Essays on new equity issues and investment banking contracts

MISSING IMAGE

Material Information

Title:
Essays on new equity issues and investment banking contracts
Physical Description:
vii, 197 leaves : ill. ; 28 cm.
Language:
English
Creator:
Bae, Sung Chul, 1955-
Publication Date:

Subjects

Subjects / Keywords:
Investment banking   ( lcsh )
Securities -- Costs   ( lcsh )
Finance   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1987.
Bibliography:
Includes bibliographical references (leaves 191-196).
Statement of Responsibility:
by Sung Chul Bae.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001030273
notis - AFB2391
oclc - 18130863
System ID:
AA00002152:00001

Full Text

















ESSAYS ON NEW EQUITY
AND INVESTMENT BANKING


ISSUES
CONTRACTS


SUNG


CHUL


BAE


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


1987























my parents


their


unending


love


and


support















ACKNOWLEDGEMENTS


am deeply


committee


indebted


chairman,


to Dr.


his


Haim


insight,


Levy,


numerous


dissertation


suggestions


corrections,


and


patient


guidance,


of which


have


greatly


contributed


quality


this


dissertation.


would


also


like


thank


other


members


committee,


. Robert


Radcliffe,


Roger


Huang,


and


Stephen


Cosslett,


their


support


and


guidance.


Special


thanks


are


also


due


to Dr.


Richard


Pettway,


who


guided


me through


early


stage


doctoral


program.


Thanks


are


also


extended


colleagues,


Young


Hoon


Byun


Pil


Lim


their


comments


numerous


discussions


and


Jim


Yoder


time


proof


reading.


have


also


benefited


from


comments


participants


finance


workshops,


University


Florida.


Finally,


but


foremost,


am indebted


my wife,


Kwangmi,


her


devoted


support


and


encouragement


throughout


academic


career


and


daughters,


Hana


and


Gina,


their


patience


with


me.
















TABLE


OF CONTENTS


PAGE


ACKNOWLEDGEMENTS


ABSTRACT


CHAPTERS


GENERAL


INTRODUCTION


. . 1


Scope
Tasks


Dissertation.
Dissertation.


THE


VALUATION OF
EQUITY ISSUES


UNDERWRITING
: AN OPTION


CONTRACT FOR NEW
PRICING APPROACH.


Introduction
Background


Objecti
The Valua
Direct
Graphic
Compari
Numeric
Und
Data .
Sample
Descrip
Empirical
Estimat
Results
Results
Results
The Eff
on
Compari
Pre
Implica


rA~


S
.I -~ -


S .


of the
on Model
rivation


Study.
Study .


O


1 Derivatio
on of the M
1 Analysis


f
f
n
od
of


nde:
he
f t]
1 w.
the


S. . 8


. . . .12
writing Contract .15
Model . . .15
he Model. . .. .23
ith Smith's Model .26
Value of


r


03
ic
Ti
03
.o
-


writing Contract. . . .3
. . . . . .3
instruction. . . . .3
on of Data. . . . .3
techniques and Results . . .4
n of Relevant Variables . .4
Upper Bound of Underwriting Value. .5
Lower Bound of Underwriting Value. .5


- Exce
cts of
nderwr
on of
ious F
ions o
SA -


SS


a


nd
E
'C


Underwr
rm/Offe
ng Valu
Empiri
.ings on
mpirica


iting Risk Premium
ring Variables
e. . .
cal Results with
Rights Offers .
1 Results. .
..amJ &.: 4: 1?.. -1 n flC


. .58

. .62

. .73
. .76


. . 1
. . . 2


. 8


S. . . 111


S . * * V









THE


CHOICE OF AN OPTIMAL FINANCING
UNDER ASYMMETRIC INFORMATION
AN EMPIRICAL INVESTIGATION .


METHOD


. . .82


Introduction
Background
Objective
Literature R
The Theory a
Data .
Sample Con
Descriptiv
Research Des
Estimation
Estimation
Measuremen
Empirical Re
Abnormal R
Results -
Results -
Results -
Implicatio
Conclusion .


o
e
n

s
e
i


t
s
e
F
S
F
n


of the Study. . . .. .8
f the Study . . . .8
view. . . . .. .8
d Testing Hypotheses. . .. .9

truction. . . .. 10
Statistics of Data . . 10
gn. . . . 11
of Firm Quality . . 11
of Investigation Costs. . 11
of Subscription Price Discount. 1
ults. . .* . 11
turns Around Announcement Day 11
irm Quality Hypothesis. . 12
ubscription Price Hypothesis. 12
irm Segment Hypothesis. . 12
,s of Empirical Results. . 13
* . . . 13


AN EMPIRICAL


INVESTIGATION


OF THE


VALUATION


RIGHTS AND
THE RIGHTS


THE EFFICIENCY
OFFERING MARKET


OF
S. . 137


Introduction
The Valuatio
The Methodol
Structure
Data and S
The Empirica
Analysis o
Analysis o
Analysis o
Market
Conclusion .


n
o0
o:
ai
1
f
f
f


. a a . . 1
Models of Rights . . 1
gy. . . . . 1
f the Empirical Study ... 1
mple Selection. . . 1
Results. . . 1
the Valuation Models of Rights 1
the Characteristics of Riahts. 1


the


Impact on


Pricing Mechanism . . 177
S . . . . 183


SUMMARY


AND


CONCLUSIONS.


. . 185


APPENDIX


PARTIAL


DERIVATIVES


OF UNDERWRITING


VALUE.


. 189


BIBLIOGRAPHY


. . . . . 1 191















Abstract


Dissertation


:he University
Requirements f


ESSAYS


AND


Presented


Florida


Degree


ON NEW


INVESTMENT


Partial


Doctor


EQUITY
BANKING


Graduate


School


Fulfillment


Philosophy


ISSUES
CONTRACTS


Sung


Chul


Bae


August


1987


Chairman:


Major


Haim


Levy


Department:


Finance,


Insurance,


and


Real


Estate


The


purpose


this


dissertation


to enhance


current

investme


understanding

nt banking co


new


ntracts.


public

Three


equity

issues


offers

are a


and


analyzed


this


ssertation.


The


first


issue


focuses


on the


relationship


between


underwriting


sk and


underwriter


s compensation


seasoned


new


stock


offerings.


A valuation


model


most


widely


used


firm


commitment


underwriting


contract


developed


employing


option


pricing


framework


Applying


model


market


data


on underwriting


spread


over


period


1982-


1985,


underwriting


value


was


found


to be highly


overpriced,


indicating


that,


on average,


underwriters


earned









underwriting


market


is less


than


perfectly


competitive


and


that


observed


excess


returns


are


mainly


monopsonistic


underwriting


business.


The


second


financing


issue


method


focuses


among


on the


standby


choice


rights


offer,


an optimal

nonstandby


rights


offer,


and


firm


commitment


cash


offer


a market


asymmetric


information.


This


study


empirically


examines


three


testable


signalling


hypotheses


equilibrium


implied


model.


Using


Heinkel


data


and


Schwartz


on primary


stock


offering


over


1976-1985


period,


we found


that


empirical


results


not


fully


support


implications


three


hypotheses.


argue


that


analytical


problems


present


Heinkel


and


Schwartz


model


induce


empirical


evidence


to deviate


from


predictions


their


model.


The


third


issue


devoted


valuation


rights


efficiency


considered


as call


rights offering

options, either


market.

on the


Rights


firm'


can D

assets


Smith


model)


or on the


firm


s stock


(Galai


and


Schneller


model)


The


predictability


these


rights


valuation


models


investigated


using


market


data


rights


over


1968


-1985


period.


We found


that


both


models


significantly


overestimate


market


value,


but


that


Smith


model


yields


slightly


better


results.


The


overall


results


suggest


that


rights


offering


market


not


as efficient


e option















CHAPTER


GENERAL


Scope


INTRODUCTION


Dissertation


Corporations,


whose


objective


is to increase


value


shareholders


stake


firm,


typically


face


basic


decisions


investment,


or capital


budgeting,


decision--


how


much


to invest


what


specific


assets


invest


financing


decision--how


raise


necessary


cash


investment.


Finance


literature


has


focused


on two


broad


issues


related


firm'


financing


decision;


firm


s choice


security


to offer


examining


market


reactions


announcement


new


security


offer


and


firm


s choice


an offering


method


analyzing


structure


various


arrangements


latter


employed


which


explicitly


capital


discussed


sing


process,


this


dissertation.


Corporations


raise


external


equity


capital


a number


different


ways.


new


equity


issue


can


offered


either


directly


firm


s existing


shareholders


1-^


rights


Lr -9 -- *


offer)
- 2 .


I


. r 1- _


I 1









Public


Rights
Offer
(3%)a
I
i
Nonunderwritten
(Uninsured)


Rights


offer


(2%)
I
-I-


Underwritten
(Standby)


offer


Rights
(1%)


Offering
I
-i _____


I
Shelf-


Underwritten


Cash


Offer


(82%)


Registered
(15%)


Firm Commitment
(Fully underwritten)
(80%)


Best
Efforts
(2%)


_____1


(78%)


I
Nonunderwritten


Offer


(2%)


Underwritten


Offer


(83%)


Notes


= competitive


contract.


= negotiated


contract.


a The proportion
method relative


the 1
drawn
the S


976-1985


from


securities


the
the


number
total


period is i
Registered


and


Exchange


number


n parenthe
Offering


offerings u
r of public


ses


The


Statistics


sing


each


offerings


over


information


provided


Commission.


Alternative


Seasoned


New


Figure
Methods


Equity


Public


Issues


Offerings of
Corporations


This


dissertation


analyzes


various


characteristics


contractual


arrangements


employed


raising


equity


capital


illustrated


Figure


1-1.


It consi


three


separate


essays


whose


topics


are


related


but


distinct.


Tasks


Dissertation


,-- An1 r.


L1-__--


-a A 4-4 :


L1 B L


r r









While


previous


studies


recognized


through


regression


analysis


effect


seven


factors


on the


overall


under-


writer


s compensation,


namely


underwriting


spread,


they


often


overlooked


fact


that


underwriting


fee,


rather


than


underwriting


tion


spread,


represents


underwriting


risk


more


involved


appropriate


in a new


compensa-


equity


offering.


Moreover,


little


analytic


attention


so far


given


magnitude


explicit


equilibrium


underwriting


risk


premium.


this


essay,


we first


employ


option


price


framework


to derive


a valuation


model


underwriting


risk


premium


most-widely


used


firm


commitment


contract.


We also


present


numerical


examples


examine


relative


effect


changes


model


parameters


interest


value


underwriting


risk


premium.


then


analyze


underwriting


risk


premium


seasoned


new


equity


issues


applying


our


model


to market


data


on underwriter


compensation-tion


over


period


1982-1985.


Specifically,


this


essay


attempts


answer


following


ques


tions.


How


should


optimal


underwriting


risk


premium


determined


and


how


much


should


issuing


firms


on average


overpay


risk


that


they


transfer


to underwriters









significantly


group,


different


type


manner


registration


depending


shelf


upon


vs.


industry


non-shelf),


method


syndication,


and


relative


position


"prestige"


often


called)


managing


underwriter


industry?


These


tioners


issues


such


are


issuing


significant


firms


importance


underwriters


to practi-


as well


regulatory


agencies


, which


have


been


concerned


about


competition


in underwriting


business


new


equity


offerings.


The


empiri


results


this


essay


will


provide


considerable


insights


into


competitiveness


new


equity


issue


market


and


into


issuing


firm/underwriter


s policy


making.


The


detailed


analysis


and


empirical


results


Essay


are


presented


Chapter


Essay


investigates


choice


an optimal


financing


method


under


asymmetric


information.


choice


underwritten


cash


offers


seasoned


new


equity


issues


vast


majority


U.S


. firms


over


the


seemingly


less


expensive


rights


offers


been


referred


so-


called


"equity


financing


paradox"


Although


some


potential


explanations


to resolve


this


paradox


have


been


provided


several


previous


studies,


theoretical


models


with


rational


expectation


formation


have









information as

investors/inve


ymmetry

stment


between


banker


issuing


which


firms


may


and


underlie


outside

the


observed


behavior


firms


choosing


their


optimal


financing


method.


Based


on the


theoreti


analysis


Heinkel


Schwartz


1986),


this


essay


empirically


examines


three


testable


hypotheses


employs


regarding


a given


whi


issue


type


under


financing


information


method


asymmetry


a firm


- (1)


firm


quality


hypothesis,


subscription


price


hypothesis


and


firm


segment


hypothesis.


This


essay


also


discusses


some


analytical


problems


pre


sent


in the


Heinkel


and


Schwartz'


signalling


equilibrium


model.


Specifically,


this


essay


seeks


answer


following


questions.


issuing


firm


s choice


a financing


method


significantly


related


quality


type)


firm?


What


type


firm


using


non-standby


uninsured)


rights


offers


sets


a higher


scription


price


Does


magnitude


investigation


costs


ass


ociated


with


standby


rights


offers


affect


number


firms


using


this


method?


What


other


alternative


type


financing


method


used


those


firms


which


avoid


standby


rights


offers


because


high


investigation


costs?








Essay


devoted


pricing


rights,


which


can


viewed


as options


such


that


holder


has


right


buy


firm


s new


shares


on specified


terms.


This


essay


has


three


distinct


parts.


First,


we develop


a per-share


version


a rights


valuation


model


derived


Smith


(1977),


which


adapts


minor


modification


boundary


condition


Black


Scholes


(1973)


simple


call


option


pricing


model


reflect


equity


dilution


induced


exercise


rights.


then


compare


per-share


version


Smith


model


with


alternative


one


derived


Galai


and


Schneller


(1978)


This


study


shows


that


Galai


and


Schneller


model


yields


identical


values


rights


Smith


model


only


certain


limited


case.


Second,


test


predictability


these


two


models


examine


effi


ciency


rights


offering


market


using


market


data


rights


issues


on both


rights-on


days


ex-right


days


over


1968-1985


period.


further


investigate


magnitude


empirical


deviations


model


price


from


observed


market


price


considering


some


potential


factors


which


may


explain


observed


biases.


