Tài liệu Review of Accounting and Finance Volume 8 ppt

Previous research has shown show that the three-day decline was the largest in more than 40 years, large enough that the drop was news itself the October 16, 1987 drop immediately before

Access this journal online _ 119

Editorial advisory board _ 120

Introduction: the 1987 market crash:

20 years later _ 121

What caused the 1987 stock market crash and

lessons for the 2008 crash

Ryan McKeon and Jeffry Netter 123

Has the 1987 crash changed the psyche

of the stock market? The evidence from

initial public offerings

James Ang and Carol Boyer _ 138

Capital market developments in the

post-October 1987 period:

a Canadian perspective

Laurence Booth and Sean Cleary _ 155

Review of Accounting and

CONTENTS

The current and past volumes of this journal are available at:

www.emeraldinsight.com/1475-7702.htm

Find more articles from this journal and search through Emerald’s

200+ journals at: www.emeraldinsight.com

See page following contents for full details of what your access includes

Access this journal electronically

This journal is a member of and subscribes to the principles of the Committee on Publication Ethics

Revisiting derivative securities and the 1987 market crash: lessons for 2009

G Glenn Baigent and Vincent G Massaro _ 176

Fraudulent financial reporting, corporate governance and ethics: 1987-2007

Lawrence P Kalbers _ 187

CONTENTS

continued

As a subscriber to this journal, you can benefit from instant,

electronic access to this title via Emerald Management Xtra Your

access includes a variety of features that increase the value of

your journal subscription.

How to access this journal electronically

To benefit from electronic access to this journal, please contact

support@emeraldinsight.com A set of login details will then be

provided to you Should you wish to access via IP, please

provide these details in your e-mail Once registration is

completed, your institution will have instant access to all articles

through the journal’s Table of Contents page at

www.emeraldinsight.com/1475-7702.htm More information

about the journal is also available at www.emeraldinsight.com/

raf.htm

Our liberal institution-wide licence allows everyone within your

institution to access your journal electronically, making your

subscription more cost-effective Our web site has been

designed to provide you with a comprehensive, simple system

that needs only minimum administration Access is available via

IP authentication or username and password.

Emerald online training services

Visit www.emeraldinsight.com/help and take an Emerald online

tour to help you get the most from your subscription.

Key features of Emerald electronic journals

Automatic permission to make up to 25 copies of individual

articles

This facility can be used for training purposes, course notes,

seminars etc This only applies to articles of which Emerald owns

copyright For further details visit www.emeraldinsight.com/

copyright

Online publishing and archiving

As well as current volumes of the journal, you can also gain

access to past volumes on the internet via Emerald Management

Xtra You can browse or search these databases for relevant

articles.

Key readings

This feature provides abstracts of related articles chosen by the

journal editor, selected to provide readers with current awareness

of interesting articles from other publications in the field.

Non-article content

Material in our journals such as product information, industry

trends, company news, conferences, etc is available online and

can be accessed by users.

Reference linking

Direct links from the journal article references to abstracts of the

most influential articles cited Where possible, this link is to the

full text of the article.

E-mail an article

Allows users to e-mail links to relevant and interesting articles to

another computer for later use, reference or printing purposes.

Structured abstracts Emerald structured abstracts provide consistent, clear and informative summaries of the content of the articles, allowing faster evaluation of papers.

Additional complimentary services available

Your access includes a variety of features that add to the functionality and value of your journal subscription:

Xtra resources and collections When you register your journal subscription online you will gain access to additional resources for Authors and Librarians, offering key information and support to subscribers In addition, our dedicated Research, Teaching and Learning Zones provide specialist ‘‘How to guides’’, case studies and interviews and you can also access Emerald Collections, including book reviews, management interviews and key readings.

E-mail alert services These services allow you to be kept up to date with the latest additions to the journal via e-mail, as soon as new material enters the database Further information about the services available can be found at www.emeraldinsight.com/alerts Emerald Research Connections

An online meeting place for the world-wide research community, offering an opportunity for researchers to present their own work and find others to participate in future projects, or simply share ideas Register yourself or search our database of researchers at www.emeraldinsight.com/connections

Choice of access

Electronic access to this journal is available via a number of channels Our web site www.emeraldinsight.com is the recommended means of electronic access, as it provides fully searchable and value added access to the complete content of the journal However, you can also access and search the article content of this journal through the following journal delivery services:

EBSCOHost Electronic Journals Service ejournals.ebsco.com

Informatics J-Gate www.j-gate.informindia.co.in Ingenta

www.ingenta.com Minerva Electronic Online Services www.minerva.at

OCLC FirstSearch www.oclc.org/firstsearch SilverLinker

www.ovid.com SwetsWise www.swetswise.com

Emerald Customer Support

For customer support and technical help contact:

E-mail support@emeraldinsight.com Web www.emeraldinsight.com/customercharter Tel +44 (0) 1274 785278

www.emeraldinsight.com/raf.htm

#Emerald Group Publishing Limited

EDITORIAL ADVISORY BOARD

Ali Abdolmohammadi Bentley College, USA Pervaiz Alam Kent State University, USA Sharad Asthana University of Texas at San Antonio, USA Tim Cairney

Georgia Southern University, USA Hsihui Chang

Drexel University, USA Rong-Ruey Duh National Taiwan University, Taiwan Mahmud Ezzamel

Cardiff University, UK Ehsan Feroz University of Washington, Tacoma, USA Liming Guan

University of Hawaii at Manoa, USA Mahendra Gujarathi

Bentley College, USA William Hopwood Florida Atlantic University, USA Marion Hutchinson

Queensland University of Technology, Australia

George Iatridis University of Thessaly, Greece Hoje Jo

Santa Clara University, USA Laurie Krigman

Babson College, USA Krishna Kumar George Washington University, USA Marc LeClere

Valparaiso University, USA Joseph McCarthy Bryant University, USA Gordian Ndubizu Drexel University, USA Hector Perera Macquarie University, Australia Alan Reinstein

Wayne State University, USA Herve´ Stolowy

Group HEC (Hautes Etudes Com), France Nikhil Varaiya

San Diego State University, USA Huai Zhang

Nanyang Technological University, Singapore

# Emerald Group Publishing Limited

Introduction: the 1987 market

crash: 20 years later

Society relies on well-functioning capital markets to promote economic progress in

businesses and households To that goal, academics argue that capital markets should

provide for price discovery and liquidity, where the best way to find out what an asset

is worth is to attempt to sell it As long as there are a large number of market

participants, bidding among them leads to price discovery, and an asset is sold quickly

resulting in liquidity Moreover, in a well functioning market the price should be close to

its intrinsic value But academic assumptions aside, is it not the case that institutional

and private investors have the same expectations of our secondary markets?

