUS 8QLYHUVLW\ 3DWHQWV ,QF 'DYLV :DWHU t :DVWH :LOVKLUH 2LO RI 7H[DV &RQTXHVW ([SORUDWLRQ &R 3UDLULH 2LO 5R\DOWLHV &R /WG 1RUWKJDWH ([SORUDWLRQ /WG $VWUH[ ,QF $GDPV 5HVRXUFHV t (QHUJ\ ,QF 'HVLJQ&UDIW ,QGXVWULHV &DOODKDQ 0LQLQJ &RUS ,QWHU*UDSK &RUS *HQHUDO $XWRPDWLRQ

PAGE 79

7DEOH FRQWLQXHG 3RSH (YDQV t 5REELQV ,QF 1HZ 0H[LFR t $UL]RQD /DQG 7UDQVFRQ ,QF&DOLIRUQLD :HVWPRUHODQG &RDO &R %61 &RUS $UWUD *URXS ,QF 065 ([SORUDWLRQV /WG 1RUG 5HVRXUFHV &RUS &RPSXWHU )DFWRU\ ,QF 7KUHH 'HSDUWPHQW *OREDO 1DWXUDO 5HVRXUFHV ,QF $PHULFDQ 6FLHQFH (QJLQHHULQJ 5 &RUS *7, &RUS &KULVWLDQD &RPSDQLHV &RQQHOO\ &RQWDLQHUV ,QF 7HOH)OH[ ,QF 3HQQ (QJLQHHULQJ t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nV DVVHW VL]H

PAGE 80

&+$37(5 5(68/76 2) (03,5,&$/ 02'(/6 2) %$1.5837&< 8VLQJ WKH VDPSOH GDWD GHVFULEHG LQ &KDSWHU ZH ZHUH DEOH WR GHULYH ILYH SUREDELOLVWLF PRGHOV GHVLJQDWHG 3 WKURXJK 3 $OO RI WKHP DUH EDVHG RQ WKH IDFWRUV GHVFULEHG LQ &KDSWHU DQG WKH\ DVVHVV EDQNUXSWF\ ULVN RQH WKURXJK ILYH \HDUV DKHDG RI WLPH FRUUHVSRQGLQJ WR WKH VXEVFULSWRQ 3Q 7KH RQO\ GLIIHUHQFH DPRQJ WKHVH GLVWLQFW PRGHOV LV LQ WKHLU FRHIILFLHQWV WKH\ DOO DVVXPH WKH IRUP GLVFXVVHG LQ &KDSWHU 7DEOH VKRZV WKH VLJQV RI WKH FRHIILFLHQWV RI WKH YDULDEOHV IRU HDFK PRGHO 3Q 0RVW RI WKHVH FRHIILFLHQWV FRQIRUP WR RXU H[SHFWDWLRQV DV H[SODLQHG LQ &KDSWHU EXW RWKHUV VKRZ VLJQLILFDQW GLIIHUHQFHV 7KH VLJQV IRU WKH FRHIILFLHQWV RI WKH EDODQFH UDWLR DQG WKH QHW UDWH RI PDQDJHPHQW VWRFN DFTXLVLWLRQV VHHP WR YDU\ UDQGRPO\ EXW OHDQ WRZDUGV EHLQJ SRVLWLYH ZKLOH ZH H[SHFWHG ERWK WR EH QHJDWLYH %RWK RI WKHVH YDULDEOHV FDQ EH SRVLWLYH RU QHJDWLYH DQG RXU H[SHFWDWLRQ ZDV VWDWHG IRU WKH QRUPDO FDVH IRU ZKLFK WKH H[SHFWHG YDOXH RI WKH YDULDEOH LV SRVLWLYH /RRNLQJ DW WKH UDZ GDWD D VLJQLILFDQW QXPEHU RI WKH YDOXHV LQ ERWK VDPSOHV ZHUH QHJDWLYH VR ZH ZRXOG KDYH WR