Since


Smith


s model


assumes


all-equity


firm


and


simultaneous


exercise


of rights,


empirical


results


this


essay


would


allow


us to indirectly


investigate


effect









Third,


test


whether


appearance


Smith


s model


Galai


and


Schneller


s model


1977


and


1978,


respectively,


have


narrowed


gap


between


theoreti


and


served


market


values


rights


The


implicit


hypothesis


that


practitioners


who


use


these


formulas


may


pushing


market


values


close


theoretical


values


rights


The


detailed


analysis


and


empirical


result


Essay


are


presented


Chapter


Chapter


V summari


zes


and


concludes


study


bringing


together


results


three


essays















CHAPTER


THE


VALUATION


EQUITY


ISSUES


OF UNDERWRITING


: AN


OPTION


CONTRACT


PRICING


FOR


NEW


APPROACH


Introduction


Background


Study


Corporations


typically


employ


investment


bankers--or


underwriters,


they


are


often


called--to


raise


new


equity


capital.


Two


alternative


forms


typical


underwriting


contracts


are


a best


efforts


contract


and


a firm


commitment


contract.


Previous


studies


provide


evidence


predominant


use


a firm


commitment


contract


seasoned


new


issues


stocks.


Booth


and


Smith


1986)


report


that


over


percent


commitment


contract


equity offers

s during the


were


period


ssued

of 19


7


via f

7-198


irm


Smith


(1977),


Bhagat


(1983),


and


periodic


Securities


and


Exchange


Commission

Unlike


studied

a best


(1986)


efforts


document

contract,


similar


which


evidence.


an investment


banker


acts


only


as a marketing


agent


to sell


issue


on a


fixed


basis


without


taking


any


risk


involved


new


issue


offering,


banker


a firm


commitment


contract


vir~nc


nflt+'


+-n cvl li


tahv


f i rmc


nrnr9 nm nHanitl


ml,,


o-hiirltr


4-virt


I1


F


I









agrees


to purchase


entire


issue


from


firm


and


bears


risk


regarding


proceeds


from


issue.


performs


three


distinctive


financing


services


originating


managing


the


new


issue


s offering


terms,


timing


issue,


and,


necessary,


recruiting


other


underwriters


, (2)


underwriting


and


absorbing)


risk


adverse


price


securities


developments,


to inve


stores.


and


The


term


distributing

"underwriting"


often


used


to refer


to all


three


financing


services.


As compensation


these


services,


an investment


banker


receives


underwriting


spread


between


price


that


issuing


firm


paid


stock


and


price


at which


stock


underwriting


sold


spread


public,


typically


"offering"


consist


price.


managing


The


fee,


underwriting


fee,


and


selling


concession.


Naturally,


issuing


firm


wants


to receive


highest


possible


price


stock,


bargains


with


banker


over


both


offering


price


and


spread.


higher


offering


price


and


lower


spread,


more


issuer


gets


per


share


stock


issued.


The


spread


percentage


generally


reflects


issue


numerous


quality


considerations


type,


such


bargaining


as market


power


conditions,


involved


: (1)









parties,


and


anticipated


difficulty


selling


proposed


issue.


The


subject


underwriter


s compensation


is permeated


with


idea


risk


that


an underwriter


bears


unsuccessful


offering.


Since


market


price


on the


offer


date


uncertain,


there


exists


a possibility


that


underwriter


offer


would


price


not


thus


be able


bear


to sell


an unexpected


whole


loss


issue


from


compensation.


three


financing


services,


frequently


noted


that


underwriting


a narrow


sense)


constitutes


taking


risk


to provide


long-term


financing


issuing


firm,


namely,


an insurance


service


of bearing


risk


of adverse


price


fluctuations.


This


study


focuses


relationship


between


under-


writing


service


new


equity


issues


and


underwriting


risk


premium


as appropriate


compensation.


From


this


relationship,

compensation

underwriting


we assess

represents


risk


whether

the fai


premium.


observed


r.


It is


competitive


particular


underwriting


price


interest


examine


put


whether


option


underwriting


premium


accounts


risk


premium


observed


represented


gross


underwriting


spread


or the


underwriting


or a certain


amount


between


these


types


underwriting


compensation.


Tn nsn nral


IT at- nrr 1 M nan


rnnqc-j 1-11t0tQP


^r\ ^


I || I9


Y


r 1|









Several


previous


studies


have


investigated


underwriter


compensation


as a function


various


characteristics


inves


tment


banking


firms


and


new


issue


offered.


They


conclude


that


a major


portion


underwriting


spread


a new


issue


represents


underwriter


s compensation


bearing


risk


an unsuccessful


offering.


Logue


Lindvall


1974)


examine


trade-off


between


offering


price


and


cash


spread


and


identify


several


factors


influencing


this


relationship


Similarly,


Stoll


(1975)


considers


determinants


underwriting


spread.


study


Ederington


1975)


shows


that


underwriting


spread


on a


competitive


bidding


corporate


bond


issue


varies


directly


with


underwriter


s uncertainty


regarding


demand


new


issue.


Bhagat,


Marr,


and


Thompson


1985


document


that


underwriting


spread


in seasoned


new


equity


issues


sk an

the


is significantly

d significantly


issuer


negatively

positively


syndicated


offerings,


related


systematic


unsystematic


but


risk


insignificant-


related


to both


in nonsyndicated


offerings.


While


regression


overall


most


these


analyses


underwriter


studies

effect


compensation,


recognized


several


namely


through

factors


underwriting


spread,


fee,


they


rather


often


than


overlooked


underwriting


fact


spread,


that


underwriting


represents


a more








writing


risk


underwriting


risk


premium


a firm


commit-


ment


equity


offers


How


should


value


underwriting


contract


determined?


Which


type


observed


underwriting


compensation


represents


fair


price


underwriting


risk


premium?


Does


issuing


firm


overpay


risk


that


transfers


underwriter?


These


issues


are


importance


practitioners


such


ssuing


firms


and


underwriters


as well


as regulatory


agencies,


which


have


been


concerned


about


market


competition


and


concentration.


Objective


Study


this


study,


above


issues


are


explicitly


examined


both


theoretically


and


empirically


through


an equilibrium


valuation


model


underwriting


contract.


For


this


purpose,


derive a

contract


we first


valuation


employ


model


context


option


a firm


pricing


commitment


a negotiated


sale.


framework


underwriting


then


analyze


underwriting


value


seasoned


new


equity


issues


applying


our


model


to market


data


period


1982-


1985.


The


value


underwriting


contract


is further


examined


partitioning


total


sample


into


several


subgroups


on the


basi


issue


character


stics.


The


feasibility


empirically


applying


our


model


to unseasoned


new


equity


issues


initial


public


offerings)


then


discussed.









factors


such


as underwriting


expenses,


issue


price,


and


price


to be


paid


issuer


held


constant,


value


underwriting


contract


can


regarded


as a contingent


claim


on the


value


new


share.


The


underwriting


combination


a long


contract


position


shown


to be


in stock,


equivalent


a certain


payment


issuing


firm,


and


a short


position


a European


call


option


with


an exercise


price


equal


public


offering


price.


It is


also


shown


to be


analogous


to a short


position


a put


option


plus


underwriting


compensation.


Thus,


derived


model


value


underwriting


contract


explicitly


yields


excess


returns


to underwriters,


namely


deviation


observed


underwriter'


compensation


from


fair


value


underwriting


risk


premium.


underwriting


which


formula,


in our


explicitly


or equivalently


model


present


put


primarily


call


option


equity


option


formula,


variance


stock


return.


Thus,


with


greater


volatility


stock


calls


prices


will


around


rise


offer


substantially,


date,


premiums


which,


turn,


puts


will


cause


a decrease


underwriting


value.


Those


factors


which n


lay


influence


underwriter


4 r n n a


C:~ nf -rnn


s payoff


nurrnu~: PA


A A r~A r I CA








Our


model


similar


one


derived


Smith


1977),


a different


approach


along


with


a simple,


graphical


explanation


employed


deriving


our


model.


We present


theoretical


similarities


practical


and


grounds


differences


as a basis


between


discussing


models.


argue


that


Smith


s model


difficult


use


context


a negotiated


examine


sale.


relative


We also


effects


present


changes


numerical


in the


examples


six


parameters


of interest


on the


value


underwriting


contract


and


to illustrate


difference


between


our


model


and


Smith


s model.


Overall,


this


study


seeks


answer


following


questions:


How


should


an investment


banker


or issuing


firm


estimate


fair


value


underwriting


risk


premium


sk that


underwriter


bears


Practically,


which


type


erved


underwriting


compensation


better


represents


fair


price


of underwriting


risk


premium?


investment


bankers


new


U.S.


equity


1ssue


significantly


offerings?


overpaid


underwriti


sk taking

ng industry


competitive


with


respect


to its


pricing


underwriting


risk


premiums?


Are


underwriters


compensated


a significantly


different


manner


depending


upon


issuing


firm


industry


sector,


method


type


of registration


syndication


syndicated


(shelf


or nonshelf),


or nonsyndicated),


and











The


Valuation


Model


Underwriting


Contract


In thi


Section,


we first


derive


valuation


model


underwriting


contract


employing


contingent


claim


technique


under


standard


assumptions


Black


and


Scholes


(1973)


option


pricing


model.


then


provide


simple,


intuitive,


graphical


derivation


model.


further


discuss


model


one


similarity


derived


and


Smith


difference


(1976).


between


This


our


part


concludes


quantitative


presenting


effects


numerical


examples


changes


to analyze


relevant


model


parameters


on the


value


underwriting


contract


and


compare


our


model


with


Smith


Direct


Derivation


Model


It is


well


known


that


when


a firm


makes


a rights


issue


through


a standby


contract,


firm


actually


buys


insurance


contract,


or equivalently


a put


option


giving


right


to sell


unsubscribed


shares


underwriter.


Since


underwriter


this


case


sells


a put


option,


underwriting


identical


compensation


value


banker,


put


or the


option.


standby


In this


fee,


form


Strictly


speaking,


underwriting


U.S.


cmnss ts


t flw


anrihv


mni


ar n i i* i nn


*nartsT


I I


CICI


. JI I W


ILJ








agreement,


risk


shared


issuing


firm


and


underwriter.


A firm


commitment


contract


similar


limiting


case


a standby


agreement


to rights


issues


where


issuing


firm


completely


fails


to sell


shares,


but


still


receives


a prespecified


amount


from


underwriter.


In a firm


commitment


contract,


issuing


firm


guaranteed


a fixed


amount


money


no matter


what


happens


share


price


on the


offer


date


simply


selling


all


shares


underwriter


offer


price


underwriting


spread.


share


price


issuing


firm


drops


unexpectedly


on the


offer


date,


new


shares


cannot


issued


offer


price


and


thus


underwriter


bears


an unexpected


loss


from


his


compensation.


share


price


rises


unexpe


ctedly,


underwriter


however,


prohibited


from


selling


new


issue


above


fixed


offer


price


National


Association


Securities


Dealers


(NASD)


Rules


Fair


Practice.


This


relationship


can


summarized


as follows.


Smith


which


1977)


and


underwriters


extract


issues
which
extrac


that


Ibbotson


may


profits


However,


suggest
t those


Smith


that


profits.


NASD'


underwriter


assoc


(1979)


1975)


discuss


to circumven


iated


underwriters


with


points


can


Throughout
a binding


a discussion


this


several
t the


l1 ways
NASD's


in
rule


oversubscribed


indirect


evidence


systematically


study,


constraint


we assume


on the


underwriter


s position


when


I 1 *I I t


ill f


narneccnri lu


3 rntnrer 4i ni-


eon


Cmi -H"


SIi c









successful


unsuccessfulu


offering)

1 offering)


(2-1)


where


= value


underwriting


contract


on the


offer


date


= stock


price


on the


offer


date


= offering


price


= price


share


that


an underwriter


obligated


pay


issuing


firm


Note


that


can


be either


positive


or negative


depending


upon


whether


or S


respectively


Merton


1974)


shown


that


any


contingent


claim


whose


value


depends


upon


underlying


asset,


, and


time,


must


sati


following


partial


differential


equation.


-- =


02V2


r V


-r f


where


V,t


value


of contingent


claim


as a function


V and


Since


value


an underwriting


contract


a function


both


time


between


contract


date


and


offer


date


and


underlying


stock


price,


solution


Equation


where


S,t)


= f(V,t


subj


ect


to a boundary


condition,


which


drawn


from


Equation


-1).


= Min


- B,


- B]


= Min


2-3)


This


boundary


condition


states


that


underwriting


contract


on the


offer


date


worth


minimum


offer








price


or the


prevailing


market


price,


net


price


to be


paid


to issuing


firm.


Assuming


that


distribution


possible


stock


price


changes


on the


offer


date


lognormal


addition


to the


usual


standard


assumptions


Black


and


Scholes


option


pricing


model


1973)


allows


value


underwriting


contract


a risk-neutral


case


to be


expressed


= E(U*)


- B]


S*)dS*


- B]


S )dS*


(2-4)


where


E denotes


expected


value


and


log-


normal


probability


density


function.


The


above


equation


can


rewritten


S*)dS*


S*)dS*


S )dS


Instead


directly


solving


Equation


one


can


easily


solve


noting


similarity


between


boundary


condition


and


expected


value


formula


underwriting


contract


(Equations


and


-5))


and


those


a pure


discount


debt.


essence,


issuing


debt


equivalent


to stockholders


selling


firm'


assets


debtholders


value


debt,


plus


a call


option


to repurchase


+ X









firm'


assets


from


the


debtholders


with


an exercise


price


equal


face


value


debt,


see


Smith


1979)).


Since


equity


option,


boundary


a firm,


analogous


condition


value


to a call


debt,


- E*


= Min


where


= value


debt


maturity


date


debt


= value


firm


assets


maturity


date


= value


equity


maturity


date


= face


value


debt


The


value


debt


solution


to Equation


where


D(V,t)


=f(V,t)


subject


to Equation


-6).