For both institutional and private investors, capital markets are the domicile of our

wealth Capital markets reflect the performance of individual firms and the investment

choices they make on behalf of shareholders Markets reflect the value of retirement

accounts such as 401 ks, 403 bs, or RRSPs in Canada On a macro scale, capital markets

are an indicator of the expectations of future earnings The well-being of capital

markets is of critical importance to all, even the US Treasury Department and Social

Security

Having turned the generational clock in 2007, it seemed appropriate to revisit the

events of 1987 The literature seemed to be mixed as to the cause of the 1987 crash, new

streams of literature such as behavioral finance have evolved, and many structural

changes have occurred Ironically, while all of the article reviews for this issue were in

progress, the factors which cause market crashes or corrections became more

important because we were witnessing a capital markets crisis in 2008 In most cases

the authors had the difficulty of drawing their analyses to a close because each day

there was more to add to the literature But here we are, and we must conclude There is

a confluence to the articles in this edition – each in some way speaks to the issue of

efficient capital markets

McKeon and Netter have provided an extension to earlier work by Mitchell and

Netter (1989) The current research reinforces the view (espoused in the 1989 paper)

that relevant news caused the market crash in 1987, but they find that significant

changes in market movements and volatility are associated with the market correction

in 2008

The ‘‘something is different now’’ theme is continued by Booth and Cleary They

report that significant structural changes have occurred in the Canadian economy

since 1987 Their analysis documents macroeconomic changes and speaks to the

impact of fiscal policy and monetary policy on the resilience of the Canadian economy,

which they describe as more resilient today than in 1987

Ang and Boyer examine an issue that was critical in 1987 – IPOs They show that

the 1987 market crash changed the psyche of the IPO market as evidenced by fewer

IPOs from riskier firms Following the crash, they find more rational pricing in the

context of smaller discounts and smaller mean reversion Clearly, this speaks to a

behavioral component in asset prices Let’s hope that the current crisis leads to more

rational pricing

The behavior of investors is continued by Baigent and Massaro who suggest that

the existence of portfolio insurance can create more aggressive trading and a moral

Lastly, there have been significant regulatory changes in the capital markets since

1987 as documented by Kalbers To the point, regulations are intended to provide formore accurate accounting information, but Kalbers opines that regulation occurs afterthe damage has occurred

In the mid 1960s Eugene Fama formulated the efficient market hypothesis in whichmarkets reflect all relevant information The problem seems to be the information, notthe markets

G Glenn Baigent and Vincent G Massaro

Guest Editors

What caused the

1987 stock market crash

What caused the 1987 stock

market crash and lessons for the

2008 crash

Ryan McKeon

School of Business Administration, University of San Diego, San Diego,

California, USA, and

Jeffry Netter

Terry College of Business, University of Georgia, Athens, Georgia, USA

Abstract

crash in 1987, update the empirical support for that argument, and compare to recent market

developments.

one-day S&P 500 drop in percentage terms in history (20.47 percent) there was also a large market

drop (10.12 percent) in the three trading days before the 1987 crash Previous research has shown

show that the three-day decline was the largest in more than 40 years, large enough that the drop was

news itself (the October 16, 1987 drop immediately before the crash was also an extremely large

one-day decline) The theoretical model of Jacklin et al show how a surprise significant drop in the

market could have provided information to the market that could directly lead to an immediate crash.

three-day period in the 20 years after 1987 had as large a market decline The paper documents the

large market movements and volatility in the period beginning in fall 2008 and suggests that this

‘‘crash’’ is different than what occurred in 1987.

about the causes of the 2008 crash.

implications for the 2008 decline The 1987 crash was due in part to characteristics news but also to the

market and trading strategy, the 2008 ‘‘crash’’ is more likely a response to fundamental economic news.

1987 crash.

Keywords Stock markets, Stock prices, Take-overs, Regulation, Financial modelling,

United States of America

Paper type Research paper

The cover story from the Newsweek (1987) issue that was released the weekend directly before

the October 19, 1987 crash was titled ‘‘Is the party over?’’ The second paragraph of the article

starts, ‘‘The cascading Dow and record trading volume marked a major shift in psychology

and sent a powerful shiver across the country’’ (Dentzer, et al., 1987)

1 Introduction

On Monday October 19, 1987, the US equity market suffered its largest single-day

percentage decline in history The S&P 500 index fell by 57.86 points, a decline of 20.46

The current issue and full text archive of this journal is available atwww.emeraldinsight.com/1475-7702.htm

The authors would like to thank Glenn Baigent, Annette Poulsen, Janis Zaima, and a referee for

helpful comments and suggestions on the paper

Mitchell and Netter (1989) argue that this three-day decline was an importantcontributing factor to the crash in fact, they describe the decline as a ‘‘trigger’’ In thispaper, we review this argument, provide simple descriptive evidence supporting theargument and suggest how October 1987 is different from the market decline in late 2008.

We report data that the drop in the stock market immediately preceding the October 19,

1987 crash that others have shown was very large in historical terms remains one of thelargest declines over the next 20 years Additionally, we document the unprecedentedlevel of volatility since August 2008 and show how it is different from 1987

The October 19, 2007 market crash of more than 20 percent did not seem to berelated to any fundamental news However, Mitchell and Netter (1989) argue that thethree-day decline preceding the crash was a large enough decline that it became thefundamental news and that shook the market The theoretical model of Jacklin et al.(1992) (among others) shows how a surprise significant drop in the market could haveprovided information to the market that would directly lead to a crash In this paper, wepresent evidence that even 20 years later, the magnitude of the market declineimmediately preceding the 1987 is still a significant outlier  only one three-dayperiod in the 20 years after 1987 had as large a market drop

Jacklin et al.’s model suggests that the sharp market decline preceding the 1987crash revealed the effects of new investment strategies by investors that had not beenfully anticipated by the market (they build on Grossman’s (1988) model of the effects ofimperfect information about portfolio insurance) This revelation to investors of theextent of dynamic hedging caused investors to dramatically revise downward theirstock valuations Other explanations of the 1987 crash include liquidity problems (thePresidential Task Force on Market Mechanisms (1988) The Brady Report) in tradingwhen volume increased tremendously (perhaps as the result of portfolio insurancetrading), or changed investor psychology or some combination of all the theories.However, each of the theories is consistent with the effects of a large downward marketmovement directly preceding the crash that was significant and unexpected, triggeringthe October 19 crash