PAGE 81

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f DUH H[SHFWHG WR EH PRVW VLJQLILFDQW FORVH WR EDQNUXSWF\ DQG WKUHH RI WKH IRXU KDYH WKH FRUUHFW VLJQV LQ WKH ILUVW WKUHH \HDUV $OVR WKH ODVW IRXU YDULDEOHV *URXS ,,f DUH H[SHFWHG WR EH PRVW VLJQLILFDQW LQ HDUOLHU \HDUV DQG WKUHH RI WKH IRXU KDYH WKH FRUUHFW VLJQ LQ WKH ODVW IRXU \HDUV :H EHOLHYH WKDW WKLV SDWWHUQ FRQILUPV WKH YDOLGLW\ RI RXU H[SHFWDWLRQV /RJLVW DQDO\VLV SURYLGHV D WHFKQLTXH WKDW DOORZV XV WR ILQG WKH PRVW VLJQLILFDQW YDULDEOHV LQ DQ\ SUHGLFWLYH PRGHO $ YDULDEOH LV PRVW VLJQLILFDQW LI E\ &KL6TXDUH 4VWDWLVWLF DQG 0/(nV VWDWLVWLF WKH\ PHHW WKH UHTXLUHPHQWV RI HQWU\ DQG VWD\ VLJQLILFDQFH OHYHOV SUHVSHFLILHG IRU WKH PRGHO 2XU HQWU\ DQG VWD\ VLJQLILFDQFH OHYHOV ZHUH VHW DW 7DEOH VKRZV WKH PRVW VLJQLILFDQW YDULDEOHV LQ HDFK RI RXU PRGHOV &DSLWDO LQWHQVLW\ LV VLJQLILFDQW LQ DOO SHULRGV 7KLV PHDQV

PAGE 82

WKDW WKH QDWXUH RI D FRPSDQ\nV OLQH RI EXVLQHVV DOZD\V SOD\V D VLJQLILFDQW UROH :KHQ EDQNUXSWF\ LV IDU LQWR WKH IXWXUH RXU GDWD LQGLFDWH WKDW WKLV IDFWRU SOD\V WKH ODUJHVW UROH RI WKH HLJKW YDULDEOHV XVHG LQ RXU PRGHO $V ZH DSSURDFK EDQNUXSWF\ HVSHFLDOO\ RQH \HDU DKHDG RI EDQNUXSWF\ KRZHYHU 52$ )&& DQG 0DUNHW%RRN 5DWLR EHFRPH LQFUHDVLQJO\ VLJQLILFDQW 7KLV FRQILUPV RXU YLHZ WKDW LQ WKH VKRUWWHUP YDULDEOHV LQ WKH FRPSDQ\ RSHUDWLQJ DQG SURILWDELOLW\ LQGLFDWRUV JURXS WKDW DUH ZHLJKWHG WRZDUG D FRPSDQ\n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

PAGE 83

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f +HU SDSHU KRZHYHU RQO\ SUHVHQWHG RULJLQDO SUREDELOLWLHV ZKLOH ZH SUHVHQW ERWK WKH RULJLQDO SUREDELOLWLHV DQG WKRVH SUREDELOLWLHV DGMXVWHG IRU VDPSOH VHOHFWLRQ DV GLVFXVVHG EHORZf $V PHQWLRQHG LQ WKH SUHYLRXV FKDSWHU EHFDXVH ZH VHOHFWHG RXU GDWD ZLWK QRQUDQGRP VWDWHEDVHG FULWHULD ZH PXVW DGMXVW WKH SUREDELOLWLHV GHULYHG IURP ORJLVW DQDO\VLV

PAGE 84

)RU DQ\ JLYHQ ILUP L LQ WKH JHQHUDO SRSXODWLRQ ZLWK D SUREDELOLW\ 3 RI EDQNUXSWF\ ORJLVW DQDO\VLV ZRXOG JLYH D SUREDELOLW\ 3n RI EDQNUXSWF\ IRU WKDW FRPSDQ\ LQ RXU VSHFLILF VDPSOH :H PXVW ILQG ZD\V RI ILQGLQJ D UHODWLRQVKLS EHWZHHQ 3 DQG 3n DQG DOVR EHWZHHQ WKH VWUXFWXUHV RI RXU VDPSOHV $VVXPLQJ WKDW WKHUH DUH 1 EDQNUXSW DQG 1 QRQEDQNUXSW ILUPV LQ WKH JHQHUDO SRSXODWLRQ DQG Q EDQNUXSW DQG Q QRQEDQNUXSW ILUPV LQ RXU EDQNUXSW DQG QRQEDQNUXSW VDPSOHV 3n DFFRUGLQJ WR %D\HVn IRUPXOD IRU FRQGLWLRQDO SUREDELOLW\ LV HTXDO WR 3n 3 r 91f ^3 r Q1f 3f r Q1f ` f 3UHYLRXV ZRUN VXFK DV 3DOHSX f ZKLFK WULHG WR SUHGLFW PHUJHU WDUJHWV XVLQJ ORJLVW DQDO\VLV PRGHOV KDYH SDUWLDOO\ H[SORUHG WKLV UHODWLRQVKLS EXW KH RQO\ JDYH D IRUPXOD IRU 3n LQ WKH VSHFLDO FDVH ZKHQ Q 1 :H KRZHYHU GHULYHG D JHQHUDO IRUPXOD IRU 3n IRU FDVHV ZKHQ Q LV QRW HTXDO WR 1 DQG Q LV QRW HTXDO WR 1 /HW D QA D f§ Q1 7KHQ IRUPXOD f FDQ UHZULWWHQ DV 3n m r 3-+93f 3f r rff f 6XEVWLWXWLQJ 3 H[S
PAGE 85