This


solution


technique


yie


V L


'(V*)dV*


L'(V)dV* ]


Thus,


value


debt


derived


see


Smith


1979))


= V N(-d1)


P N(d2)


where


- V N(dl)

= [in(V/P)


P N(


(2-8)


aJT


= dl


= time


to maturity


debt


= variance


returns


on the


firm


total


assets


Note


that


first


two


terms


on the


right


side


Equation


are


analogous


to Equation


except


counting


factor.


V*


+ P


= V


, P]


- aoT









Equation


(2-5),


we obtain


following


closed


form


solution


value


= erT


underwriting


- S N(gl)


contract


X N(g


= erT


- erT


N(g1


- e


X N(g2)]


2-9)


where


[in(


S/X)


aJT


- aJT


= riskless


rate


interest


= time


between


contract


date


and


offer


date


= variance


stock


returns


issuing


firm,


and


other


variable


are


defined


as before.


Note


that


since,


unlike


valuation


debt


as an


option


Equation


current


discounting


value


underwriting


not


needed


contract


see


to find


Equation


-5)),


we simply


multiply


Equation


(2-8)


the


reciprocal


discounting


factor


-rT


to offset


discounting


effect.


value


an underwriting


contract


can


be shown


to be


equivalent


to a combination


a long


position


new


share,


a cash


position


payment


a European


issuing


call


option


firm,


with


and


a short


an exercise


price


equal


offer


= erT


price.

SerT


2-10


Note


that


underwriter


sell


a "synthetic"


call


option


public


public


offering


price


and


implici


= gl91


, IT;









option


values,


call


option


Equation


is quite


sensitive


volatility


stock


prices.


Thus,


with


greater


volatility


stock


prices,


call


premiums


will


rise


substantially,


which


will,


in turn,


cause


a decrease


value


contract,


and


vice


versa.


In general


form,


this


can


expressed


2-11)


where


SU/ as,


aU/ax


aU/802


aU/ar


and


aU/ aB


The


partial


effects


are


expected


direction


have


intuitive


interpretations:


An increase


stock


price


dire


ctly


increases


probability


successful


offering


thus


offering


increases


price,


value


expected


contract.


payoff


With


call


a higher


option


decreases


thus


underwriter


s payoff


increases


because


higher


offering


price


directly


widens


underwriting


spread.


An increase


in the


stock


volatility


decrea


ses


underwriter


s payoff,


which


can


seen


directly


from


Equation


2-10);


underwriter


a short


position


tion


in a call


option.


an underwriter


Since


s payoff


probability


positively


distribu-


skewed


and


limited


underwriting


spread,


increase


stock


volatility


affects


left


side


stribution


more


significantly


and


thus


increases


underwriting


risk.


= U(


, X,


, r,


, SU/ T









greater


value


contract.


A dollar


increase


payment


issuing


firm


directly


causes


a dollar


decrease


underwriter


s payoff.


There


are


two


effects


value


underwriting


contract


an increase


time


to offer


date,


either


which


can


dominate:


increase


time


to offer


date


increases


future


value


underwriting


spread;


an increase


time


to offer


clearing


date


price,


increa


and


ses


thu


uncertainty


worsens


about


underwriter


market


s position.


now


deriving


investigate


valuation


standard


formula


assumptions


an underwriting


employed


contract


see


whether


these


assumptions


are


reasonable


context


underwriting


a new


equity


issue.


Since


final


paid


typically


underwriting


issuer


signed


and


day


agreement


on the


intended


before


final


public

actual


price


offering

offering


to be


price

date,


underwriters


are,


in general,


required


pay


issuing


firm


within


four


days


time


when


offering


actually


begins


see


Perez


(1984,


68)),


underwriting


call


option


a short


duration.


Thus,


seems


reasonable


assume


that


stock


return


variance


constant


that


interest


rate


known


and


constant


during


this


short


period.


Also,


firms


pay


no dividends


during


life


contract.


Therefore,


assumptions


employed









Furthermore,


previous


studies


have


shown


that


Black


Scholes


call


option


pricing


model


quite


robust


with


respect


underlying


assumptions


and


that


discre-


pancies


between


predi


cted


call


prices


and


observed


market


prices


are


signifi


cant.


Graphical


Derivation


Model


In this


subpart,


we present


a simple


, graphical


approach


to deriving


our


model


without


explic


itly


employing


Equations


2-3)


Following


Equation


2-1),


relationship


between


value


underwriting


contract,


and


stock


price,


offer


date


given


the


line


segments


and


Figure


2-1.


Note


that


s i nce


underwriter


cannot


sell


new


issue


above


offer


price,


even


when


greater


than


value


underwriting


contract


limited


line


segment


which


represents


difference


between


or the


underwriting


spread.


Galai


1983)


of empirical


Black


and


tests


Scholes


Smith


option


simple


1976)


call


provide


pricing
option


excellent


models
pricing


surveys


including
formula.















X-B






a


B X


Notes


= value
= value


= offering


= price


per


an underwriting


a new
price


share


share


to be


contract


on the


paid


offer


date


issuing


firm


The


Value


Figure 2-1
>f Underwriting


Contract


From


Figure


2-1,


obvious


see


that


value


underwriting


contract


a portfolio


a short


position


a put


option


and


a fixed


amount,


- B,


contract


date.


put


option


is written


new


share.


expires


on the


offer


date


and


has


an exercise


price


equal


offer


price,


It follows


that


value


underwriting


contract


contract


date


= erT


- B)


(2-12)


where


value


a put


option.


It is


quite


important


to note


difference


discounting


between


typical


put


option


contract


on the









contract.


term


on the


right-hand


side


Equation


(2-12)

short


represents


position


the u

a put


sual


payoff


option.


function


This


payoff


involving

function


profiles


an immediate


cash


receipt


from


writing


a put.


However,


unlike


usual


option


trading,


underwriting

an amount eq


business


does


uivalent


not


involve


value


a cash


put


transfer


option


on the


contract


date.


Rather,


proceeds


are


deferred


until


after


offer


date,


when


deal


between


issuer


and


underwrite

process,


Equation


are


settled.


discounting


2-12),


Because


is not


we need


this


required.


to adjust


underwriting


Therefore,


usual


as shown


payoff


function


a put


option


multiplying


value


option


reciprocal


scount


factor.


In addition,


since


underwriting


spread


determined


contract


date


setting


offering


price,


and


price


that


underwriter


promises


pay


issuing


firm,


no discounting


needed.


This


relationship


can


further


explained


following


manner:


The


current


put


price


discounted


expected


terminal


price;


E(-p


where


a negative


sign


means


a short


position.


From


this,


= erT


(-p).


Since


discounting


simply


equal


is not


needed,


to E(-p


underwriting


(X-B).


Substituting


contract,


E(-p








equilibrium,


riskless


rate


instead


risk


adjusted


rate


employed


see


Cox


and


Ross


1976)).


The


implic


assumption


employing


riskless


rate


a discount


factor


to derive


value


an underwriting


contract


that


Black


Scholes


option


valuation


model


1973)


valid


and


that


value


put


option


can


derived


with


certainty,


using


five


input


variables.


Employing


put-call


parity


relationship,


Equation


can


rewritten


- e


S,T;X)


-13)


The


10),


above


which


equation


we derived


exactly


using


same


a differential


as Equation


equation


previous


part


Comparison


Our


Model


with


Smith


s Model


Smith


1977)


derives


a similar


valuation


model


underwriting


contract


employing


option


pricing


framework.


Assuming


that


an underwriter


submits


a bid,


Btotal


, today


and


promises


pay


Btotal


dollars


offer


date


undertaking


the


new


shares


which


represent


fraction


a of


total


shares


ssuing


firm,


boundary


condition


underwriter


s position


offer


date


shown


to be


U*total


= Min


[a(V *+


Btotal)


_ Btotal


- Btotal]


(2-14)









= no. of


Btotal


new


= total


shares

amount


no.


that


total


shares


an underwriter


promises


pay


= total


offering


issuing

amount,


firm

and


denotes


value


on the


offer


date.


above


condition


indicates


that


total


offering


amount,


above


value


shares


firm


on the


offer


date,


a(V*+Btotal),


underwriter


cannot


sell


new


shares


offer


price,


hence


receives


portion


total


firm,


total


amount


value


that


Btotal


firm,


guarantees


other


hand,


a(V*+Btotal),


pay


K is


minus


issuing


below


value


shares,


underwriter


able


to sell


new


shares


and


receive


a total


amount


of K,


which


equal


offer


shares,


price


minus


per


share


Btotal


It i


multiplied


implicitly


number


assumed


new


that


offering


is successful,


a(V +Btotal)


equal


to K and


that


not,


Employing


former


Equation


is strictly


2-3)


less


with


than


above


latter.


boundary


condition,


following


Smith


closed


derives


form


risk-neutral


solution


case


value


underwriting


contract.


Total


- erT


a V


- a)Btotal


- erT


aV N(hl)


*r-nt- 1 S-


CL


It)









= time


other


to until


variables


offer


are


date,


as defined


and

previously.


Equation


analogous


reveals


to a portfolio


that


an underwriting


consisting


a long


contract


position


assets


a call


firm,


option


a cash


on a of


payment,


firm


and


value


a short


with


position


an exercise


price


equal


- a Btotal)


Total


- erTaV


- erTc(V,T


K-aBtotal


-16)


now


compare


our


model


with


Smith


s on both


theoretical


practical


grounds.


argue


that


Smith


s model


difficult


to apply


operationally


to a negotiated


underwriting


sale.


The


equations


are


rewritten


below.


= erTs


- erT


S,T;


(2-15)


total=


At a first


a erTV+Btotal]


glance,


_Btotal_


major


erTc(V,T;K-aBtotal)


difference


2-16)


between


formulas


seems


to stem


from


fact


that


while


value


underwriting


contract


Equation


-15)


express


sed


terms


share,


terms


total


analyzed


value


Equation


shares


can


eas


ily


shown


that


number


new


shares


Equation


-16)


small


relative


that


outs


standing


shares,


two


equations


will


give


close


values


underwriting


contract.


see


this,


we first


rewrite


(l-a)Btotal









l/Qn){[Qn/(Qn+Qs)][


SosQ


+ B Qn]


- B Qn


- erT[Qn/(Qn+Qs


SoQnN(hI


+rx


Qn/(Qn+Qs))]N(h2))


Qn+Qs


- {[Q


s/(Qn+Qs)]e


SoN(hl)+[Qn/(Qn+Qs


]BN(h


2)}+XN(h2)


2-17)


where


= number


new


shares


being


issued


= number


outstanding


shares


= stock


price


per


share


outstanding


shares


and


subscript


denotes


"Smith"


small


relative


to Q


Qs/(Qn+Qs)


close


one


and


zero.


exactly


same


Then


as Equation


above


2-15)


equation


except


becomes


that


replaced


From


this


relationship,


obvious


that


larger


relative


to Qs,


bigger


discrepancy


between


two


model


values.


Another

distribution


similarity


lies


underlying


assumption


asset.


While


on the


our


return


model


assumes


a lognormal


distribution


stock


price


per


share,


Smith


s model


assumes


same


distribution


on total


value


outstanding


shares


, V.


Since


firm


Smith


model


an all


equity


firm,


V is


fact


nothing


but


stock


price


per


share


multiplied


number


outstanding


shares,


where


stock


price


also


lognormally


distributed.


16 c


q *


Ia


r- -a -a a -- .- -- ---- -"


[Qn/


On+Qs


Qn+Qs


1


ft


R


P


L~rL* ~L1 rLr


1


-- t


--









In addition,


models


are


compatible


with


each


other.


we close


examine


equations,


we can


see


that


stock


price


Equation


(2-15


actually


certain


weighted


outstanding

information


share,


regarding


the


average


value


stock


new


the

price


offering


stock price

before the


is revealed


market,


price


per


share


to be


paid


issuing


firm,


As shown


Equation


17),


weighting


ctors


are


proportion


new


shares


and


outstanding


shares


relative


total


number


shar


es.


This


because,


an efficient


offering


capital


(namely,


market,


size


information


ering)


on a new


should


issue


already


reflected


stock


price


before


contract


date.


typical


negotiated


sale,


offering


information


available


market


least


several


days


before


contract


date


even


though


exact


offer


price


and


number


new


shares


being


issued


announced


until


offer


date.


other


hand,


this


process


new


issue


offering


makes


difficult


to employ


Smith


s model


widely


used


negotiated


sale.


Since


Smith


s model


requires


stock


price


outstanding


shares,


which


does


reflect


offering


information,


practically









impossible


to observe


this


price


underwriting


contract


date.


Suppose


that


an underwriter


uses


Smith


s model


evaluate


underwriting


contract


on the


contract


date,


one


day


before


offer


date


His


estimation


will


be based


closing


stock


price


previous


day


or the


prevail-


stock


price


market


time


contract.


However,


since


either


price


reflects


information


on the


new


issue


offering,


namely,


expected


proceeds


from


issuing


new


shares,


using


prevailing


market


price


would


result


a double


counting


effect


new


offering.


Smith


s model


implicitly


assumes


that


information


regarding


new


issue


is not


revealed


until


offer


date


thus


V in


his


formula


Equation


16)),


Equation


-17),


does


include


any


information


regarding


new


Smith


issue.


s model


Therefore,


using


order


we should


to properly


back


employ


pre-


announcement


date


and


get


stock


price


that


date.


This


problem


strictly


limits


applicability


of Smith


model


valuation


an underwriting


contract


negotiated


sale


time


contracting.


Unlike


Smith


s model,


our


model


free


from


this


information


problem


since


observed


stock


price


efficient


or So









market


Numerical


Analysis


Value


Underwriting


Contract


In the


illustrations


that


follow,


offer


price,


price


to be


paid


issuing


firm,


are


assumed


$100


and


$94,


respectively.


Thus,


underwriting


spread


case


a successful


offering


or 6%


offer


examples


values


price.


are


are


The


arbitrary,


likely


parameter


but


span


values


selected


relevant


chosen


these


ranges


range


these


each


parameter


value


reality.