The paper proceeds as follows In section 2, we examine possible reasons for the

1987 crash, providing a general discussion on what causes large market movements,and reviewing the Mitchell and Netter work on the 1987 crash In section 3, we examinetrading volume and market volatility since the 1987 crash, including the extraordinarymarket events of the fall of 2008 We conclude in section 4

2 Explanations for the October 1987 crashThere are at least three general views of the causes of the stock market crash onOctober 19, 1987 The views are not mutually exclusive One is the efficient marketstory the market reacted to some fundamental news that led market participants torevalue stocks down by more than 20 percent in one day A second is a liquiditystory for some reason, probably a large number of sell orders, liquidity declinedsignificantly, depressing prices A third is some variant of a behavioral financestory investors acting irrationally either drive prices up too high, followed by asignificant fall, or panic and sell for some reason, significantly depressing prices

What caused the

1987 stock market crash

125

2.1 Explanations of large market-wide stock-price movements

Cutler et al (1989) analyze the question of what fundamentally causes large stock price

movements in a paper that followed soon after the 1987 crash Their general conclusion is

that we are not very good at explaining large stock price movements, questioning the

‘‘efficiency’’ of the market They first consider the impact of macroeconomic news on the

stock market The paper examines the relation of macroeconomic fundamentals as

dividend payments, industrial production, real money supply, long and short-term interest

rates, inflation and stock market volatility They conclude that these macroeconomic

variables are not statistically meaningful in explaining stock market returns

Cutler et al also conduct a less formal analysis of the impact of ‘‘big news’’ on the

stock market Using the World Almanac as a source of significant news stories, the paper

narrows its selection of ‘‘big news’’ to those stories which were featured on the front page

of the New York Times or were the lead story in the business section of the paper Few if

any of the stock market returns on these days are comparable to the returns seen in

October 1987 The paper then reverses the analysis, examining the largest single-day

movements of the S&P 500 index and examining the New York Times for an explanation

of the event The explanation they find for the October 19, 1987 crash is ‘‘Worry over

dollar decline and trade deficit; Fear of US not supporting the dollar’’

Cutler et al conclude that it is difficult to explain large price movements even after

the fact An interpretation of their results is that if one cannot explain market

movements after the fact, when news has been revealed, and markets do not move

much in response to large news stories, it is difficult to argue that fundamentals drive

markets, at least in the time of extreme movements

Haugen et al (1991) perform a similar analysis, but focus on stock market volatility

rather than returns The authors document the largest single day shifts in stock market

volatility and search for contributing explanations When the authors are able to match

large increases in volatility with a well-documented event, the event tends to be an act

of warfare, a natural disaster or an assassination The paper also notes that large

decreases in volatility can generally be matched to political enactments or

proclamations This latter finding suggests a possible role for regulatory entities in

‘‘calming’’ markets during volatile times

2.2 The cause of the market decline October 14-16, 1987

Mitchell and Netter (1989) provide a case study of one market movement in contrast to

the more general but less detailed analysis of Cutler et al Mitchell and Netter provide

both cross-sectional and time series evidence supporting their hypothesis that the very

large October 14-16, 1987 market decline was due to an unexpected proposal in the

House Ways and Means Committee to end the tax deductibility of debt used in

takeovers Noting that the October 14-16, 1987 period represented the largest one-,

two-and three-day declines (–5.16,8.11 and 10.44 percent, respectively) since the French

army’s defensive position was unexpectedly compromised in May 1941, in WW II, the

paper examines the possible causes of this decline The paper concludes that the

market reacted negatively to news that a bill ending the interest deductibility of debt in

takeovers was unexpectedly proposed by the US House Ways and Means Committee

and was likely to pass Therefore, it was the prospect of this bill being signed into law

that may have caused the market decline of October 14-16, 1987, which then led to the

subsequent more drastic decline on Monday October 19

On the night of October 13, 1987 the House Ways and Means Committee introduced

a tax bill that had several provisions designed to restrict takeovers, especially ending

Two other papers, Miller and Mitchell (1999) and Mitchell et al (2007), examine inmore detail how the news about the tax could have caused a major decline in themarket Miller and Mitchell (1999) examine whether fundamental news couldconceivably explain the market movements of October 1987 and show that fairlymodest changes in expected future cash flows or discount rates can result in largemarket revaluations The authors state, ‘‘while it might first seem that one should beable to identify the shocks to fundamental factors that can cause such a dramatic pricedecline, the above analysis suggests that these shocks do not necessarily have to bedramatic themselves’’ Mitchell et al (2007) analyze how costly arbitrage mispricing

‘‘can be large and can extend for a long period’’ They consider several examples,include the stock market crash They show how the tax bill caused merger arbitragers

to sell on October 14, 1987 through October 16, 1987 and the selling increased on theOctober 19 accelerating the price decline

Roll (1989), however, argues that the international nature of the decline over theweekend of October 17 and 18 is not consistent with the takeover-tax story Mitchelland Netter note however that the world decline occurred after the US decline and wasmuch smaller in magnitude (an equally weighted world index fell 2.03 percent).Further, as we discuss below, the world movement may be consistent with the triggerstory told my Mitchell and Netter the large three-day decline started the wholemarket downward

2.3 The effect of the market decline immediately before October 19, 1987Grossman (1988) models a situation where the amount of dynamic hedging undertaken

by traders is not public knowledge until they trade on these strategies Whencoordinated selling occurred based on these strategies (e.g during a big market decline)liquidity issues will further depress the market Grossman notes that if there were moredynamic hedging strategies in place than anticipated by traders, traders might beunable to execute all the dynamic hedging they had planned, increasing marketvolatility Jacklin et al (1992) also address the situation where a market hasunderestimated the amount of dynamic hedging strategies such as portfolio insurance,and the true amount if the hedging is revealed However, unlike Grossman theyconcentrate on the effect the revelation of the information about the extent of dynamichedging will have on traders’ valuation of securities In this case, the market willdecline to reflect the information that the market was overvalued Jacklin et al note thelarge decline on October 16, 1987 fits with their model While the liquidity problemspointed out by the Brady Commission (1988) and Grossman played a role in the crash,

it is unlikely they were enough to cause the crash on their own

What caused the

1987 stock market crash

127

3 Large market movements since 1950

In this section we report on market movements and volatility since 1950, concentrating

on evidence post-1987 Our goal is two-fold First, we provide new evidence that

supports the argument that the market drop in the trading period immediately before

the 1987 crash was an unusually large decline Second, we report recent evidence on the

unprecedented nature of the September–November 2008 ‘‘crash’’ Note here our single

most important piece of evidence (contained in Table IV) about the 1987 crash is that

the magnitude of the three-day decline immediately preceding the 1987 crash was

larger than any three-day decline in the 20 years after the 1987 crash

The years since the market crash of October 1987 have served to further strengthen

the argument that the period from October 14-16, 1987 was an unusually large market

decline, which very likely then precipitated the crash on Monday, October 19, 1987