,I D D 3n 3 7\SH HUURU ZLOO LQFUHDVH 7\SH ,, HUURU ZLOO GHFUHDVH ,I D D 3n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nV =(7$ PRGHO XQGHU RSWLPDO FRQGLWLRQV LV b DFFXUDWH RQH \HDU EHIRUH EDQNUXSWF\ EXW RQO\ b DFFXUDWH ILYH \HDUV EHIRUH EDQNUXSWF\ RXU PRGHOnV

PAGE 86

SUHGLFWLYH SRZHU EDVHG RQ GDWD IURP RQH H[SHULPHQWDO VDPSOH LV UHODWLYH VWDEOH ,W LV PRVW HIIHFWLYH RQH \HDU EHIRUH EDQNUXSWF\ ZLWK b DFFXUDF\ DV PHDVXUHG E\ WKH &,QGH[ RI ORJLVW DQDO\VLVf EXW HYHQ DW LWV ORZHVW SRLQW WKUHH \HDUV EHIRUH EDQNUXSWF\ LW VWLOO KDV b DFFXUDF\ DQG ILYH \HDUV EHIRUH EDQNUXSWF\ LW LV b DFFXUDWH 6HFRQG ZH KDYH IRXQG WKDW GXULQJ GLIIHUHQW SHULRGV EHIRUH EDQNUXSWF\ GLIIHUHQW JURXSV RI YDULDEOHV EHFRPH LPSRUWDQW LQ SUHGLFWLQJ EDQNUXSWF\ :KHQ FORVH WR EDQNUXSWF\ WKH *URXS ILQDQFLDO YDULDEOHV KDYH YHU\ KLJK SUHGLFWLYH SRZHU $V ZH PRYH IDUWKHU EDFN LQ WLPH DQG WU\ WR SUHGLFW EDQNUXSWF\ IDUWKHU DKHDG RI WLPH ZH IRXQG WKDW WKHVH YDULDEOHVn SUHGLFWLYH SRZHUV VWDUWHG WR ZDQH 7KH *URXS ,, IXQGDPHQWDO YDULDEOHV EHFDPH LQFUHDVLQJO\ PRUH SRZHUIXO DV ZH PRYHG IDUWKHU EDFN LQ WLPH 7KLV GHPRQVWUDWHV WKDW WKHUH LV D WLPH UHODWLRQVKLS EHWZHHQ RXU IXQGDPHQWDO YDULDEOHV DQG WKH IXWXUH ILQDQFLDO SRVLWLRQ RI WKH FRPSDQ\

PAGE 87

7DEOH 6LJQV RI &RHIILFLHQWV 0RGHO E E E A E E E EJ 3 3 S 3 WKURXJK 3 DUH PRGHOV IRU SUHGLFWLQJ EDQNUXSWF\ RQH WR ILYH \HDUV DKHDG RI WLPH E WKURXJK E DUH FRHIILFLHQWV RI WKH IROORZLQJ YDULDEOHV E 52$ (%,77$f E )&& E %DODQFH 5DWLR E 0DUNHW%RRN 5DWLR E 5HODWHGQHVV 5DWLR E 1HW 5DWH RI 0DQDJHPHQW 6WRFN $FTXLVLWLRQV E 5HODWLYH 6DOHV *URZWK 5DWH E &DSLWDO ,QWHQVLW\