Table


focuses


on the


effects


of changes


stock


price


on the


contract


date


contract


and


date,


closing


, changes


date,


time


, changes


between


standard


deviation


stock


returns,


and


changes


risk-


free


rate


interest,


on the


value


an underwriting


contract


based


on our


model


(Equation


10)).


The


results


consistent


a sensitivity


with


partial


analysis


effects


Table


derived


are


previous


subsection:


stock


price


increases


so does


value


contract


sk-free


rate


interest


increases,


does


value


contract


standard


deviation


stock


returns


increases,


value


contract


decreases.









Unlike


these


obvious


relationships,


value


under-


writing


contract


does


move


in one


direction


time


closing


date


increases.


appears


from


Table


that


value


underwriting


contract


sensitive


to changes


standard


deviation


stock


returns.


However,


the


stock


price


less


than


offer


price,


effect


changes


standard


deviation


stock


returns


not


noticeable.


Table


focuses


effect


changes


number


new


shares


being


issued


relative


number


outstanding


share


value


underwriting


contract


based


on the


model


derived


Smith


1977).


Note


that


employ


same


stock


prices


as we did


Table


order


to make


model


value


an underwriting


contract


easily


comparable


to each


other.


The


responses


value


underwriting


contract


based


on Smith


s model


changes


values


other


four


arguments,


namely,


' a,


and


whose


results


brevity


sake


are


not


reported


here


are


expected


direction


and


qualitatively


similar


ones


shown


Table


As demonstrated


before,


value


an underwriting


contract


decreases


relative


proportion


new


shares


to outs


standing


shares,


or to


total


shares,


increases.


exceptions


occur


when


the


stock


price


ess


than


he price









Table


Value of


An Underwriting Contract Based on


Our Model:
Price At th


Sensitivity Analysis with respect


le Contract Date,


Standard Deviation of


Risk-free Rate of


Stock Returns,


and Time


to Stock
Interest,


to Offer


Date


Model


= erT


- erT


c(S,T;X)


Risk-free Rate of


Interest


Stock Price


Contract


Standard Deviation of


Stock Returns


Date


Time


to Offer


Date


day


6.000
5.798


.013


0.013
-2.988


.999
.589
.103
.013
.988


5.17


2.945
0.012
-2.988


.000
.805


.975


.999
.596
.026
.026
.975


.926
.178


2.958


-2.975


. Time


to Offer


Date


days


.000
.718
.027


0.026
-2.975


.988
.423


5.753


3.015


0.02


.975


0.006
-2.976


.000
.731
.053
.052
.950


.988
.436
.041


0.052


.758
.846


2.815
0.031


.950


.951


Time


to Offer


Date


= 4 days


.999
.609


3.053
0.052
-2.950


5.928


.192


2.982


.051
.950


.412
.356
.478


-0.077
-2.968


.000
.635
.106
.103
.900


.218
.031
.102
.900


.425
.383
.518


-0.030
-2.918


Notes:


Offer price


= $100.


Price


to be


paid


to issuer


= $94.









Table


The


Value of


An Underwriting Contract Based on


Smith's Model:


Sensitivity with respect


to Stock


Price At
the Number


Model:


the Contract Date,


of New Shares


{[Qs/(Qn+Qs)]


Time


Relative


erT


to Offer


Date,


and


to Outstanding Shares


[Qn/(Qn+Qs)3


- {[Qs/


Qn+Qs)]erTSoN(hl)+[Qn/(Qn+Qs)]BN(h2)}+XN(h2)


Number


of New Shares


Stock Price


100,000


00,000


400,000


1,000,000


Contract


Date


.40)


.00)


Time


to Offer


Date


day


.000
.435


.012


.000
.010
.511
.011
.490


.976


.009


-2.134


.507
.007
.507
.006
.494


. Time


to Offer


Date


= 2 days


6.000
5.394
2.751


.998
.008


.938
.305
.162


.514
.014


1.513


0.018


-2.70


.479


.125


.013
.488


Time


to Offer


Date


= 4 days


5.997


5.985


.873


.528


0.047


.984
.544
.043
.458


3.027


.181


0.037
-2.107


.475


Notes:


Outstanding shares


per


share


of interest


to be p
t = 5%.


aid


= 1,000,000


to issuing firm


. Offer price


= $94.


Standard deviation of


= $100.


Price


Risk-free rate


stock return


= 10%.


aa --- - - -1


.10)a


a ~,,,.,,


u A A









Table


Relative Difference


Based on Our
(Equations


.e Value of An Underwriting Contract
Model and Smith's Model
s (2-9) and (2-16))a


Number


of New Shares


Stock Price


100,000

(0.10)b


200,000


400,000


1,000,000


Contract


Date


.40)


.00)


Time


Time


Time


to Offer


to Offer


to Offer


Date


Date


Date


day


days


= 4 days


.022
.298
.400
.400
.400


Notes:


Outstanding shares


share t
interest


o be pai
= 5%; s


= 1,000,000; offer
d to issuing firm =
standard deviation of


price
$94;
stoc


= $100; price
risk-free rate
k return = 10%


a Relative difference =
16)) / (Us in Equation


(U in Equation
(2-16))


2-9)


- Us


in Equation


b denotes the


ratio of


new shares to outstanding


shares.









than


what


gets


from


selling


shares,


incurs


a loss.


A close


loss


examination


decreases


Table


firm


shows


issues


that


- and


magnitude


underwriter


takes


- a higher


proportion


new


shares,


which


appears


contrary


our


intuition.


now


compare


values


an underwriting


contract


based


on our


model


Table


with


those


based


on Smith


model


(Table


The


numbers


Table


represent


relative


difference


between


values


underwriting


contract


Table


2-1 and


those


Table


four


different


values;


proportions


= 5%,


new


= 10%.


shares


Table


under


shows


same


that


parameter


larger


number


new


shares


relative


to outstanding


shares,


bigger


discrepancy


between


two


model


values


and


that


when


relative


proportion


new


shares


smallest,


two


models


give


closer


values


underwriting


contract.


In each


column


differences


Panels


increase


and


stock


these


price


relative


decreases.


Note


that


regard


ess


time


to offer


date,


value


contract


our


model


almost


twice


large


that


Smith


s model


when


a firm


issues


same


number


new


shares


outs


standing


shares.


Data









December


1985.


This


sample


represents


issues


reported


Directory


Corporate


Financing


published


Investment


Dealer


s Digest,


which


provides


data


on issuer


name,


issue


date,


public


offering


price,


issue


size,


underwriting


spread


and


three


components


selling


concession,


underwriting


fee,


and


managing


fee),


type


offering


(negotiated


or competitive),


type


contract


firm


commitment


or best


effort).


The


sample


is restricted


negotiated


offerings


common


stock.


Thus,


issues


comprised


more


than


one


class


common


stock,


or combined


units


common


stock


and


other


securities,


are


excluded


from


sample


along


with


competitive-bid


offerings.


The


sample


further


restricted


to offerings


those


firms


whose


common


stock


traded


New


York


Stock


Exchange


or the


American


Stock


Exchange


which


at least


daily


stock


returns


preceding


issue


date


are


available


from


Center


Research


Security


Prices


CRSP)


daily


return


tape.


The


data


each


issue


cross-checked


using


data


from


Securities


Exchange


Statistics


Commi


Tape.


ssion


For


s 1985


each


Registered


selected


Offering


issues,


share


price


previous


day


taken


was


before


closing


offer


date


market


price


from


CRSP


on the


daily


master


data


file


and/or


NYSE


and


AMEX


Daily


Stock


Price









The


final


sample


which


necessary


data


were


available


consists


seasoned


new


equity


issues.


Description


of Data


Tables


2-4 and


present


descriptive


statistics


sample.


several


Table


variable


summarizes


interest


mean


three


median


industry


values


subgroups


These


groupings


seem


to be


useful


since


effects


various


firm-related


characteristics


on the


underwriting


compensation


could


quite


different


regulated


than


unregulated


industries.


As Table


-4 shows,


average


gross


spread


as a


percentage


public


offering


price


.37%


issues

(62%).


mately


with

The


selling


concession


underwriting


equal


each).


and


The


taking


managing


gross


the

fees


largest


are


spread


portion


approxl-


shown


to be,


on average,


much


lower


utility


companies


than


industrial


financial


companies


(3.05%


vs.


and


5.01%).


This


result


may


least


part,


due


significantly


lower


underlying


risk


stock


proxied by

returns of


the

the


standard


new


deviation


utility


It is


new


frequently


equity


by inform
investors


may


also


noted


issues,


national


and


that


new


differences


investment


capture


information


asymmetry


effect
among


with
issue


between


bankers)


regard
market


insiders


Thus, t
different


three


_ .. _ _


in


-


valuation


characterized


and


hese


magn


outside
groupings
itude of


austry groups o
j_ -c C a-


n the


'I


1









offerings.


Table


also


shows


that


seasoned


new


equity


issues


are,


on average,


underpriced


.3%.


Out


issues


our


sample,


a higher


offering


price


than


last


trade


price


was


only


issues


20%)


and


vice-versa


issues


price


20%),


same


while


issues


level


last


52%)


trade


offering


price.


interesting


price


finding


utility


that,


companies


on average,


is set


public


significantly


offering


higher


than


last


trade


price,


namely


overpriced,


.48%.


Considering


magnitude


underpricing


together


with


risk


underlying


new


issue,


is evident


from


Table


that


underwriters


offering


price


lower


than


last


trade


price


more


risky


and


volatile


issues.


This


result


on the


underpricing


of seasoned


new


equity


issues


consistent


with


previous


findings.


Smith


(1977)


finds


that


these


issues


during


period


1971


-1975


underpriced


and


Bhagat


and


Fro


(1986)


find


that


seasoned


new


equity


issues


NYSE


AMEX


utility


companies


using


negotiated


offerings


from


January


1973


through


September


1980


overpriced


.3%.


Our


findings


also


lend


support


hypothesis


advanced


Parsons


and


Raviv


(1985).


They


predict


that


due


asymmetry


information


among


inve


stores


offering


price









can


seen


from


Table


that


offering


price


in our


sample


was


below


both


market


price


prevailing


prior


offer


date


and


price


following


distribution


new


issue


MPP


.23%),


even


though


magnitude


underpricing


was


slightly


reduced


time


market


closing


offer


date


Table


values


-5 provides


relevant


results


variables


comparing


based


average


type


registration


shelf


. nonshelf),


method


syndication


(syndicated


. nonsyndicated),


and


relative


position


managing


underwriter.


can


spread


nonshelf


This


seen


on shelf


issues


finding


from


issues


more


appears


Table


average


significantly


than


to be


lower


underwriting


than


offering


attributable


price


increased


competition


among


underwriters


shelf


offerings.


Even


though


underpri


cing


greater


with


nonshelf


issues


than


shelf

diffe


issues,


rence


there


does


in underpri


seem


cing


to be

two


any

types


significant


offerings.


Overall,


our


consistent


findings


with


those


on shelf


vs.


Bhagat,


nonshelf


Marr,


and


issues

Thompson


are


1985).


Smith


(1977)


obtained


similar


results.


LL Shelf r


egi


station


-A- 4


under
-' 1


Securities


r*


*


and


Exchange


- a


m


__


CI








A close


underpricing


examination


was


MPP


significantly


Table


reduced


reveals


after


that


shelf


issues


were


offered,


implying


that


the


post-offer


stock


price


recovered


support


substantially


investment


banking


This


industry


evidence


s view


does


that


not


shelf


registration


depr


esses


price


firm'


outstanding


shares


Our


findings


on underpric


shelf


nonshelf


issues


indicate


that


shelf


registration


indeed


allows


ssuing


firms


take


advantage


favorable


market


conditions


minimi


ing


time


interval


between


decision


to make


offering


and


actual


sale.


The


syndication


method


new


issue


appears


to affect


magnitude


underwriter


s compensation.


Bhagat,


Marr,


and


Thompson


(1985)


explain


this


as follows


cost


savings


that


result


from


not


forming


a syndicate


form


writer


spread


offering.
particular
likely to


are


a lower


will


Also,
issue
form a


passed
spread,


lower


an underwriter


will


very


syndicate


risky,
(P.


then


the
the


issuer
under-


nonsyndicated


knows


that


more


1391)


It follows


are,


from


general,


their


likely


argument

to have


that


a higher


syndicated


offerings


underwriting


spread


than


non-syndicated


Looking


syndication


mean


in Table


offerings


value


2-5,


because


based


average


their


on the


higher


method


underwriting


spread


risk


erings


are,


as expected,


lower








nonsyndicated


offerings


than


syndicated


offerings,


even


though


differences


are


not


significantly


different


from


zero


level.


It is


interesting


to note


that


while


post-issue


share


price


syndicated


close


offering


offerings


price,


rose


substantially


after-issue


share


to be


price


non-syndicated


offering


fell


down


further


below


offering


price.


this


reason,


one


may


conjecture


that


underwriters,


including


syndicate


members


as a


whole,


provide


more


effective


services


marketing


and


stabilization


activities


in syndicated


offerings.


investment


banking


firms


are


competitive,


at leas


within


a segmented


market


that


they


belong


would


useful


to examine


underwriter


s compensation


partitioning


sample


issues


into


several


subgroups


based


upon


underwriter'


relative


position


the


industry.


is reasonable


to expect


that


while


underwriting


firms


different


groups


certainly


compete


with


one


another,


they


seem


to compete


more


frequently


with


other


firms


same


group.


Thus,


we should


able


to analyze


whether


and


how


much


a specific


banking


firm


or a group


several


firms


compensated


underwriting


risk


differently


than


other


members


same


group


or than


other


groups


investment


banking


firms.









this


distinct


purpose,


subgroups


sample


- "special"


divided


"bulge"),


into


three


"maj or"


and


"submaj or" 2 5


following


usual


breakdown


securities


industry


large


investment


banking


firms


see


Perez


(1984,


136))


This


ranking


underwriters,


sometimes


underwriter


called


"pecking


s overall


syndi


order,


cate


" shows


standing


each


as a symbol


real


power


industry.