While the 20 years since the crash have seen the volume of trading on the NYSE

increase dramatically, and while there have been episodes or days with large market

movements, the October 14-16, 1987 period still ranks high amongst periods of severe

market decline While an argument can be made that the infrastructure of the market

has improved over the last 20 years so that the market is better able to cope with

episodes of high volume and illiquidity, it is also fair to say that the three-days

preceding Monday October 19, 1987 remain an extraordinary period of decline in the

history of the stock market In this section we explore both of these issues: the volatility

and resiliency of the market since the crash, and the unusual nature of trading on

October 14-16, 1987, which immediately preceded the crash

3.1 Volatility and volume

Table I reports the largest single-day negative returns for the S&P 500 index since

1950, over various intervals This table concentrates on single day movements, not as

much a focus of Mitchell and Netter (1989) who concentrate on the three-day window

when tax news reached the market, but potentially related to the ‘‘trigger argument’’ In

the period from 1950 through July 30, 2008 October 16, 1987 still appears to have been a

very large decline There are nine days (excluding October 19, 1987) on which the

market suffered a greater decrease than October 16, 1987 However, these decreases

were not followed by the kind of crash seen in October 1987 Indeed, with only one

exception, the market rebounded on the following days, as indicated in the table The

story is different for the market conditions seen since August 1, 2008, when there are

ten days with a greater decline than the October 16, 1987 decline We discuss this

period in greater detail in section 3.3

Table II further illustrates this point The Table highlights all the single days since

1987 in which the market (as measured by the S&P500 Index) declined by 3 percent or

more Note that in this table the period since the end of July 2008 is considered

separately As of November 30, 2008, there had been 51 such negative-return days (less

than 3 percent), and 20 of these days have occurred since the beginning of August 2008

Furthermore, there were 121 two-day periods between January 1988 and November 30,

2008, which saw a cumulative market return of 3 percent or less The number of

three-day periods where the market fell by 3 percent or more during this same period is

216 In other words, the market has certainly seen significant swings since October

1987, but without generating another similar crash

Notable features of the October 1987 market crash are the volume of trading and the

level of volatility exhibited by the market Figure 1 illustrates the time series of daily

trading volume from January 1968 to December 1987 It illustrates the dramatic spike

The unusually high levels of trading in October 1987 are again illustrated in Figures

2 and 3, which use October 1987 as a starting point and chart the daily volume oftrading in S&P 500 stocks until November 30, 2008 Although there is a clear upward

The data are split into a Pre-August 2008 period, which covers the period from January 1st 1950

to July 30, 2008, and a Post-July 2008 period, which covers the period from August 1, 2008S to November 30, 2008; October 19 and October 16, 1987 rank 1st and 11th, respectively on the Pre-August 2008 list

What caused the

1987 stock market crash

129

Table II.Days post-1987 in which the S&P 500 index has registered a decline of 3 percent or more

Since 1987, noticeable spikes in trading volume have been observed on several days:

. January 4, 2001: volume reached 2,131,000,064 shares as the S&P returned1.06percent This followed a relatively large rise in the market on January 3, 2001of5.01 percent (volume of 1,880,700,032)

. September 1, 1998: volume reached 1,216,600,064 shares as the S&P returned3.86 percent This followed a relatively large fall in the market on August 30 of

Notes: This table reports the details for each trading day on the S&P 500 since November 1,

is the trading volume as number of shares traded; one day return is the return for the trading day calculated from the previous trading day’s closing level to the actual trading day’s closing level

Figure 1.

Daily volume of S&P 500

shares January 1, 1968

-October 31, 1987

What caused the

1987 stock market crash

131

. August 16, 2007: volume reached 6,509,300,000 as the S&P returned 0.32 percent

This followed a relatively large drop in the market on August 14 and 15 of1.82

percent and 1.39 percent, respectively (volume of 3,814,630,000 and

4,290,930,000)

. September 18, 2008: volume reached 10,082,689,600 as the S&P returned 4.33

percent This followed a relatively large drop in the market on September 17 of

4.71 percent (volume of 9,431,870,400 shares)

Figure 2.Daily volume of S&P 500 shares October 1, 1987 September 30, 2001

Figure 3.Daily volume of S&P 500 shares 1st October, 2001 November 30, 2008

Figure 4 illustrates the monthly variance of the S&P 500, using daily-realizedreturns to compute the measure October 1987 stands out as the most volatile month oftrading during the period covered ( January 1950 to November 30, 2008) It is alsointeresting to note that the second half of the 1990s saw increased levels of stockmarket volatility, although no period has ever approached the level of October 1987.Market volatility over the September–November 2008 period saw levels of volatility torival October 1987 Further discussion of this period follows in section 3.3.

Table III reports the most volatile months of trading in the S&P 500 since January

1950 October 1987 is the most volatile month of trading over this period, but all othermonths except one listed on the table have occurred since October 1987 Once againthis is an indication that the market has been comparatively volatile in periods sinceOctober 1987, without producing the crash seen in 1987 Even so, prior to September

2008, no other month on this list had exhibited even half the level of volatility ofOctober 1987

Events in September 2008 and since have produced a highly volatile market Theweek of September 15 to 19, 2008 saw firms such as Merrill Lynch, AIG and LehmanBrothers reveal deep financial troubles within their operations This followed therevelation of severe financial difficulties for Fannie Mae and Freddie Mac, which sawthe Federal government promise taxpayer-funded assistance estimated to be as much

as $100 billion for each entity (Labaton and Andrews, 2008) The months of September,October and November 2008 exhibited historic levels of volatility, even though they didnot eclipse October 1987 We will have more to say on the market volatility of thesethree months in section 3.3

What caused the

1987 stock market crash

133

3.2 October 14-16, 1987 in historical context: the following 20 years

Despite this evidence that the stock market has recently exhibited comparatively high

volatility and large single-day movements, Table IV illustrates how truly

extraordinary the market movements of October 14-16, 1987 were Table IV first

reports the ten largest three-day cumulative decreases in the S&P 500 in the years after

October 1987, starting from November 1, 1987 Data for the three months starting at

the beginning of September 2008 is presented separately

As reported in Mitchell and Netter (1989), the decline in the value of the stock

market between the October 14 and 16 was the largest three-day decrease since the