PAGE 88

7DEOH 0RVW 6LJQLILFDQW 9DULDEOHV LQ )RUHFDVWLQJ %DQNUXSWF\ 0RGHO 0RVW 6LJQLILFDQW 9DULDEOHV 3 52$ )&& %DODQFH 5DWLR &DSLWDO ,QWHQVLW\ 0DUNHW%RRN 5DWLR 3 5HODWLYH 6DOHV *URZWK )&& &DSLWDO ,QWHQVLW\ 3 52$ 1HW 5DWH RI 0DQDJHPHQW 6WRFN $FTXLVLWLRQV 5HODWLYH 6DOHV *URZWK &DSLWDO ,QWHQVLW\ 3 &DSLWDO ,QWHQVLW\ 3 &DSLWDO ,QWHQVLW\ 3 WKURXJK 3 DUH PRGHOV IRU IRUHFDVWLQJ EDQNUXSWF\ RQH WKURXJK ILYH \HDUV DKHDG RI WLPH UHVSHFWLYHO\

PAGE 89

7DEOH 3UHGLFWLYH 3RZHU DQG (IIHFWLYHQHVV RI 'LIIHUHQW 0RGHOV
PAGE 90

$FFXUDF\ EDVHG RQ & f§ ,QGH[ .r ;$FFXUDF\
PAGE 91


PAGE 92

7DEOH (PSLULFDO 3UREDELOLW\ 'HQVLW\ )XQFWLRQ RI 8QDGMXVWHG %DQNUXSW 3UREDELOLWLHV 7LPH 0LGYDOXH 1RQENSW b b b b b b b b b b %DQNUXSW b b b b b b b b b b 7LPH 0LGYDOXH 1RQENSW b b b b b b b b b b %DQNUXSW b b b b b b b b b b 7LPH 0LGYDOXH 1RQENSW b b b b b b b b b b %DQNUXSW b b b b b b b b b b 7LPH 0LGYDOXH 1RQENSW b b b b b b b b b b %DQNUXSW b b b b b b b b b b 7LPH 0LGYDOXH 1RQENSW b b b b b b b b b b %DQNUXSW b b b b b b b b b b 0LGYDOXH LV WKH PLGYDOXH RI WKH SUREDELOLW\ LQWHUYDOV 1RQENSW DQG %DQNUXSW WKH SHUFHQW RI QRQEDQNUXSW DQG EDQNUXSW REVHUYDWLRQV ZKLFK IHOO ZLWKLQ WKDW LQWHUYDO DUH -?

PAGE 93

7DEOH (PSLULFDO 3UREDELOLW\ 'HQVLW\ )XQFWLRQ RI $GMXVWHG %DQNUXSW 3UREDELOLWLHV 7LPH 0LGYDOXH 1RQENSW b b b b b b b b b b %DQNUXSW b b b b b b b b b b 7LPH 0LGYDOXH 1RQENSW b b b b b b b b b b %DQNUXSW b b b b b b b b b b 7LPH 0LGYDOXH 1RQENSW b b b b b b b b b b %DQNUXSW b b b b b b b b b b 7LPH 0LGYDOXH 1RQENSW b b b b b b b b b b %DQNUXSW b b b b b b b b b b 7LPH 0LGYDOXH 1RQENSW b b b b b b b b b b %DQNUXSW b b b b b b b b b b 0LGYDOXH LV WKH PLGYDOXH RI WKH SUREDELOLW\ LQWHUYDOV 1RQENSW DQG %DQNUXSW WKH SHUFHQW RI QRQEDQNUXSW DQG EDQNUXSW REVHUYDWLRQV ZKLFK IHOO ZLWKLQ WKDW LQWHUYDO DUH FR

PAGE 94

)UDFWLRQ RI &RPSDQLHV 3UREDELOLW\ ,QWHUYDO 0LG9DOXHV )LJXUH 8QDGMXVWHG (PSLULFDO 3UREDELOLW\ 'HQVLW\ )XQFWLRQ
PAGE 95

)UDFWLRQ RI &RPSDQLHV )LJXUH f§ n 2LO 6LOO 1RQENSW %DQNUXSW 3UREDELOLW\ ,QWHUYDO 0LG9DOXHV 8QDGMXVWHG (PSLULFDO 3UREDELOLW\ 'HQVLW\ )XQFWLRQ
PAGE 96

)UDFWLRQ RI &RPSDQLHV 1RQENSW %DQNUXSW 3UREDELOLW\ ,QWHUYDO 0LG9DOXHV )LJXUH 8QDGMXVWHG (PSLULFDO 3UREDELOLW\ 'HQVLW\ )XQFWLRQ
PAGE 97