Since


each


firm


s standing


governs


size


of its


underwriting


participation,


especially


in a negotiated


sale,


bracket


position


important


impact


on the


underwriter


s profits.


The


second


column


Table


shows


that


gross


spread


among


three


subgroups


indeed


significantly


different


from


each


other


with


"submaj or"


group


having


earned


significant


highest


spread


difference


clear


in underwriting


spread


whether


due


difference


stock


volatility


new


issues


underwritten


each


group


offering


size


or both.


It is


commonly


believed


that


there


are


economies


scale


associated


with


issuing


new


stock.


Previous


studies


found


'3 Hayes (1971)
characteristics


provides
of each


an excellent
underwriting


discussion


bracket.


about


Goldm
Broth'
HISjr -S a


largest
an Sachs
ers are


-II


investment
, Merrill


the "


Sp


banking


Lynch,
ecial"


- r


firms


Morgan S
bracket,


_- E_ I


such


as First


;tanley, a
15 firms
i j- *


nd
in


Boston,


Salomon
the


t


L,,,1,, I


E!






45


Table


Sampl
During


Statistics


January


1982


New


through


Equity


Issues


December


Industrial


Variable


Companies
es (n=426)b


Financial
Companies
(n=96)


Utility
Companies
(n=157)


Total
(n=679)


Mean


Median


Mean


Median


Mean


Median


Mean


Median


Gross


Spread


.14)


Selling


Conc


ess


Underwriting
Fee (
Managing


0.99
0.48)
0.98


1.01)
2.14


.66)


.80)


1.06


.47)


.70)


.22)


.17)


0.49)
0.89


1.13)
0.92


.41)


.45)


Reallowances


.58)


.39)


.37)


.66)


0.69


.19)


.44)


0.67


.67)


.68)


.34)


.48)

.65


.53)


Underpricing


-0.49


.91)


.87)


.81)


.44)


MPP


0.00


.14)


.07)


.59)


.94)


.80)


.93)


.24)


Relative


13.13


.10)


.37)(15


8.04


.03)


.37)


9)(10


.89)


Offer


$million)


4)(44


.07)(68


69.84)(43


7)(5


.62)


.47)


.68)


.60)(16


.08)


.11)


.13)(14


Notes
fee,


: Gross


and


offering
price)-1
price at
is expre
deviatio:


spread,


reallowances


price.


.0.
the
ssed
n of


selling


are


concess


express


Underpricing=(last


is scaled


offering dat
as a percent


stock


e/public


ion,


underwriting


as a percentage


trade


a factor
offering


of outstanding


returns.


price


/public


fee, managing
the public
offering


MPP=(closing


price


-1.0


shares


stock


Relati


standard


1985a


.88)d(il






46


Table


Compari


son


Mean


Values


Variables


on Several


Categories


Type


Regi


Method


station


Syndication


Variables


Shelf
n=49)b


Nonshelf
(n=630)


Syndicated
(n=130)


Nonsyndicated
(n=67)


Gross


Spread


.91*
.00*


Reallowances
Underpricing
MPP
STD
Share Size


.41*
.90*


Offer


29.64


11.79


53.30


2.61*
-1.11
-3.44*
0.43
-1.77
3.11*


Relative


ition


of Managing


Underwriters


Variabi


cial


Major
(n=307)


Submajor
(n=43)


F-value


Gross


Spread


Reallowances
Underpricing
MPP


0.55


1.32
1.81


57.09**
77.47**
19.82


STD


Share
Offer


24.07
38.35**
28.77**


10.12


Notes
tage


: Gross


price/public


of 100.
offering


Share
Offer


size


MPP=(
price


spread
public


and


reallowances


offering


ering


do
)-1


sing
.0.


is expressed


price
stock
STD =


price


are


expre


ssed


Underpricing=(l


This


price at
standard


as a percentage


millions.


scaled


the offering
deviation of


u


as a percen-
ast trade
p by a factor


date/public


stock


outstanding


returns
shares.


n denote
denotes
denotes


figure
s the


are


number


t-value.


signifi


cant


significance


percentage


terms


observations


difference


except for
the sample


means


level.


Offer


groups


n 4 a --.





U1II1


An~nCA~


3:


Lt 1_ _


--AA--


L









evidence


significant


economies


scale


equity


offerings


Smith


(1977))


bond


rings


Ederington


1975)).


Recently,


Booth


and


Smith


(1986),


Bhagat


and


Frost


1986),


Bhagat,


Marr,


and


Thompson


1985)


also


documented


that


issue


size


significantly


negatively


related


to issuing


cost


(per


dollar


equity


sed)


and


thus


underwriting


spread.


Since


sk of


offering


increases


with


very


large


issues,


issuing


cost


and


also


gross


spread)


seems


to be


affected


combined


effects


both


underwriting


risk


and


offering


size.


Note


that


average


magnitude


underpricing


and


MPP


the "

other


special"


group


groups.


significantly


A plausible


smaller


explanation


than


this


evidence


that


"special"


group,


comprised


five


largest


investment


banking


houses


a far


clearer


picture


market


condition


and


thus


faces


more


certain


demand


new


share.


Empirical


Techniques


and


Results


For


each


seasoned


new


equity


issue,


model


value


underwriting


contract,


which


we denote


Umodel,


calculated.


Umodel


expressed


as a percentage


public


offering


price.


It is


important


to note


that


Umodel


represents


excess


returns


to underwriters


plus


under-









Umodel


= erT


SerT


2-10)


- erT


2-12)


where


G denotes


observed


underwriting


spread


X-B).


Therefore,


the


underwriting


activity


did


not


involve


any


expenses,


a competitive


equilibrium,


value


Model


should


be equal


zero.


Putting


this


another


way,


observed


underwriter


s compensation,


should


significantly


different


from


value


erT


which


nothing


but


fair


value


underwriting


risk


premium.


Any


deviation


between


these


two


values


represent


excess


returns


to underwriters.


Estimation


Relevant


Variables


calculation


Umodel


using


Equation


(2-10),


need


to estimate


input


variables,


four


which,


namely


stock


rate


price


and


exercise


underwriting


price


spread


riskles


are


interest


directly


observable.


The


other


variables


time


between


contract


date


and


final


settlement


date


and


standard


deviation


Since


stock


final


returns


underwriting


can


reasonably


contract


estimated.


is normally


signed


on the


day


before


actual


offer


date,


closing


stock


price


day


before


actual


offer


taken


input


variable


+ G









Since


difficult


to observe


actual


offering


period,


a range


from


a minimum


single


day


to maximum


five


days


taken


duration


contract.


Normally,


underwriter


obligated


pay


issuer


within


four


days


time


offering


offi


cially


begins


(Perez


(1984,


p.68)).


The


less


rate


interest


estimated


using


average


bid


and


asked


quotes


on the


U.S.


Treasury


Bills


reported


Wall


Street


Journal


approximately


same


duration


To estimate


underwriting


instantaneous


contract.


standard


deviation


relevant


stock


prior


returns,


offer


date


daily


are


stock


collected


return


from


data


CRSP


days


daily


stock


return


tape.


These


returns


are


used


first


compute


continuously


compounded


returns


taking


natural


of return


plus


The


computed


standard


deviation


daily


rates


are


then


converted


an annual


basis


stock price
slightly bi


the
the


a


value
true


e


input


underwriting


variable,


value.


seems


L" Option
standard


standard


result


values


are


deviation.


deviation


may


methodology.


estimates
a71e a..I Ii


weaken


seem


However,
to have


quite


estimates


case
with


fairly
k 1-


sensitive
extent t


due


the
dail


low


to estimates


hat


variation


to sampling


option-pricing


standard


stock


.


error,


returns,
errors,


a .I ,


this


which


1 I -


i


y


which


.I









The


last


input


variable,


measured


difference


between


offering


price


and


observed


underwriting


spread.


Employing


estimated


variables,


value


underwriting


contract,


Umodel,


estimated


two


different


ways


first,


under


ass


umption


no underwriting


expenses


involved


second,


incorporating


some


proxies


underwriting


expenses.


It is


important


to note


that


given


limited


data


underwriting


value


expenses


, these


underwriting


approaches


contract


correspond


to estimating


to extreme


cases.


In this


constitute


respect,


upper


predicted


lower


bound


underwriting


true


values


value


underwriting


contract.


In addition,


premium


we also


estimated


compare


with


underwriting


observed


risk


underwriting


fee,


which


paid


as a risk


premium


risk


assumed


underwriter.


Results--Upper


Bound


Underwriting


Value


first


task


to examine


direction


and


magnitude


predicted


value


underwriting


contract,


Umodel,


under


assumption


that


underwriting


activity


costless


assuming


that


out-of-pocket


expenses


are


reimbursed


issuing


firm)










: Umodel


vs.


:Umodel


where


Umodel


denotes


average


value


Umode


entire


sample.


reje


action


null


hypothesis


a statistical


test


and


a significantly


positive


average


value


Umodel


would


provide


evidence


that


issuing


firms,


on average,


overpay


offerings


sk that


that


underwriter


underwriters,


bears


on average,


in new


earn


issue


excess


returns.


In order


to determine


rejected,


whether


nonparametric


or not


test


null


techniques


hypothesis


Wilcoxon


Signed


Rank


test,


Mann-Whitney


test


are


used


to complement


conventional


t-test.


can


seen


from


Figure


, the


distribution


value


underwriting


contract


(the


excess


returns


underwriters),


estimated


Equation


-12),


positively


skewed


with


a mean


value


2.89%,


a median


value


.53%,


and


a couple


extreme


values


when


time


final


settlement


seems


to be


date


days


inappropriate


skewed


standard

nature o


stribution.


underwriting


risk


The


premium,


estimated


of

by


the
erT


model


more


positively


skewed


(see


Figure


-3).


can


such


t-test


Thus,


given


distribution












of -"o


Issues


o~o oo~ oo!?????????????oco o
S* 0 0 0 0 0 1 1 1 1 1 2 2 3 3 2 3 3 i
1 0 0 2 4 6 8 2 4 6 2 8 2 4 6 8
42 66 42


Value


As % of


Underwriting


Public


Contract


Offering


Distribution


When


Duration


Figure
Value


Contract


Underwriting


is Equal


to 5


Contract


Days












Issues


Si o o o o o o o o oi i i i i i i i i i i i


Model


As % of


Underwriting


Publi


Risk
ering


Premium


Price


Figure


Distribution


Model


Underwriting


Risk


Premium


When


Duration


Contract


is Equal


to 5


Days









nonparametric


standard


tests


parametric


have


tests.


a number


Unlike


advantages


parametric


over


tests,


nonparametric


tests


are


stribution


-free


since


they


assume


which


nothing


about


observed


stribution


sample


drawn.


population


In addition,


from


while


parametric


tests


typically


give


greatest


weight


to outliers,


which


may


bias


true


mean


values,


nonparametric


tests


weight


servations


equally.


Table


reports


stati


stical


results


predicted


value


underwriting


contract


using


both


t-test


and


Wilcoxon


Signed


Rank


test.


At first


glance,


there


seems


to be


substantial


excess


returns


to underwriters


over


period


examined


in our


study.


Regard


ess


time


final


settlement


date


and


states


tical


test


employed,


underwriting


values


are


significant,


ranging


from


.89%


to 3


.64%


expressed


as a per


centage


public

there

(97.8%


offering


were

) and


price.


positive

negative


In fact,


underwriting

underwriting


values

values


whol

for

for


issues


664 i

only


issues


issues


As expected,


value


underwriting


contract


decreases


time


final


settlement


date


lengthens.


This


because


an increase


in the


duration


offering


period


increases


uncertainty


about th









The


apparent,


significantly


large


mean


underwriting


values


are


at all


surprising


and


are


mainly


ignorance


involved


underwriting


expenses.


In this


regard,


mean


values


excess


returns


Table


can


regarded


upper


bound


true


excess


returns


earned


underwriters.


Results--Lower


Bound


Underwriting


Value


With


this


mind,


we now


turn


second


approach


estimating


value


underwriting


contract.


practice,


certain


underwriting


expenses


are


deducted


from


the g

heavy


ross


spread


losses


when


and

his


on occasion


compensation


underwriter


could


not


may


suffer


cover


expenses


incurred.


However,


overall


profit


net


underwriting


expenses


underwriting


activity


would


difficult


to ascertain,


because


nobody


can


agree


on how


much


underwriting


firm'


overhead


should


allocated


activity.


In addition,


data


on underwriting


expenses


incurred


investment


banking


firms


each


offer


are


public


cly


available.


These


problems


restrict


direct


measurement


relevant


underwriting


expenses.


The


total


costs


underwriters


their


issue


ocess


mainly


involve


expenses


providing


originating,


managing,


and


distributing


services.


The


originating


expenses


may


depend


upon


number and


number


and









cost


keeping


records


involved


agreements,


prospectus


a syndicate,


managing


preparation


and


costs,


so on are


services


are


cost


related


likely


involved


managing


originating


to be


and


invariant


additional


dollar


equity


underwritten.


The


distributing


expenses


are


largely


affected


time


required


to sell


out


issue,


whi


ch is,


turn,


closely


overhead


associated


or fixed


with


size


stribution


issue.


costs


such


Other


advertising


costs


are


also


involved.


Based


on the


above


discussion,


we assume


that


the


components


observed


underwriting


spread,


namely,


managing


and


selling


concession,


are


just


enough


cover


underwriting


expenses


to handle


offering.


Recall


that


managing


paid


lead


manager


and


co-


managers,


any,


services


originating


and


managing


offering


underwriters,


that


including


selling


managing


concession


underwriter,


paid


who


those


distribute


issue


paid


that


directly.


Unlike


as an insurance


underwriter


these


premium


assumes


underwriting


risks


guaranteeing


offering


with


commitment


own


capital.


It is


unlikely


that


this


activity


involves


substantial


expenses.


Another


factor


that


we take


into


account


"dealer









underwriter,


they


sell


those


shares


public


offering


price


less


"reallowances"


which


a fraction


selling co

concession


ncesslion.


this


Thus,


case


actual


reduces


underwriter


s selling


amount


reallowances.