Second World War Table IV illustrates that despite the recent increases in trading

volume and the relatively high levels of market volatility from 1997 to 2003, three-day

declines of similar magnitude were rare in the 20 years following the Crash Only one

period, in late August 1998 outpaces the steep market decline of October 14-16, 1987

The August 1998 period relates to the Russian default/Long Term Capital Management

crisis[2]

Prior to the October 1987 crash, three-day negative cumulative returns in the order

of magnitude of those documented above were rare Periods of decline which rivaled

October 14-16, 1987 were largely confined to the Great Depression years of 1929-1933

As reported in Mitchell and Netter (1989) the largest three-day decline in the stock

market prior to the 1987 crash occurred in May of 1940 ‘‘when German tanks broke

through the French armies, sealing France’s fate in World War II’’ We report in Table IV

the largest three-day declines since 1950, but before October 1987 No three-day decline

is as large as October 14-16, 1987, and only one three-day period listed in this table even

makes the top 10 list reported in the first set reported in Table IV (all dates up to

September 2008 ) Note we choose the three-day period because that was the period

chosen by Mitchell and Netter since it was the period after the first tax announcement

Table III.Most volatile months of trading in the S&P 500, January 1950-November

Notes: This table reports the most volatile months of trading in S&P 500 stocks since January

1950; the measure of volatility is standard deviation, calculated from daily realized close-to-close

returns from within the calendar month; Month is the relevant calendar month and SD is

from the previous day’s closing level to day t’s closing level, and R-bar is the average daily return

within the calendar month

3.3 Market crisis in fall 2008The evidence shows that the three trading days prior to the crash on Monday October

19, 1987 was a historically large market decline This continues to be true even with 20additional years of data since the 1987 crash studied by Mitchell and Netter (1989)

Table IV.

Largest three-day

declines in the S&P 500

post 1950 for various

periods compared to

October 14-16, 1987

Largest three-day cumulative decreases in the S&P 500 since October 1987

measured from the previous day’s closing level to the day t closing level

What caused the

1987 stock market crash

135

However, market conditions in late 2008 have seen market declines and volatility,

which rival, and in some cases surpass, those of October 1987 In fact, the recent

market conditions only serve to confirm how extraordinary October 1987 was Table I

includes August to November 2008 data for the largest single-day negative returns on

the S&P 500 Note especially two fact First, there were a comparatively large number

of significant single-day market declines in the fall of 2008 Many days from August,

September, October and November would rank near the top of the list of largest

single-day negative returns on the S&P 500 since 1950 Secondly, none of these single-single-day

declines is of the magnitude of the Monday October 19, 1987 crash Therefore, the

one-day data shows that the fall of 2008 was a particularly bad time for the market, but that

the one-day market crash of October 19, 1987 is still unique

However, Table IV illustrates that the market crisis of fall 2008 has produced

three-day declines comparable to October 12-16, 1987 In particular, late August and the early

part of October 2008 saw three three-day declines larger than October 14-16, 1987, all in

a similar, over-lapping period Specifically these periods were October 3-7, 6-8 and 7-9,

2008 The decline over the November 18-20 period was also noticeably large

The months of August, September, October and November 2008 saw an unusually

high number of significant three-day declines The magnitude of the decline over the

October 14-16, 1987 period is still historically high, however, even though it has since

been joined by other comparable three-day declines The fact that none of these more

recent three-day declines triggered a market crash similar to October 19, 1987

reinforces the notion that the market is now different to how it was back in 1987

Table II provides evidence on just how volatile the market conditions of fall 2008

have been In terms of day-to-day closing levels of the S&P 500, the volatility exhibited

in September, October and November 2008 are historically high The standard

deviation (SD) of intra-month daily returns shows that these months rank 4th, 2nd and

3rd, respectively on the list of most volatile months of trading in the market since

January 1950 However, importantly, October 1987 still remains at the top of the list as

the most volatile month of trading in the market since 1950

The market crisis of 2008 has clearly led to sustained levels of high market

volatility, as shown by the high ranking of all months in this period on Table II To

some extent this was also true of the 1987 crash The subsequent months in 1987 also

saw comparatively high levels of volatility, although not on the same scale as

September–November 2008 November 1987, December 1987 and January 1988 saw the

following levels of intra-month volatility: 1.84, 1.77 and 2.18 percent, respectively

These levels would rank 20th, 23rd and 9th on the updated post-1950 list of monthly

volatility After October 1987, February 1988 saw a return to less volatility; it’s level of

0.99 percent ranks 160th It remains to be seen what will happen to the market in this

current crisis

4 Conclusion

In this paper we use the 20 years since the market crash of October 1987 to further

strengthen the argument that the period immediately before the crash (October 14-16)

period saw an unusually large market decline We review the arguments that this initial

market decline of October 14-16 precipitated the crash on Monday, October 19 We

argue that the news of the large three-day drop from the October 14 to October 16 led to

the crash on October19

Mitchell and Netter (1989) perform a detailed case study analysis of the causes of

the market decline from October 14 to16 They argue that while several factors matter,

There are several theories on how the large three-day decline could lead to a crash

on the next trading day (October 19) They center on the idea that either a significantdecline revealed negative information about the market, led to liquidity problems intrading, or changed investor psychology However, the theories rest to some extent onthe premise that the downward market movement was significant and unexpected.Critical to all the theories is that the decline of over 10 percent in the three-daysbefore the crash was very unusual Mitchell and Netter note it was the biggest one-,two-, or three-day decline since the unexpected victory of Germany over France in WW

II Here we examine the 20 years since and find that the market decline the week beforethe crash was indeed an unusually large decline in market history

While the 20 years since the crash have seen the volume of trading on the NYSEincrease dramatically, and while there have been episodes or days with large marketmovements, the October 14-16, 1987 period still ranks high amongst periods of severemarket decline While an argument can be made that the infrastructure of the markethas improved over the last 20 years so that the market is better able to cope withepisodes of high volume and illiquidity, it is also fair to say that the three-dayspreceding Monday October 19, 1987 remain an extraordinary period of decline in thehistory of the stock market

Finally, we suggest that the 1987 October Crash was caused by fundamentallydifferent dynamics than the fall 2008 market decline In the October 1987 crash, wereview the argument that fundamental news moved the market down over 10 percent.This significant market decline changed traders’ view of the viability of dynamictrading strategies, which affected market operations, causing a downward revaluation

of stock prices and leading to a ‘‘crash’’ on October 19 In fall 2008, there was asignificant stock market drop related to fundamental factors of the seizing up of thecredit markets, major declines in the price of housing and resulting foreclosures,declines in the value of CMOs, and the extent of bank leverage, bank failures, bailouts,recession and overleveraging, but unlike 1987 there was little or no informationrevealed about the trading strategies of stock market participants Thus, we shouldperhaps take little solace that within a year the market had recovered the value lost inthe crash of 1987 and the crash was not followed by a recession Things may be verydifferent now