)UDFWLRQ RI &RPSDQLHV 1RQENSW %DQNUXSW 3UREDELOLW\ ,QWHUYDO 0LG9DOXHV )LJXUH 8QDGMXVWHG (PSLULFDO 3UREDELOLW\ 'HQVLW\ )XQFWLRQ
PAGE 98

)UDFWLRQ RI &RPSDQLHV 3UREDELOLW\ ,QWHUYDO 0LG9DOXHV )LJXUH 8QDGMXVWHG (PSLULFDO 3UREDELOLW\ 'HQVLW\ )XQFWLRQ
PAGE 99

)UDFWLRQ RI &RPSDQLHV 1RQENSW %DQNUXSW 3UREDELOLW\ ,QWHUYDO 0LG9DOXHV )LJXUH $GMXVWHG (PSLULFDO 3UREDELOLW\ 'HQVLW\ )XQFWLRQ
PAGE 100

)UDFWLRQ RI &RPSDQLHV 1RQENSW %DQNUXSW 3UREDELOLW\ ,QWHUYDO 0LG9DOXHV )LJXUH $GMXVWHG (PSLULFDO 3UREDELOLW\ 'HQVLW\ )XQFWLRQ
PAGE 101

)UDFWLRQ RI &RPSDQLHV 3UREDELOLW\ ,QWHUYDO 0LG9DOXHV )LJXUH $GMXVWHG (PSLULFDO 3UREDELOLW\ 'HQVLW\ )XQFWLRQ
PAGE 102

)UDFWLRQ RI &RPSDQLHV 3UREDELOLW\ ,QWHUYDO 0LG9DOXHV )LJXUH $GMXVWHG (PSLULFDO 3UREDELOLW\ 'HQVLW\ )XQFWLRQ
PAGE 103

)UDFWLRQ RI &RPSDQLHV 1RQENSW %DQNUXSW 3UREDELOLW\ ,QWHUYDO 0LG9DOXHV )LJXUH $GMXVWHG (PSLULFDO 3UREDELOLW\ 'HQVLW\ )XQFWLRQ
PAGE 104

7DEOH $GMXVWHG DQG 8QDGMXVWHG &RUUHFW 5DWLR 7\SH (UURU DQG 7\SH ,, (UURU IRU RXU 6DPSOH 8QDGMXVWHG f§ W W 7 W &RUUHFW 5DWLR b b b b b 7\SH (UURU b b b b b 7\SH ,, (UURU b b b b b $GMXVWHG f§ r W QL Q D 1 f D 1 f &RUUHFW 5DWLR b b b b b 7\SH (UURU b b b b b 7\SH ,, (UURU b b b b b

PAGE 105

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nV GDLO\ VWRFN UDWH RI UHWXUQ 6HFRQG ZH EXLOW D PRGHO RI EDQNUXSWF\ ZKLFK DVVHVVHV EDQNUXSWF\ ULVN EDVHG QRW RQO\ RQ ILQDQFLDO GDWD EXW DOVR RQ PRUH IXQGDPHQWDO FKDUDFWHULVWLFV RI D FRPSDQ\ 7KLV PRGHO LQFRUSRUDWHV QHZ GLPHQVLRQV LQWR EDQNUXSWF\ UHVHDUFK WKDW KDYH WR GDWH EHHQ OHIW RXW ,Q WKH DUHD RI ILQDQFLDO PDUNHWV DQG EDQNUXSWF\ RXU UHVHDUFK UHDFKHG WZR PDMRU FRQFOXVLRQV )LUVW ZH IRXQG WKDW WKH PDUNHW FDQ DGHTXDWHO\ DVVHVV WKH ULVN RI EDQNUXSWF\ LQ LWV