Considering


above


factors


value


underwriting


contract


re-estimated


using


following


equation.


AUmodel


= -erT


- RA)


- MF]


(2-18


where


AUmode 1


= adjusted


value


underwriting


contract


SC = selling


concession


RA = reallowances


MF = managing


and


other


It should


variables


be noted


that


are


our


defined


as before.


conjecture


under-


writing


expenses


appears


to be


quite


extreme


Since


this


could


lead


an overestimation


actual


underwriting


expenses


and


thus


an underestimation


true


under-


writing


values,


estimated


values


using


Equation


(2-18)


should


regarded


lower


bound.


Compared

predicted u


figures


underwriting


reported


value


Table


Table


are


2-6,


quite


small,


although


they


are


still


significantly


different


from


zero


any


conventional


level


significance.


The mean


excess


The


mean









returns


range


from


.15%


5-day


duration


to 0


.91%


single


day


duration,


indicating


that


underwriters


are


significantly


overpaid


their


assumed


underwriting


risk


regardless


However,


time


a close


final


examination


settlement


number


date.


issues


which


underwriters


earned


positive


or negative


returns


tells


a somewhat


different


story.


While


there


were


negative


excess


returns


issues


total


issues


when


day,


approximately


half


total


issues


issues)


generated


negative


excess


returns


under-


writers


when


days.


This


appears


to imply


that


on occasion


underwriter


realized


compensation


much


lower


than


fixed


under-


writing


spread,


possibly


due


fact


that


new


issue


simply


does


not


appeal


market.


However,


this


conclusion


does


seem


valid,


primarily


because


the


market


quite


predictable


seasoned


new


stock


issues.


Another


plausible


reason


evidence


Table


that


many


underwriters,


selling


concession


constitutes


their


major

Result


source


- Excess


excess


underwriting


Underwriting


Risk


income.

Premium


Finally,


we compare


predicted


underwriting


risk


premium


estimated


using


formula


erT


with


observed


underwriting


fee,


which


paid


as a ri


premium









fair


price


underwriting


risk


premium


new


equity


issue


offering.


It is


well


known


that


insurance


aspect


investment


banking


use


term


"underwriting"


describe


risk-taking


behavior


investment


banker


and


that


fees


paid


to compensate


this


risk


are


analogous


put


option


premium.


Note


that


usual


insurance


premium


represented


value


a simple


put


option,


Thus,


underwriting


premium


in a new


issue


offering


estimated


using


then


this


would


underestimate


true


value


underwriting


risk


premium


since


multiplying


factor,


erT


ignored.


The


degree


underestimation


mainly


depends


upon


duration


underwriting


contract,


that


time


interval


between


contract


date


and


final


settlement


date


new


issue


offering.


Based


on the


above


discus


sion,


excess


underwriting


sk premium


(EURP),


difference


between


observed


underwriting


and


predicted


put


option


premium


(multiplied


erT),


calculated


from


following


equation.


EURP


SerT


S,T;


+ UF


2-19)


where


EURP


= excess


underwriting


risk


premium









UF = underwriting


narrow


term)


and


other


variables


are


defined


as before.


It is


worth


mentioning


that


Equation


(2-19)


a special


case


Equation


-18),


where


reallowances


(RA)


are


zero.


This


case


where


underwriters


a new


issue


process


are


able


to sell


shares


that


they


are


undertaking


and


thus


there


is no buy


or sell


between


underwriters


and/or


other


dealers.


The


results


are


shown


Table


2-8.


In contrast


to the


previous


findings,


we observe


negative


average


values


excess


risk


premium,


which


are


all


significant,


except


case


where


day.


In fact,


there


were


negative


average


values


excess


risk premium


issues


when


= 1 day


when


days.


An interesting


finding


from


Table


that


even


though


discrepancies


actual


underwriting


from


predicted


risk


premium


are


significantly


different


from


zero,


there


seems


to be


an average


excess


risk


premium


close


zero


case


where


duration


underwriting


contract


lies


between


one


and


days.


Supposing


that


option


premium


is a reasonable


measure


fair


underwriting


underwriters


risk premium,


are


this


practice


evidence


able


may


to sell


indicate


out


that


whole


issue in


a couple


days


and thus


they


are


compensated


and


thus


are


_eta









properly


cover


risk


that


they


assumed.


Rather,


negative


excess


returns


to underwriters


are


as expected.


important


recognize


that


underwriting


risk


Equation


2-19


expli


citly


present


put


option


variance


stock


returns.


Accordingly,


put


option


premium


represents


fair


insurance


premium,


covering


only


underwriting


in a narrow


sense.


However,


total


flotation


risk


in a new


equity


offering


more


likely


to involve


both


pri


cing


risk


underwriting


risk


in a narrow


sense)


and


stributing


inventory


Bhagat


Frost


1986)


postulate


that


since


investment


banking


firms


are


well


diversified,


they


are


not


able


diversify


away


unsystematic


sk associated


with


temporarily

underwriters


holding

receive


newly

some


issued


stock.


this


compensation


uns,


reason,

ystematic


risk


that


they


accept.


Booth


and


Smith


1986)


also


find


that


unsystematic


sk is


significantly


related


underwriter


In practice


s compensation,


, underwriting


at least


activity


industrial


intertwined


firms.


with


stribution


activity


and


risk


taking


only


incidental


distribution


function


see


Christensen


(1965))


other


hand,


distribution


risk


more


related


unsystematic


risk


associated


with


temporarily


holding


unsold








isolate


underwriting


risk


a narrow


sense)


and


proper


compensation


from


total


flotation


risk


and


gross


underwriting


spread.


Effects


Firm/Offering


Variables


on Underwriting


Value


we find


that


underwriters,


on average,


earned


significant


excess


returns


over


four-year


period,


interest


investigate


whether


and


how


magnitude


underwriting


values


varies


with


firm


and


issue


characteristics


such


stock


price


volatility,


offering


size


, industrial


sector,


type


registration


shelf


or non-


shelf


method


syndication


syndicated


or nonsyndicated),


relative


pos


ition


managing


underwriter


"special,


" "major,


"submaj or"


bracket).


To address


this


issue,


different


approaches


are


employed.


three


First,


subgroup


the

the


sample

basis


partitioned


into


categories


two


discussed


above.


Second,


effects


these


variables


and


factors


on the


underwriting


values


are


further


analyzed


using


following


cross-sectional


regression


equation.


+ al(STDi)


a3(Shelfi)


[log(Offer


+ a4(Industry-Ai


Size)i]


Industry-


+ a6(Lead


-Ai)


Lead-Bi)


(2-20)


where


Umodel(5)i


= the


value


underwriting


contract


Umodel(5)i


= a,









including


underwriting


expenses


(T=5


days)


STDi


= standard


Offer


Size)i


deviation


= natural


stock


logarithm


returns


offer


issue


size


issue


i in


millions


Shelfi


issue


i is


shelf


registered


and


otherwise


Industry


issue


made


an industrial


otherwi


Industry


issue


made


an financial


0 otherwise


Lead


lead


manager


issue


i iS


"special"


bracket


otherwise


Lead


lead


manager


issue


"major"


bracket


otherwise


= the


regression


error


term


The


result


are


summarized


Tables


and


-10.


Looking


first


figures


in Table


reveal


that


average


underwriting


value


as a percentage


offering


price


is significantly


lower


utility


firms,


shelf


registered


underwriters


issues,


syndicated


belonging


offerings,


"special"


group.


and


These


results


are


based


significance


level


of nonparametric


tests


because


skewed


nature


stribution


underwriting


value.


The


findings


Table


are









their


magnitude


are,


general,


as expected


and


consistent


with


results


volatility


stock


Table


returns


2-9,


although


and


different


effects


types


registration


results


underwriting


shown


value


Tables


are


and


not


significant.


-10,


results


on three


subgroups


underwriters


are


somewhat


surprising,


since


services


underwriters


"special"


group


seem


to be


more


valuable


issuing


firms


and


thus


their


services


are


inherently


more


costly


than


services


underwriters


other


groups.


One


possible


and


plausible)


reason


higher


excess


returns


"submajor"


group


that


they


are


mainly


attributable


to both


higher


risk


issues


that


this


group


undertaken


and


smaller


size


offering.


To investigate


this


issue


further,


we run


a regression


model


underwriting


value


against


standard


deviation


stock


returns


and


offer


each


bracket


group.


results


are


presented


Table


While


coeffi


clients


stock


volatility


"special"


and


"major"'


groups


are


positive


and


insignificant,


"submajor"


group,


has


a negative


value,


which


significant


level.


Furthermore,


stock


volatility


offer


size


"submaj or"


group


are


negatively


correlated


with


each


other.









certain


demand


new


shares


and


have


enough


capital


capacity

"submajor


to absorb the

" group, which


underlying


has


risk


relatively


new


small


issues

capital


the

base


limited


underwriting


skills,


significantly


affected


risk


new


ering


that


undertakes.


It is


also


conceivable


that


major


customers


underwriters


"special"


and


"major"


groups


are


larger


and


more


established


firms


and,


thus,


have


a good


deal


bargaining


power


new


issue


process.


This


would


make


offering


writing


value


more c

among


competitive,


bringing


underwriters


down

the


under-


"special"


and


"major"


groups.


The


above


results


together


with


findings


that


average

spread


excess


are


returns


significantly


as well


different


observed


among


three


underwriting

brackets


indirectly


imply


that


investment


banking


market


fact


segmented


according


investment


banker


s relative


position,


namely


"prestige,


" in


that


industry.


also


interest


see


whether


underwriters


are


competitive


within


a segmented


market


that


they


belong


To do


this,


we compare


mean


excess


returns


five


largest


firms


"special"


group.


It is


evident


from


Table


2-12


that


mean


returns


among


five


largest


underwriters


are


significantly


different


from


each










The


66


Table 2


Value


to Underwriters


Underwriting


Umodel(T)


assuming

= erT


Contract


(Excess


No Underwriting


Returns


Expenses


S,T


Duration
(day) Mean(%) T-value Pr>JTI Median(%) SRb Pr>JSRI


1 3.64 43.44 0.0001 3.41 114700 0.0001
(2.18)c (2.33)d
2 3.39 41.16 0.0001 3.12 114669 0.0001
(2.15) (2.16)
3 3.20 39.25 0.0001 2.90 114621 0.0001
(2.12) (2.12)
4 3.03 37.55 0.0001 2.70 114570 0.0001
(2.10) (2.08)
5 2.89 35.98 0.0001 2.53 114481 0.0001
(2.09) (2.02)


Notes


contract
where T


: Umodel(T)


expre
denote


days).


a predicted


ssed
s the


value


as a percentage


time


to final


underwriting


public


settlement


offering


date


price


,3,4,5


number


SR denote
Standard


of observation


Wil


coxon


deviation


Interquartile


range


signed


rank


t


parenthe


in pare


nthe


sample
est.
ses.


issues.


ses


Keys


: Umodel(T)


where


= ri


= erT
sk-free


time


= put


rate


to final


option


inter


settlement


date


= stock


= public
= observe


price


per


offering


share
price


underwriting


contract


per


date


share


spread


+ G


+ G









Table


Value


The


to Underwriters


AUmodel(T)


Contract


Underwriting


- e


(Excess


- RA)


Returns


Duration
(day) Mean(%) T-value Pr>IT! Median(%) SRb Pr>jSRJ


1 0.91 19.90 0.0001 0.79 91499.5 0.0001
(1.16)c (1.03)d
2 0.66 14.65 0.0001 0.53 78704.5 0.0001
(1.14) (1.01)
3 0.46 10.32 0.0001 0.36 58517.5 0.0001
(1.13) (1.00)
4 0.29 6.58 0.0001 0.21 36370.5 0.0001
(1.14) (1.03)
5 0.15 3.27 0.001 0.21 15521.5 0.001
(1.14) (1.04)


Notes
contri


: AUmodel(T)
act net of u


public


settlement


an adjusted


underwriting


offering


date


T=l,


price


2,3,4,5


model


expenses
where T


value
express


denote


underwriting
a percentage


time


days).


to final


number


SR denote
Standard


of
Wil


observations


coxon


signed


deviation


Interquartile


range


rank


parent


in pare


nt


e sample
test.
eses.
heses.


issues.


Keys: U
where


- e


= ri


p (S,T;
sk-free


- (SC


rate


of interest


- RA)


- MF]


= time
= put


to final


settlement


option


date


= stock
= public


= obs


SC = selling


price


per


concess


RA = reallowances


MF = managing


share
price


per


share


Underwriting


Net


Expenses


- MF]


offering


erved


underwriting


contract


spread


date









Table


Excess


Underwriting


EURP


- ert


sk Premium(EURP)a


+ UF


Duration
(day) Mean(%) T-value Pr>ITI Median(%) SRb Pr>ISRI


1 0.13 3.61 0.0004 0.16 32395.5 0.0001
(0.92)c (0.72)d
2 -0.12 -3.29 0.001 -0.10 -19859 0.0001
(0.91) (0.74)
3 -0.31 -8.67 0.0001 -0.30 -51725 0.0001
(0.92) (0.77)
4 -0.48 -13.09 0.0001 -0.47 -69388 0.0001
(0.93) (0.79)
5 -0.63 -16.81 0.0001 -0.61 -79953 0.0001
(0.95) (0.86)


Notes


: EURP


is an excess


as a percentage


time


to final


public


settlement


date


underwriting
c offering p


T=l,


sk premium


rice,


,3,4,5


where


expre


ssed


T denotes


days)


number


SR denote
Standard


observation


s Wilcoxon
deviation


signed


in par


the sample
ank test.
entheses.


issues


Interquartile


range


in parenthe


ses


Keys


: EURP


where


- e


= risk
= time
= put


p(S


-free


rate


to final


+ UF


interest


settlement


date


option


= stock
= public


UF = observed


price


per


offering


share
price


contract


per


date


share


underwriting









Table


Comparisc
Several


Mean


Categories


Values


time


Underwriting


to final


Contract


settlement


date=


Based


days)


Meana


Nonparametric


Sample


Significance


Level(%)b


Industry


Industrial
Financial


Utility


Type


of Regi


station


Shelf


Non


-shelf


Type


Distribution


Syndic


ated


Non-syndicated


3.55


Relative


Position


Managing


Underwriters


Special


Major


Group


Group


Submajor


Group


Means
These


are


tests
sample


compare s


per


centage


include
e means
on of m


the
and


ore


terms


Mann
the
than


-Whitne
Kruskal


U-test for


-Wallis


sample


sample


comparison
test for


means


Test


v w









Tabl


2-10


Regression


Analysis


Underwriting


Value


Explanatory
Variables


Dependent


Underwriting


Spread


Variable


Underwriting


Value


Coefficient


T-value


Coeffi


cient


T-value


Constant


.180


.67*


.183


14.14*


.027


.38*


.010


Offer


Size


.50*


.008


.009


.41*


Shelf


.006


Industry


0.007


.59*


.30*


.005


.010


Industry


.013


Lead


0.016


.73*

.76*


.008


Lead


.005


-0.008


Adjusted
F-value


Dep.
Root


.731


Mean


MSE


0.044
0.015


.317
.920


.017


Notes
The s
value


: For
ample
is u


expenses
returns


offer
shelf


made
i is


regi


by a
made


simplicity,
consists of


nder


and


subs
new


assumption


days.


issue


issue


stered


n indu


manager
Lead-Bi


otherwise


and
trial


STDi


script
equity


i for


ssue


issues


involved


standard


Log(Offer


in $
0 oth


and


an financial


ssue


lead
ui


i


The


is dropped.
underwriting


underwriting


devia


natural


millions
erwise.
0 otherwi


and
"spe


manager


regr


Shelfi


Indu
se.