Notes

1 There are minor discrepancies in returns data reported for the S&P 500, depending onsource and data series used For example, the center for research in security prices(CRSP) value-weighted return including distributions series shows the returns from14th-16th October as 2.79, 52 and 5.15 percent, respectively By contrast, theReturn on the S&P 500 Index (NYSE/AMEX only) series on CRSP reports returns of

2.95, 2.34 and 5.16 percent, respectively In the paper our numbers correspond tothe latter series Note that in either case the three-day cumulative return is 10.12percent, which represents a minor difference to the10.44 percent reported in Mitchelland Netter (1989) based on the data available at that time

2 See Lowenstein (2000) for a colorful account of this episode in the market, and its effect

on the hedge fund Long Term Capital Management (chapter 7 in particular)

What caused the

1987 stock market crash

Gerety, M and Harold Mulherin, J (1991), ‘‘Patterns in intraday stock market volatility, past and

present’’, Financial Analysts Journal, September-October, pp 71-9

Grossman, S (1988), ‘‘Analysis of the implications for stock and futures price volatility of

program trading and dynamic hedging strategies’’, Journal of Business, Vol 61, pp 275-98

Haugen, R Talmor, E and Torous, W (1991), ‘‘The effect of volatility changes on the level of stock

prices and subsequent expected returns’’, Journal of Finance, Vol 46, pp 985-1007

Jacklin, C., Kleidon, A and Pfleiderer, P (1992), ‘‘Underestimation of portfolio insurance and the

crash of October 1987’’, Review of Financial Studies, Vol 5, pp 35-63

Labaton, S and Andrews, E.L (2008), ‘‘In rescue to stabilize lending, US takes over mortgage

finance titans’’, New York Times, September 7, available at: www.nytimes.com/2008/09/08/

business/08fannie.html

Lem, G (1987), ‘‘Changes in markets add to risk, NYSE boss warns’’, The Globe and Mail,

September 28

Lowenstein, R (2000), When Genius Failed: The Rise and Fall of Long-term Capital Management,

Random House Trade Paperbacks, New York, NY

Miller, M and Mitchell, M (1999), ‘‘The stock market crash of 1987: what was all the fuss about?’’,

unpublished paper, University of Chicago GSB, Chicago, IL

Mitchell, M and Netter, J (1989), ‘‘Triggering the 1987 stock market crash anti-takeover

provisions in the proposed house ways and means tax bill?’’, Journal of Financial

Economics, Vol 24, pp 37-68

Mitchell, M., Pedersen, L and Pulvino, T (2007), ‘‘Slow moving capital’’, American Economic

Review, Vol 97, pp 215-20

Presidential Task Force on Market Mechanisms (Brady Commission) (1988), Report, US

Government Printing Office, Washington, DC

Roll, R (1989), ‘‘The international crash of October 1987’’, in Kamphuis, R.W Jr., Kormendi, R.C

and Watson, J.W.H (Eds), Black Monday and the Future of Financial Markets,

Mid-America Institute for Public Policy Research, Irwin, Homewood, IL

Wikipedia (1987), ‘‘Black Monday’’, available at: http://en.wikipedia.org/wiki/Black_Monday_(1987)

About the authors

Ryan McKeon received his PhD from the University of Georgia in December of 2008 He will be

an Assistant Professor of Finance at the University of San Diego in the fall of 2009 His research

is in asset pricing

Jeffry Netter is the C Herman and Mary Virginia Terry Chair of Business Administration and

a Josiah Meigs Professor at the University of Georgia He has a PhD from The Ohio State

University, J.D from Emory University, and a BA from Northwestern University He has taught

at the University of North Carolina and the University of Michigan His research concentrates on

the interactions of law, economics, and finance in areas such as corporate control, corporate

governance, and politics and governance He is the Managing Editor of the Journal of Corporate

Finance Jeffry Netter is the corresponding author and can be contacted at: jnetter@terry.uga.edu

To purchase reprints of this article please e-mail: reprints@emeraldinsight.com

Or visit our web site for further details: www.emeraldinsight.com/reprints

Department of Finance, College of Business, Florida State University,

Tallahassee, Florida, USA, and

Design/methodology/approach – The paper compares the number of IPOs, as well as accounting data during the years surrounding the 1987 crash to determine if there is a change in financial quality The underwriting fee structure, underpricing and short term price changes during one year prior to and one year following the 1987 crash are examined, as well as the long term returns surrounding the crash.

Findings – The stock market crash of 1987 did change the market psyche in the short to medium term Results show greater risk aversion in the post crash period, as evidenced by fewer IPOs from riskier firms Pricing is found to be more rational – less one day run-up, less upward adjustment from offering range, and less likely to be overpriced in intermediate and longer terms.

Originality/value – The paper demonstrates the importance of market sentiment and may illuminate the causes of market cycles.

Keywords Flotation, Stock markets, Stock returns, United States of America Paper type Research paper

Introduction

On October 19, 1987 the Dow Jones Industrial Average declined 508.32 points (22.6percent) which equates to a loss of $500 billion This paper is an examination ofwhether the 1987 stock market crash affected the psyche of the investing public Inparticular, we investigate whether the crash caused a structural change in the market

or a shorter-term behavioral change The latter may include a drastic but short-termshift in risk aversion by investors resulting in a flight to quality type of phenomenon.Although flight to quality could be tested in principle with stocks that are traded, theprice run up prior to the crash may make the notion of quality difficult to define andmeasure A more direct approach is to examine the initial public offering (IPO) market,where the test for flight to quality is simplified to whether only better quality (larger,more profitable, longer history) unlisted firms offer IPO We use characteristics of IPOs

to see if there are changes in the quality of IPO issuers due to the stock market crash of

1987 This paper seeks to explain how the market crash affected the stock market withrespect to IPOs, both immediately and in the longer term IPOs are interesting take onthe market psyche in that it involves both the response of supply from issuers anddemand by the investors The event of the crash if resulted in a flight to quality couldmanifest in the withdrawal from the market of certain potential issuers (lower quality)

as well as investor types (less risk averse/more optimistic)

The current issue and full text archive of this journal is available atwww.emeraldinsight.com/1475-7702.htm