PAGE 106

RYHUDOO DVVHVVPHQW RI ULVN 2XU GDWD VKRZHG KRZ DV FRPSDQLHV PRYHG WRZDUGV ILOLQJ IRU &KDSWHU WKH PDUNHW VXFFHVVLYHO\ LQFUHDVHG WKH ULVN SUHPLXP RQ WKHLU ERQGV WKXV GHPRQVWUDWLQJ WKDW WKH PDUNHW DSSUHFLDWHG WKH LQFUHDVLQJ ULVN RI EDQNUXSWF\ DQG DFFRXQWHG IRU LW LQ LWV YDOXDWLRQ RI VHFXULWLHV 6HFRQG ZH VKRZHG WKDW D ERQG UDWLQJ GRZQJUDGH KDV D QRWDEOH LPSDFW RQ WKH PDUNHW DV PHDVXUHG E\ WKH GDLO\ UDWHV RI UHWXUQ RI D FRPSDQ\nV VWRFN :KHQ D FRPSDQ\nV ERQG LV GRZQJUDGHG LWV VWRFNnV GDLO\ UDWHV RI UHWXUQ WHQG WR IDOO SUHFLSLWRXVO\ 7KLV LQGLFDWHV WKDW ERQG GRZQJUDGHV SURYLGH QHZ LQIRUPDWLRQ WR WKH PDUNHW ZKLFK WKHQ XVHV LW WR HYDOXDWH WKH FRPSDQ\nV ULVN $OWKRXJK VRPH SDVW UHVHDUFK VWXGLHV KDV QRW UHDFKHG WKH VDPH FRQFOXVLRQ ZH EHOLHYH WKDW RXU UHVHDUFK KDV FRQWULEXWHG VLJQLILFDQWO\ WR WKLV DUHD RI UHVHDUFK EHFDXVH LW LV XQLJXHf§WKLV LV WKH ILUVW UHVHDUFK VWXG\ WR LQFRUSRUDWH ERWK WKH HYHQW VWXG\ PHWKRGRORJ\ DQG GDLO\ UDWHV RI UHWXUQ RQ D FRPSDQ\n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

PAGE 107

DERXW D OHDG RU ODJ SKHQRPHQRQ 5DWKHU LW LV PRUH LPSRUWDQW WR FRQVLGHU WKH ORQJWHUP HIIHFWV RI D ERQG GRZQJUDGH ZKLFK DUH ILQDQFLDO GLVWUHVV DQG SHUKDSV HYHQWXDO EDQNUXSWF\ 2XU UHVHDUFK KDV VKRZQ KRZ ERQG GRZQJUDGHV DFWXDOO\ GR SURYLGH WKH PDUNHW ZLWK QHZ LQIRUPDWLRQ RQ ILQDQFLDO GLVWUHVV DQG EDQNUXSWF\ 7KHUHIRUH UHJDUGOHVV RI ZKHWKHU WKHUH LV D ODJ RU QRW EHWZHHQ D ULVH LQ D ULVN SUHPLXP DQG D ERQG GRZQJUDGH ERQG UDWLQJV DQG GRZQJUDGHV DUH VLJQLILFDQW WR WKH PDUNHWn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f DUH HIIHFWLYH DW SUHGLFWLQJ EDQNUXSWF\ WKHLU SUHGLFWLYH SRZHU GHFUHDVHV DV ZH ORRN IDUWKHU DKHDG RI WLPH 7KXV PRGHOV EDVHG RQ VXFK YDULDEOHV VXFK DV $OWPDQnV =(7$ PRGHO ZLOO LQHYLWDEO\ ORVH SUHGLFWLYH SRZHU DV RQH WULHV WR ORRN IDUWKHU DKHDG RI WLPH

PAGE 108

0HDQZKLOH KRZHYHU YDULDEOHV WKDW DVVHVV D FRPSDQ\nV PRUH IXQGDPHQWDO FKDUDFWHULVWLFV RXU *URXS ,, YDULDEOHVf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nV GDLO\ UDWHV RI UHWXUQ WR GHWHUPLQH WKH

PAGE 109

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n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

PAGE 110

$33(1',; 5,6. 35(0,$ &859(6 ,Q WKH IROORZLQJ SDJHV ILIW\ FXUYHV DUH SUHVHQWHG 7KHVH FXUYHV VKRZ WKH ULVN SUHPLD WUHQGV RI FRUSRUDWH ERQGV LQ WKH VDPSOH RI FRPSDQLHV WKDW ZH XVHG WR VWXG\ WKH UHODWLRQVKLS EHWZHHQ ULVN SUHPLXP DQG EDQNUXSWF\ $ VROLG OLQH RQ D FXUYH LQGLFDWHV D UDWLQJ FKDQJH E\ 6WDQGDUG t 3RRUnV D GRWWHG OLQH RQH E\ 0RRG\nV 8QOHVV RWKHUZLVH LQGLFDWHG E\ DQ DUURZ SRLQWLQJ XSZDUG DOO UDWLQJ FKDQJHV DUH GRZQZDUG FKDQJHV

PAGE 111

5LVN 3UHPL?LP $OOHJKHQ\ ,QWnO 6XE 6) 'HE b f6) 'HE b f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $O