0 otherwi
cial" bra
or issue


ess


tion of
logarith


SS


stry-Ai is 1
Industry-Bi


cket


Lea
and


i is in
error t


d-Ai is
0 othe
"maj or"


r


erm.


stock
m of
ue i is


issue


issue


lead


wise.
bracket


indicates


a significant


difference


level.


-- 1 1 .. ----


f









Table
Regression Analysis c
By The Relative Posi


of Underwriting Value
tion of Underwriters


Dependent


Underwriting Spread


Variables


Underwriting Value


Explanatory
Variables


Special


Major


Submajor


Special


Major


Submajor


Constant


Log(Offer
Size)


.34)*

.041


.007
.44)*


.142
.48)


.63)*

.046
.84)*

.009


.008


.010
.87)


.012
.97)*


.007
.39)*


.188
.42)*


.014
.64)


-0.009
-6.57)*(


.300
.81)


-0.026
-2.23)


.015
.55)*


Adjusted R2
F-value
Correlation


0.14


47.79


.194


.227


.134


.194


.227


-0.134


Sample


Notes:


The
valu


simplicity,


sample consists of


e


the subscript i
679 new equity


is under the assumption of


involved and T


returns of


= 5 days.


issue


STDi


Log(Offer


issue


issues.


The


is dropped.
underwriting


no underwriting expenses


standard deviation of


Size)


is natural


stock


logarithm of


offer


size of


issue


i in $ millions.


indicates a significant


difference at


level.









Table


Compari
Largest


son


Underwriting


Investment


Bankers


Value


Among


"Special"


Five


Group


Investment


Banking


Firm


Gross


Spread


Underwriting
Value(%)


Sample


Merrill


Salomo

Goldma

First

Morgan


Lynch


Brothers


Sachs


Boston


2.03


Stanley


1.99


F-value


.008


Kruskal


-Wallis


k-sample


test


value


.117


Notes


: The


underwriting


cons


underwriting


expenses
645 issues


value


involved


under


and


assumption


days.


Total


sample










compete


with


one


another,


they


compete


more


frequently


with


other


firms


same


bracket.


Comparison
on Rights


Empirical


Offers


Findings


In thi


subpart,


magnitude


overpricing


underwriting


services


found


our


study


is compared


with


previous


findings


on standby


rights


offerings.


Overall,


average


excess


returns


to underwriters


seem


to lie


between


.15%


underwriting


expenses


and


.89%


including


such


expenses


when


duration


underwriting


contract


5 days


and


between


.91%


and


.64%


when


duration


contract


day.


In a study


United


Kingdom


standby


rights


issues,


Marsh


1980


finds


that,


on average,


underwriters


earned


significant


excess


returns


.83%


as a percentage


offering


price,


providing


evidence


that


underwriting


was


cons


iderably


overpriced.


rights


issues


made


U.S.A.,


degree


overpricing


underwriting


premium


was


shown


to be


even


more


signifi


cant


with


excess


returns


underwriters


being


approximat


ely


.08%.


The


purchase


a standby


contract


a rights


issue


new


stock


same


effect


as a firm


commitment


cash


offer.


In a standby


agreement,


underwriting


simply


a put


option


giving


issuer


right


to put


unsubscribed


with


Results


Previous








Why


then


we observe


a significant


difference


under-


writing


value


between


standby


rights


offers


and


firm


commitment


worth


cash


examining


offers


a number


answer


important


this


question,


differences


between


these


types


offers.


First


all,


there


is a significant


difference


duration


underwriting


contracts.


In both


cases


underwriting


agreement


with


issuer


normally


signed


afternoon


before


offer


announcement.


However,


unlike


final


standby


on the


rights


last


offering,


acceptance


which


date


contract


rights


normally


three


weeks


after


initial


offer


date),


underwriting


contract


underwritten


cash


offers


lasts,


most,


four


to five


days.


shown


before,


underwriting


value


decreases


duration


of underwriting


contract


becomes


longer


and,


thus,


underwriters


earn


higher


excess


returns


on shorter-lived


contract.


Increase


in the


duration


contract


equivalent


an increase


in underwriting


This


would


increase


put


option


premium


and,


turn,


decrease


underwriter


s excess


returns.


Therefore,


difference


duration


contract


should,


at least


in part,


account


difference


underwriting


values


between


these


types


offerings.









. S .


In the


U.K.,


more


common


issuing


firms


offer


sec


uriti


directly


the


public,


with


investment


bankers


standing


to guarantee


a minimum


price


ssuer,


not


helping


distribute


issue


(Christensen


Therefore,


standby


U.K.


rights


offers


represents


a fee


underwriting


in a narrow


sense.


However,


underwriter


s compensation


U.S.


underwritten


distributing


cash


offers


services


includes


as well


fees


as for


advice


underwriting


services.


long


as the


underwriter


takes


distribution


risk


involved


a new


issue


offer


and


compensated


such


risk,


underwriting


value,


or excess


returns


to under-


writers


, would


much


higher


U.S.


firm


commitment


cash


offers


than


U.K.


standby


rights


offers.


Third,


a firm


commitment


contract


is similar


to the


limiting


firm


case


completely


a standby


fails


rights


to sell


offer


shares.


where


issuing


While


form


a standby


and


agreement


underwriter,


sk is


whole


shared


underwriting


both


risk


issuer


taken


solely


underwriter


firm


commitment


contract.


In this


respect,


it is


more


likely


that


underwriters


takes


higher


flotation


risk


latter


case


than


former


case.


Considering


differences


above,


underwriters









that


standby


rights


offers,


even


when


underwriting


risks


these


offers


are


supposedly


comparable.


Implications


Empirical


Results


Having


observed


significant


excess


returns


underwriters,


natural


question


- why


issuing


firms


overpay,


on average,


underwriting


activity?


answer


that,


this


typical


question,


negotiate


important


ed offering,


to recognize


underwriting


spread


may


also


include


payment


past,


present,


and


future


consulting


advice


given


investment


banker


unrelated


particular


offering.


The


true


value


underwriting


contract


should


equal


difference


between


gross


underwriting


spread


and


expenses


incurred


issuing


process


that


values


advice,


managing,


underwriting,


and


any


legal


services


involved


process.


The


difficulty


measuring


true


value


underwriting


primarily


stems


from


fact


that


is not


possible


from


our


data


to identify


value


advice


and


consulting


services


underwriters.


As a plausible


explanation


excess


returns


underwriters


found


study


U.K.


standby


rights


offers,


Marsh


1980)


issue


side


payment


underwriters


issuing


firms


subunderwriters.


However,


new


issue


market


competitive


underwriter









Another


argument,


as Marsh


(1980)


suggests,


that


because

exceed


the

the


costs


associated


overpayment


with


nonunderwritten


underwriting,


issuing


issues


firms


may


are


prepared


pay


high


underwriting


ees.


a non-


underwritten


issue


rearranging


fails,


issue


administrative


as well


cost


opportunity


cost


being


able


to utilize


otherwise


available


proceeds


can


quite


substantial.


However,


this


argument


is not


quite

equity


convincing


offers


since


to fail


the

seems


probability for

to be trivial.


seasoned

Further,


new

under-


writing


deals,


fact,


are


eas


arranged


and


costs


involved


this


process


not


appear


to be


substantial.


This


leads


inescapable


question


- Is


under-


writing


market


new


issue


rings


competitive?


believe


that


not.


underwriting


services


are


competitively


after


supplied,


underwriting


then


expenses,


the

just


underwriter

enough to


should


compensa


receive,

te him


risk


that


bears.


this


were


not


case,


new


banking


firms


would


enter


market


and


extract


any


excess


profit


market


paying


process


a higher


would


price


continue


issuing


until


firm.


profit


This


driven


down


zero


each


parti


cipant.


Traditionally,


investment


banking


long


tended


assume


a pyramidal


competitive


structure


with


a few


with









degree

Vinson


a monopsonistic


1970))


industry


a more


see


extensive


West


study,


1965,1966)


Hayes,


Spence


Marks


(1983)


show


that


examination


underwriting


volume


data


[and


gross


underwriting


revenue]


during


1970s


suggests


that


there


was


actually


tendency


toward


increased


concentration


investment


banking.


" (p.


They


further


show


that


"there


are


indeed


distinct


market


segments


within


which


stantial


competition


takes


place,


but


between


which


competition


may


much


ess


robust


sistent


this


study


with


strongly


these studies,

suggest that


the

the


overall


results


underwriting


market


less


than


perfectly


competitive


and


that


the


excess


returns


earned


underwriters


underwriting


business


are,


being


at least


part,


monopsonistic


due


in a new


equity


issue


would


market.


seem


This


to stem


underwriter


from


s monopolistic


nature


advantage


long-standing


relationship


between


issuing


firms


and


investment


bankers


The


Application


to Initial


-- - .


Public


Offerings


Although


this


study


does


empirically


examine


underwriting

(unseasoned


risk

new e


premium


quity


on initial


issues),


public


equity


we discuss


issues

subpart


t S S S


" (p.


Underwriting


Valuation


Model


J ^.


w


I 1 .









As shown


previously,


in order


to value


underwriting


risk


premium,


we need


measure


or estimate


input


variables;


time


between


contract


date


and


offer


date),


offering


paid


price),


to issuing


(risk-free


firm),


rate


standard


interest),


deviation


(price


stock


returns),


and


(stock


price


on the


contract


date)


For


initial


public


issues,


first


four


variabi


and


are


directly


observable.


a can


estimated


using


earnings


variability


or ex-post


stock


price


volatility


issuer


However,


as a proxy


is impossible


sk of


to observe


new


last


issue.


variable,


since


firms


going


public


do not


have


a pre-offering


trading


record.

estimate


Due

the


this


optimal


obstacle


underwriting


difficult


risk


to directly


premium.


Instead,


equilibrium


seems


stock


feasible


price,


to derive


, by


an explicit


standard


solution


numeri


analysis


techniques


such


iteration


method)36


given


the a

input


ctual


underwriting


variables


sk premium


compare


other


with


observed


five

market


price


first


day


trading


new


shares.


This


analysis


reflects


especially


this


enables


see


information


values


empirical


analysis


whether


regarding


X and


would


market


new


The


tha


correctly


equity


explicit

t the un


offering,


assumptions


derwriting


, X,









An extension


above


empirical


analysis


examine


relationship


among


stock


price,


offering


price,


underwriting


risk


premium


without


using


observed


underwriting


as a measure


input


variable,


Suppose


that


to issuer,


an underwriter


Since


return


proceeds


from


to be


underwriting


paid


new


issue


strictly


depends


upon


offering


price


and


stock


price


on the


offer


date,


would


able


to gain


some


insights


about


his


own


return


examining


how


estimation


S and


X affects


underwriting


sk premium


on a particular


issue.


Conclusion


Investment


services;


bankers


provide


originating


three


managing,


distinctive

underwriting


financing


a narrow


sense,


and


stributing


services.


Underwriting


service


constitutes


taking


of risk


to provide


long-term


financing


issuing


firm,


namely


an insurance


service


bearing


risk


adverse


price


fluctuations.


This


study


focuses


relationship


between


under-


writing


service


and


underwriter


appropriate


compensation


widely-used


negotiated


new


equity


offerings.


commitment


this


underwriting


purpose,


contract


a valuation


was


model


developed


the


and


firm


applied


was









This


study


finds


that


value


underwriting


contracts


was


significantly


overpriced,


even


in the


extreme


case


when


underwriting


expenses


were


taken


into


account.


argue


that


these


excess


returns


are,


at least


in part,


due


monopolistic


nature


underwriting


busine


ss.


This


study


also


finds


that


predicted


underwriting


value


varies


significantly


with


issues


made


different


types


different

different


firms


types


methods


industrial,


financial,


registration


syndication


(shelf


or utility),

or nonshelf),


syndicated


by

by


or nonsyndi-


cated),


and


managing


underwriter


s prestige


(special,


major,


or submajor


bracket).


was


particularly


interesting


to find


that


while


underwriting


firms


different


prestige


groups


compete


with


one


another,


they


seem


to compete


more


frequently


with


other


firms


same


group.


This


lends


support


to previous


findings


on the


competitiveness


investment


banking


industry


sum,


particularly


underwriting


those


business


investment


yields


bankers


a handsome


able


profit,


to obtain


large


underwriting


participation


(typically,


managing


underwriters


or lead


underwriters)


It also


appears


that


underwriting


is not


only


a profitable


business


but


also


growing


sources


revenue


investment


banking


firms.