The psyche of the stock market

139

Data and theory

If the market crash caused the psyche of the market to change in the short term,

irrational behavior may be observed In particular, having observed the occurrence of a

relatively rare event of over a 20 percent decline in a single day, the market would

assign much higher probability for a similar decline to occur in the near future, a result

of a small sample or representativeness bias In addition, the exit of the most optimistic

investors would further reinforce a greater market discount for risks Since the riskiest

stocks tend to decline more than the average, these effects would be felt mostly on high

risk stocks, and investors in these stocks were burned and would avoid them

As a result of these factors, there should be a flight to quality in the following

months and years This ‘‘flight to quality’’ hypothesis (Bernanke et al., 1996) states that

adverse shocks to the economy may be accelerated by worsening market conditions

Essentially, the financial accelerator implies that borrowers with severe agency

problems have reduced access to credit during economic downturns When applied to

IPOs, the ‘‘flight to quality’’ has the following predictions:

. Only high quality firms issue IPOs This means larger, more profitable, less

levered firms, less need for certification from third parties such as venture

capital firms, and from more established industries[1]

. The pricing of IPOs could be lower than those issued prior to the crash due to the

withdrawal of the less risk averse investors; the discount is greater for riskier

shares

. The aggregate dollar and number of IPOs will decline, as lower quality issues are

being shut out of the market However, the dollar value per IPO will rise

reflecting larger, less risky firms dominating the issue market

. Investment banks would have harder time marketing new IPOs, as reflected in

their greater selling expense

. The greater aversion to risk would also reduce the demand for IPOs from new

industries, a negative externality in reduced external funding for new ventures

The null hypothesis is that the market was not affected by the crash, and there would

be no difference in the pre to post crash comparison Underlying the hypothesis are the

assumptions that the market did not underestimate the probability of a crash earlier,

nor did it overestimate its probability afterward They are simply regarded as what

they are, rare events Investor sentiment is quite a relevant issue, as Siegel (1992) notes

that shifts in investor sentiment, perhaps induced by noise traders, were a factor in the

1987 stock decline Our study examines investor sentiment and market psyche through

the IPO market Similarly, Seyhut (1990) found evidence that suggests overreaction

was an important part of the1987 crash In a related issue, Shiller (1989) surveyed

investor behavior around crashes and argued there were no changes in economic

fundamentals, and that investors merely trade based on price changes

Results

Number of IPOs

The number of IPOs gives a sense of the IPO market momentum Table I shows the

aggregate number of IPO’s in terms of issues and principal amount, both the sum and

average size of the given year We can see that in the year before the crash (1986) that

the number of IPOs peaks at 708 for the eleven year period surrounding the 1987 crash

The year of 1987 had a smaller number of IPOs at 531, but this is still larger than the

8,2

140

preceding years of 1985 with 322 and 1984 with 343 IPOs The relatively high number

of offerings in 1987 reflects the fact that there was a robust market for 9-10 monthsprior to the crash Clearly, the years of 1986 and 1987 had significantly larger quantity

of IPOs After the 1987 crash, the quantity of IPOs did not reach the 1987 level until fiveyears later in 1992 when the number of the IPOs hit 561 In the three years followingthe 1987 crash, the number of IPOs did not exceed 300 per year Clearly, the 1987 crashhad an effect upon the quantity of IPO for several following years Lowry (2002) findsthat investor sentiment is a determinant of aggregate IPO volume The quantity ofIPOs had risen prior to the market crash, reflecting the sentiment driving the market tonew heights and overvaluation, which also drove the IPO market The year 1987 is ofparticular interest, as the table shows the momentum of IPOs were much in line withthe general stock market overheating at the time Although we feel investor sentimentinfluenced the IPO market following the 1987 crash, managerial timing to avoid thedepressed stock market may also be a plausible explanation

In terms of the principal amount, Table I shows both the sum and average principalamount for each of the eleven years surrounding the 1987 crash The principal amountnumbers are not inflation adjusted The years 1986 and 1987 were peak years with thesum in 1986 at $22 billion and 1987 at $24 billion The numbers in 1987 are the sum oftwo regimes – the uptrend in principal continued up to the October crash, and declinesafterward The decline in the number of IPOs immediately after the crash may be aresult of investment bankers concern over their reputation Dunbar (2000) studied howthe withdrawal of IPO from October 1987 to December 1987 affected market share ofinvestment banks later, their ability to complete IPOs did impress potential issuers, sothis factor may be more important than cost of issuance In the year 1988, although wesee a similar aggregate level at $22 billion, the issue market was dominated by largeissues, with average issue size almost twice that of previous years This supports thetendency of investors’ flight to quality immediately after the crash But in the yearsfollowing a decline to $13 billion in 1989 and $11 billion in 1990 However, the marketrecuperates by 1991 as the sum resumes to the 1987 level of $24 million The experience

of the crash appears to have left the market with a sour taste, or bad memory for at

Table I.

Aggregate IPOs in

number of issues and

the sum and average

principal amount

Year

Principal amount Number of IPOs

Sum of all markets ($ mil)

Average of all mkts ($ mil)

The psyche of the stock market

141

least 2-4 years, before a gradual recovery due to penned up supply by firms It is

possible given the momentum of the upward trend from 1985 to 1986, the number of

IPOs in 1987, 1988 would have been higher

Size

We observe another interesting issue is in terms of the average size of the IPO It

appears as though after the 1987 crash, the average size of IPOs increased In 1988, the

average size increased to 83 million, up from 45 million in 1987 Furthermore, within

the year of 1987, before the crash the average IPO size was $44.7 million, whereas after

the crash, the average IPO size increased to $59.1 million This is evidence that the

market would only accept IPOs of larger, more established companies This is

consistent with the prediction of the hypothesis We can also see that in the years

following the 1987 crash, the average IPO size stays above the 1987 levels In 1989, the

average IPO size was $56 million and in 1990, the average IPO size was $53 million

The decline in average issue size could be consistent to a reduction of investors’ flight

to quality, allowing smaller issues to be offered

Volume

We examine the quantity and volume of IPOs per month in 1987, as well as the monthly

figures 5 years prior to the crash to see if there is a decline in IPOs in months prior to

10/19, to enable us to determine whether the IPO market anticipated the market

decline Furthermore, we seasonally adjust the data by dividing the IPOs number and

volume in each month by its previous five years of monthly average as shown in

Table II The seasonally adjusted figure represents the IPOs number and volume in

each month divided by its previous five years average of the same month In the year

1987, the seasonally adjusted monthly number of IPOs is 1.31 in January, 2.44 in

February, 1.88 in March and 1.89 in April In May, June, July and August, the numbers

drop slightly to 1.46, 1.70, 1.47 and 1.34 In September, the number increases to 1.88, but

drops to 0.70 in October and declines greatly in November and December to 0.16 and