PAGE 112

5LVN 3UHPLXP $OOLV &KDOPHUV 0IJf &RUS 6) 'HE b f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 113

5LVN 3UHPLXP $PDUH[ ,QF 6XE 6) 'HE r n RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 114

5LVN 3UHPLXP $PHU +HDOWK &DUH 0JPW 6XE 1W b n )LJXUH $

PAGE 115

5LVN 3 U H P L YL P $QJOR &R ,QF *WG 6XE 6) 'HE r f )LJXUH $

PAGE 116

5LVLF 3UHPLXP $UJR 3HWUROHXP RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 117

5LVN 3UHPLXP %DOGZLQ 8QLWHG FRUS 6XE 6) 'HE b n RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 118

5LVN 3UHPLXP %$6,; &RUS 6XE 'HE r f )LJXUH $

PAGE 119

5LVN 3UHPLXP %HNHU ,QGXV 6HU 6XE 6) 'HE r n RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 120

5LVN 3UHPLXP %UDQLII $LUZD\V 6) 'HE b n RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 121

5LVN 3UHPLXP %XWWHV *DV N 2LO 'HE b f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $OO

PAGE 122

5LVN 3UHPLXP &KDUWHU &R 6XE 'HE r f W RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 123

5LVN 3UHPLXP &KHPHWURQ &RUS 6) 'HE b n )LJXUH $

PAGE 124

5LVN 3UHPLXP &ROHFR ,QGXV 6XE 6) 'HE b f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 125

5LVN 3UHPLXP &U\VWDO 2LO &R 6XE 6) 'HE f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 126

5LVN 3UHPLXP (PRQV ,QGXV ,QF (T 7U &WIV 6HU f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 127

5LVN 3UHPLXP )LUVW 5HSXEOLF %DQN 6) 'HE b f )LJXUH $

PAGE 128

5LVN 3UHPLXP )RRG )DLU 6WRUHVf 6) 'HE f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 129

5LVN 3 U H P L ?BL P )6& &RUS 6U 6) 'HE r f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 130

5LVN 3UHPLXP *DPEOHV &UHGLW 6U 1RWHV b f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 131

5LVN 3UHPLXP *OREDO 0DULQH ,QF D RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 132

5LVLF 3UHPLXP +DUGZLFNH &RnV ,QF 6XE 6) 'HE r f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 133

5LVN 3UHPL[LP ,QIRUH[ ,QF 6XE 'HE 6 f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 134

5LVN 3UHPLXP ,WHO &RUS ;V 6XE 'HE b n RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 135

5LVN 3UHLQL?LP -RKQV 0DQYLOOH 6) 'HE r f r RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 136

5LVN 3UHPLXP -RQHV t /DXJKOLQ 6WHHO 6XE 'HE b f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 137

5LVN 3UHPLXP /LRQHO &RUS 6XE 6) 'HE b f )LJXUH $

PAGE 138

5LVN 3UHPLXP /79 &RUS ;V 6) 'HE b f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 139

5LVN 3UHPLXP 0F/HDQ ,QGXV 6XE 'HE b f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 140

5LVN 3UHPLXP 0HJR ,QWnO 6XE 6) 'HE b f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 141

5LVN 3UHPLXP 0LFKLJDQ *HQHUDO ;V 6U 6) 'HE r f )LJXUH $

PAGE 142

5LVN 3UHPLXP 0LVVLRQ ,QV *URXS 6) 'HE b f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 143

5LVN 3UHPLXP 0RUWRQ 6KRH &RV ,QF 6U 6) 'HE r f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 144

5LVN 3UHPLXP 1RUWK $PHULFDQ &DU &RUS (TXLS 7U b n RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 145

5LVN 3 U H P L ?BL P 3HWWLERQH &RUS ;V 6XE 6) 'HE r f )LJXUH $

PAGE 146

5LVN 3UHPLXP 3XEOLF 6HUYLFH 1HZ +DPS VW 9 n RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 147

5LVLF 3UHPLXP 5DGLFH &RUS 6XE 6) 'HE b f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 148

5LVN 3UHPLXP 6KDURQ 6WHHO 6XE 6) 'HE f m RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 149

5LVN 3UHPLXP 6PLWK ,QWnO 6) 'HE r f )LJXUH $

PAGE 150

5LVN 3UHPLXP 6WRUDJH 7HFKQRORJ\ 1WV b f r RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 151