CHAPTER


THE


UNDER


CHOICE


ASYMMETRIC


OF AN OPTIMAL


INFORMATION


: AN


FINANCING
EMPIRICAL


METHOD


INVESTIGATION


Introduction


Background


Study


The


question


method


optimal


issuing


firm


choice


long


new


been


equity


financing


a controversial


one


has


received


considerable


attention


finance


literature.


Firms


are


faced


with


choice


to offer


a new


equity


issue


either


directly


firm'


existing


share-


holders


a right


offer


or to investors


at large


underwritten


cash


offer.


The


former


method


includes


standby


underwritten)


rights


ers


and


non


-standby


(non-


underwritten


or uninsured)


rights


offers


latter


method


includes


firm


commitment


fully


underwritten)


cas


h offers


and


best


effort


offers


The


main


question


involved


why


majority


. firms


employ


underwritten


cash


offers


rather


than


seemingly


ess


expensive


rights


offers


The


U.S.


predominant


firms


use


been


underwritten


documented


cash


several


offers


previous


studies.


Smith


1977)


reports


that


over


% of equity









offerings


over


period


1971-1985


were


underwritten


even


though


out-of-pocket


expenses


an underwritten


equity


issue


were


three


thirty


times


higher


than


those


a nonunderwritten


rights


issue.


Booth


Smith


(1986)


present


similar


evidence


that


over


1977


-1982


period


underwritten


issues


equity


offerings


are


more


common


than


other


issues


combined,


regardless


whether


they


are


initial


or non-initial


registered


offerings.


The


choice


issues


vast


underwritten


majority


cash


U.S.


offers


firms


new


over


equity


less


expensive


rights


offers


has been


referred


so-


called


"equity


financing


paradox.


A number


paradoxical


questions


choice


regarding


issue


this


arrangement


apparently


arise:


observed


choice


financing


method


inconsistent


with


rational,


expected-utility


maximizing


behavior


owners


issuing


offering


firms


method


Should


order


issuing


to avoid


firms


high


employ


underwriting


rights


and


tribution


expenses


associated


with


underwritten


cash


offering?


Although


some


potential


explanations


to resolve


this


paradox


have


been


provided


several


empirical


studies


(among


others


Smith


1977),


Hansen


and


Pinkerton


1982),


Bhagat


1983)),


theoretical


models


with


rational









expectation


formation3


have


been


tested


to date.


Most


prior


studies

or more


lack

forms


convincing


market


theoretical analysis

imperfections which


and

may


ignore

underlie


one

the


observed


behavior


firms


choosing


their


optimal


financing


method.


The


observed


behavior


most


firms


choosing


between


rights


offer


and


underwritten


cash


offer


is not


easily


justified


a market


symmetric


information.


It i


frequently


noted


that


with


regards


valuation


new


equity


issues


, the


new


issue


market


characterized


informational


asymmetry


between


insiders/managers


and


outside


investors.


The


fundamental


problem


with


rai


new


equity


capital


that


insider


and


shareholders


act


potential


opportunism


exploiting


better


information


firm


(see


Myers


and


Majluf


1984)).


In this


regard,


without


incorporating


information


asymmetry


problem


new


tory


equity


enough


market,


to explain


analysis


observed


would


satisfac-


behavior


firms.


' According *
expectations
the available


to the general
, expectations
e relevant inf


hypothesis


are


formed


ormation


rational


on the


concerning


basis
the


variable


being
agents


predicted


use


and,


available


furthermore,
information


individuals
intelligently


or economic
This









Objective


Study


purpose


this


study


to empirically


examine


table


hypothe


ses


regarding


which


type


financing


method


a firm


employs


a given


issue


under


asymmetric


informa-


tion


between


issuers


and


investors/investment


bankers.


There


a growing


body


literature


which


analyzes


various issues

Recently, the


related


role


asymmetr


information as

ic information


ymmetry


has


problem.


been


expanded


Heinkel


and


Schwartz


1986;


H&S,


hereafter)


justify


firm


s behavior


choice


a financing


method.


H&S


focus


on a information


problem


whi


issuer


knows


more


about


firm


quality


than


either


investors


or investment


bankers.


Their


model


predicts


cross-sectional


relationships


between


firm


quality


and


choice


a financing


method


among


three


alternative


types;


standby


rights


offer,


non-s


tandby


rights


offer,


firm


commitment


cash


offer.


Their


analysis


also


helps


to explain


simultaneous


existence


above


three


types


financing


standby


arrangement


rights


and


offers


reason


to set


firms


subscription


using


pri


non-


ces


enough


ensure


success


offer.


Resting


on the


theoreti


analyse


presented


H&S


signaling


equilibrium


model,


this


study


analyzes


whether


observed


behavior


firms


consistent


with


mpirical


__


--


_


_ __


rw








sectional


relationship


between


firm


quality


and


firm'


choice


a financing


method.


(2) S

using


with


ubscription


price


a non-standby


subscription


hypothesis


rights


price


offer


that


The

pos


each


quality

itively


firm


a firm


correlated


sets


Firm


segment


hypothesis


The


proportion


firms


using


standby


varies


rights


inversely


offers


with


population


magnitude


issuing


investigation


firms


cost.


Note


that


since


these


three


hypotheses


are


not


mutually


exclusive,


could


that


each


hypothesis


works


well


whole


or only


a subset


sample.


empirically


investigating


above


hypotheses,


this


study


attempts


answer


following


questions


issuing


firm


s choice


of a financing


method


significantly


related


quality


(type)


each


firm?


What


type


firms


using


uninsured


rights


offers


higher


subscription


prices


How


does


magnitude


investigation


costs


associated


with


standby


rights


offers


affect


number


firms


using


this


financing


method?


rest


this


study


organized


follows


Part


reviews


financing


previous


method.


work


A brief


related


choice


discussion


theoretical


model


developed


Heinkel


and


Schwartz


1986)


their


empirically


testable


hypotheses


are


presented


Part


III.











Literature


Review


There


are


bodies


literature


which


examine


various


aspects


sing


new


equity


capital


firms


The


st body


literature


(Mandelker


and


Raviv


(1977),


Baron


(1979),


Baron


and


Holmstrom


1980),


Baron


(1982))


characterizes


optimal


contract


between


issuer


and


investment


banker


context


of a negotiated


sale.


Their


models


focus


on the


uncertainty


associated


with


new


issue


and


on the


conflict


interest


between


parties


involved.


Mandelker


an investment


Raviv


banking


analyze


contract


risk


under


sharing


features


assumption


that


parties


have


symmetric


information


time


contracting.


They


derive


certain


conditions


under


which


various


issue


arrangements


firm


commitment,


best


efforts,


standby


contracts4)


are


Pareto


optimal.


It is


shown


that


optimal


contract


a function


both


riskiness


a particular


issue


and


risk


attitudes


parties


involved.


Baron


(1979)


also


assumes


symmetric


information


between


contracting


parties


but


considers


incentive


problem


resulting


from


issuer


inability


to observe


investment


banker


s distribution


effort.


argues


that








order


to mitigate


incentive


problem,


issuer


must


sacrifice


some


gains


from


optimal


risk


sharing


first-


best contract,


which


would


induce


banker


to expend


more


effort.


The


optimal


commis


sion


function


this


case


shown


to be


an increasing


function


the


net


proceeds


from


suing


securities


with


a bonus


involved


issue


sold


offer


price.


Baron


and


Holmstrom


1980)


assume


symmetric


information


time


contracting,


but


allow


investment


banker


to obtain


better


information


about


demand


issue


conducting


preselling


activities


during


registration


period.


Their


model


shows


that


due


to information


dis-


advantage,


issuer


must


design


contract


that


distribution


effort


and


offer


price


decisions


made


banker


serve


interests


issuer


and


that


optimal


to delegate


offer


price


deci


sion


investment


banker.


Baron


1982)


analyzes


an optimal


delegation


contract


advi


sing


and


distribution


services


between


issuing


firm


and


investment


banker


better


banker


informed


under


about


assumption


market


that


demand


firm


s securities


than


issuer.


analysis


predicts


that


price


value


decision


issuer


banker


of delegating


an increasing


offer


function









papers


discussed


above


mainly


analyze


optimal


contract


between


issuing


firm


and


underwriter


seeking


answer


following


question:


Given


fact


that


firm


needs


underwriting,


offer


pricing,


or distributing


services


underwriter,


how


different


characteris-


tics


both


parties


and


issue


itself


determine


optimal


form


contract?


Since


these


studies


not


explicitly


consider


firm


s behavior


choosing


between


rights


offer


and


underwritten


offer,


they


not


provide


any


potential


explanation


to resolve


equity


financing


paradox.


second


body


literature,


focusing


mainly


difference


in issuing


costs


rights


offers


versus


underwritten


cash


offers,


puts


forth


some


potential


explanations


problem


equity


financing


paradox.


Smith


1977)


suggests


a monitoring


cost


hypothesis


that


"from


shareholder


s standpoint,


firm


use


underwriter


optimal


because,


even


though


managers


may


receive


benefits


that


not


accrue


owners


firm,


financing


cost


method


monitoring


incurred


management


shareholders]


choice


exceeds


savings


out-of-pocket


expenses


from


using


rights.


298)


Thus,


argument


that


less


costly


rights


offering


method


bypassed


managers at


ne expense


" (p.









existing


shareholders,


who


must


otherwise


incur


even


greater


monitoring


costs


to limit


manager


s behavior.


However,


monitoring


First,


as Smith


cost


since


1977)


hypothesis


costs


himself


presents


monitoring


point


out,


a couple


management


problems.


are


observable,


hypothesis


not


directly


testable.


Second,


consideration


competition


market


management


should


reduce


required


monitoring


expen


ses


see


Fama


(1980


and


Jensen


and


Meckling


(1976)),


and


thus


plausibility


hypothesis.


Contending


that


Smith


s estimation


out


-pocket


expenses


misleading


due


to a systematic


difference


samples


approach


firms


this


involved,


paradoxical


Hansen


problem


Pinkerton


examining


1982


share-


ownership


concentration


issuing


firms.


They


argue


that


right


offers


are


advantageous


only


closely-held


firms,


which


have


a substantial


comparative


flotation


cost


advantage


due


presence


a large


block


centrally


controlled


common


stocks.


They


show


that


majority


firms


using


underwritten


cash


offers


would


have


incurred


significantly


higher


costs


these


firms


had


used


alternative


rights


offerings.


Their


results


indicate


that


issuing


firm'


choice


between


rights


offers


and


underwritten


offers


can


explained


different


degree


share-









However,


Hansen


and


Pinkerton'


comparative


cost


hypothesis


has


been


fully


supported


subsequent


studies


Smith


and


Dhatt


1984),


Bhagat


and


Frost


1986),


Bhagat


(1983).


Smith


and


Dhatt


(1984)


doubt


about


cost


function


used


Hansen


and


Pinkerton


arguing


that


functional


form


significantly


overstates


forecasted


issuing


cost


rights


offerings


firms


with


low


shares


central


control.


Their


results,


using


alternative


specifications,


reject


comparative


cost


hypothesis


Hansen


and


Pinkerton.


Recognizing


that


"cost


differences


between


rights


vs.


underwritten


cash


offering


methods


expected


future


earnings


would


capitalized


into


share


price


the


time


announcement


pre-emptive


right


amendment"


296),


Bhagat


(1983)


investigates


effect


pre-emptive


right


amendments


estimate


on shareholder


financing


cost


wealth


differences


to indirectly


between


methods.


He finds


that


amendments,


on average,


decrease


shareholder


wealth,


implying


that


net


benefits


to existing


shareholders


are


greater


from


rights


offerings


than


from


underwritten


cash


offering


method.


This


result


again


indicates


that


financing


costs


rights


offers


are


less


than


those


alternative


method.









The


empirical


evidence


Bhagat


does


not


support


Hansen


Pinkerton


s hypothesis,


which


can


be viewed


consistent


with


shareholder


wealth


maximization


hypothesis


amendment


Bhagat


should


lead


s study


that


an increase


pre-emptive


in shareholder


right


wealth.


Bhagat


and


Frost


1986)


find


that


issuing


costs


under-


written


equity


cash


offerings


to exi


sting


shareholders


are


determined


risk


offerings


and


information


costs,


well


as by


only


size


determinant


offering,


reported


which


flotation


identified


costs


in the


Hansen


Pinkertons'


study.


this


regard,


Bhagat


and


Frost


argue


that


since


Hansen


Pinkerton


ignore


first


determinants


issuing


costs,


their


model


misspecified.


A major


drawback


on the


studies


discussed


above


that


their


analyses


not


incorporate


an important


characteris-


new


equity


market,


namely,


market


imperfection


asymmetric


information


investors/investment


banker,


between


which


issuer


may


and


underlie


outside


observed


behavior


firms


in choosing


their


optimal


financing


method.


In this


regard,


their


analyses


seem


to be


unsatisfactory


enough


to explain


observed


behavior


firms.


__











The


Theory


and


Testing


Hypotheses


A number


previous


studied


second


body


literature


reviewed


Section


II),


which


lack


convincing


theoretical


analysis


and


suffer


some


potential


problems


with


sample


selection


attempted


bias


or inappropriate


to investigate


functional


equity


forms,


financing


have


paradox,


with


inconclusive


results.


As set


forth,


argument


prior


studies


centers


cost


difference


between


rights


offers


and


underwritten


cash


offers


without


explicit


consideration


on information


asymmetry


problem


new


issue


market.


extent


that


is difficult,


possible,


measure


relevant


may


financing


quite


costs


leading


a new


there


equity


are


issue,


omitted


result


or misestimated


costs


and


benefits.


The


analysis


Heinkel 1


and


Schwartz


(1986)


does


not


fully


rely


on estimation


financing


costs


and


thus


relatively


free


from


possible


cost


estimation


problem


when


one


investigates


their


empirical


implications.


This


part


briefly


discus


ses


rationale


behind


information


equilibrium


model


developed


H&S


and


presents


empirical


hypotheses


about


choice


a financing


method.