5LVOF 3UHPLXP 6XQEHDP &RUS 6) 'HE r f )LJXUH $

PAGE 152

5LVN 3UHPLXP 7H[DFR &DSLWDO ([WnG 1W r f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 153

5LVOF 3UHPLXP 7H[DFR ,QF 'HE f )LJXUH $

PAGE 154

5LVN 3UHPLXP 7H[DV ,QWL $LUOLQHV ;V 6XE 'HE r n RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 155

5LVN 3UHPLXP 7RGG 6KLS\DUGV 6U 6XE 1WV b f r RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 156

5 L V OF 3 U H P L ?L P :HGWHFK &RUS 6U 6XE 1WV b n RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 157

5LVN 3UHPLXP :HVWHUQ &R 1R $PHULFD [V 6XE 'HE r f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 158

5LVN 3UHPLXP :KLWH 0RWRU &RUS 6) 'HE b f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 159

5LVN 3UHPLXP :LFNHV &RUS 6) 'HE Eb f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 160

5LVOF 3UHPLXP :LOVRQ )RRG 'HE f RI 0RQWKV EHIRUH )LOLQJ &KDSWHU )LJXUH $

PAGE 161

5()(5(1&(6 $OWPDQ ( f &RUSRUDWH %DQNUXSWF\ 3RWHQWLDO 6WRFNKROGHU 5HWXUQV DQG 6KDUH 9DOXDWLRQ -RXUQDO RI )LQDQFH $OWPDQ ( 5 +DOGHPDQ DQG 3 1DUD\DQDQ f =(7$ $QDO\VLV $ 1HZ 0RGHO WR ,GHQWLI\ 5LVN RI &RUSRUDWLRQV -RXUQDO RI %DQNLQJ DQG )LQDQFH %HDYHU : f )LQDQFLDO 5DWLRV DV 3UHGLFWRUV RI )DLOXUH (PSLULFDO 5HVHDUFK LQ $FFRXQWLQJ 6HOHFWHG 6WXGLHV 'HUNLQGHUHQ ) DQG 5 &UXP f 'HYHORSPHQW DQG (PSLULFDO 9DOLGDWLRQ RI 6WUDWHJLF 'HFLVLRQ 0RGHOV ,QWHUQDWLRQDO 6WXGLHV RI 0DQDJHPHQW DQG 2UJDQL]DWLRQ 6SHFLDO ,VVXH RQ 6WUDWHJLF 0DQDJHPHQW )DPD ( ) f )RXQGDWLRQV RI )LQDQFH %DVLF %RRNV ,QF 1HZ
PAGE 162

.DW] 6 f 7KH 3ULFH $GMXVWPHQW 3URFHVV RI %RQGV WR 5DWLQJ 5HFODVVLILFDWLRQV $ 7HVW RI %RQG 0DUNHW (IILFLHQF\ -RXUQDO RI )LQDQFH 0DVXOLV 5 f 7KH (IIHFWV RI &DSLWDO 6WUXFWXUH &KDQJH RQ 6HFXULW\ 3ULFHV $ 6WXG\ RI ([FKDQJH 2IIHUV -RXUQDO RI )LQDQFLDO (FRQRPLFV 0RQWJRPHU\ & f 'HVLJQ DQG $QDO\VLV RI ([SHULPHQWV QG (GLWLRQ -RKQ :LOH\ t 6RQV ,QF 1HZ
PAGE 163

fA:LOFR[ f $ *DPEOHUnV 5XLQ 3UHGLFWLRQ RI %XVLQHVV )DLOXUH 8VLQJ $FFRXQWLQJ 'DWD 6ORDQ 0DQDJHPHQW 5HYLHZ =DYJUHQ & f $VVHVVLQJ WKH 9XOQHUDELOLW\ WR )DLOXUH RI $PHULFDQ ,QGXVWULDO )LUPV $ /RJLVWLF $QDO\VLV -RXUQDO RI %XVLQHVV )LQDQFH t $FFRXQWLQJ =PLMHZVNL 0 f 0HWKRGRORJLFDO ,VVXHV 5HODWHG WR WKH (VWLPDWLRQ RI )LQDQFLDO 'LVWUHVV 3UHGLFWLRQ 0RGHOV -RXUQDO RI $FFRXQWLQJ 5HVHDUFK 6XSSOHPHQW

PAGE 164

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

PAGE 165

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

PAGE 166

81,9(56,7< 2) )/25,'$