0.17

The same trend can be seen in terms of volume In the year 1987, the seasonally

adjusted monthly sum of volume of IPOs is 6.40 in January, 4.84 in February, 6.57 in

March and 2.44 in April In May, June, July and August, the numbers drop slightly to

3.05, 3.50, 3.57 and 1.96 In September, the number increases to 4.49, but drops to 1.21 in

October and declines greatly in November and December to 0.26 and 0.70

We split the sample into three-time periods: the pre 1987 crash time period from

1982 to 19 October, 1987, the post crash from 20 October, 1987 to 1989 and the post

crash 1990-1992 This is to separate the data into the pre crash period, the near term

after crash period, and longer term after crash period

Financial data

We examine the accounting data of IPOs during the years surrounding the 1987 crash

to determine if there is a change in quality from a financial standpoint If we observe a

change in the quality of financial indicators after the 1987 crash, it is evidence that the

market would only absorb more financially stable, higher quality IPOs To differentiate

whether the change in market’s attitude is short term or structural, we examine

characteristics of issues (as indicators of quality) for the immediate two years post

crash period (October 1987 to December 1989), and the period after (1990-1992) (see

Tables III and IV) In general, we find more homogenous and higher quality in the first

In terms of revenue before the 1987 crash, the average revenue was $197 million,

$182 million two years after the crash, and $122 million during the three to five yearsafter the crash In particular, we examined 1987 and noted that the average revenuewas $119 million while the median was only $23 million Thus, the year was dominated

by a few large revenue issuers The median revenue, however, before the 1987 crash are

$18.6 million, and increased to $34.7 million two years after the crash, and $26.95million during the three to five years after the crash The standard deviation of revenuewas $276 million before the crash, $327 million two years after the crash, and $469million during the three to five years after the crash

A similar pattern can be seen with respect to net income with both the mean andmedian, as one observes a larger dollar amount of net income after the 1987 crash Themean net income prior to the 1987 crash is $3.38 million, $5.13 million post crash to

1989, and $4.38 million three to five years after the crash The same trend can be seen inthe median with IPOs of less profitable firms with median net income of $0.90 million

Table II.

Number and principal

amount per month for

The psyche of the stock market

143

Table III.Financial data per firm

The psyche of the stock market

145

occurring prior to the crash, whereas after the crash the market required firms to be

more profitable with median net income of $1.4 million during the two years after the

crash, with a slight decline of $1.35 median net income three to five years after the

crash This is consistent with the hypothesis of the market fleeing to larger, more

profitable firms after the crash The standard deviation in net income is also larger

prior to the 1987 crash, at $21.25 million, $16.56 post crash to 89, and $19.18 million

three to five years after the crash

The amount of leverage that companies are willing to take on, or that the market

may be willing to bear with respect to an IPO may have changed after the 1987 crash

We examine both the dollar amount of debt as well as the percent of debt to equity In

terms of the dollar amount of debt, from 1984, there is a steady increase in the debt

leading up to 1987 There had been a steady increase in debt/ leverage too, by issuers

prior to the crash, however, the trend was reversed in 1988 It is interesting to note at in

1987, before the crash the average debt was $63 million, but in the months after the

crash, the average level of debt was only $23 million, indicating that the market was

not willing to accept an IPO from a risky company with a lot of debt

In terms of averages prior to the 1987 crash long term debt is $85 million, yet in the

two years following the crash, long term debt declines to $55 million, yet creeps

upwards to $70 million three to five years after the crash indicating that the market

impact is temporary

A similar pattern can be seen for the debt to equity ratio starting in 1984, with a

visible increase in the debt leading up to 1987 In 1984, debt to equity ratio for

companies conducting an IPO was 45 percent, increasing to 69 percent in 1985,

78 percent in 1986 and 95 percent in 1987 Once again, we see a retrenchment in terms

of the market’s ability to digest a company with more debt after the crash During 1987

before the crash, the debt to equity ratio was 97 percent, after the crash it was 47

percent indicating that companies conducting an IPO after the crash were much less

leveraged and thus less risky

These are pre IPO leverage ratio which reflects, consistent with the flight to quality

hypothesis, less willingness for the market to bear risks, or demand greater quality, as

the equity base is a function of past profits The mean and median debt/equity ratio

also show the same trend of more debt/equity prior to the 1987 crash, with a decline in

debt/equity after the crash, with the numbers increasing in the 3-5 year period after the

crash According to Wolfson (1996), the 1987 stock market crash demonstrates the

modern debt-deflation process encompasses falling asset prices and debt repayment

difficulties

In terms of the return on equity, we see a trend of less profitable firms able to

conduct an IPO prior to the crash with an average ROE of 15.51 percent prior to the

crash, yet after the 1987 crash, the average ROE increases to 17.17 percent indicating

that the market was risk adverse, requiring higher profitability after the crash The

median numbers show a similar trend with ROE of 11.78 percent prior to the crash,

12.14 percent in the two years after the crash and declining to 11.39 percent three to five

years after the crash

This is consistent with the hypothesis of the market fleeing to more profitable firms

after the crash The most divergence we see is during 1987 prior to the crash, the return

to equity ratio is 16.88 percent, whereas after the 1987 crash, the ratio doubles to 33.27

percent This indicates that in the short term, the market required substantially higher

profitability to accept an IPO In addition, return on equity is an after leverage number,

thus highly levered firm’s high ROE is due to use of debt at interest cost lower than

a related issue, Lauterbach and Ben-Zion (1993) study the effects of the 1987 crash onthe Tel-Aviv Stock Exchange and find that the crash and its aftershocks lasted for aweek and selling pressure was concentrated in higher beta, larger capitalization andlower leverage firm stocks.

During the 1982-1992 period the number of employees increases on average, from

140 employees in 1982 to 562 in 1983, dipping in 1984 to 182, increasing to 765 in 1986.The average number of employees at a company conducting an IPO in 1987 is 505 and

we observe a decline in the following two years to 475 in 1988 and 293 in 1989

The number increases substantially to 1109 in 1990 and 1218 in 1991 During theyear of 1987 before the crash, the number of employees is 527, but for firms conducting

an IPO after the crash, the number declines to 214 Thus, there is a post crash decline inaverage number of employees per firm and yet an increase in revenue per share Sowhat kind of industry dominates these firms? They could not have been service, orretail industries with many employees, nor very high tech with few employees, as weknow there would be few new industries, or risky firms So, they are established andcapital intensive firms The predominant industry having an IPO after the 1987 crashuntil the end of 1988 was manufacturing, with 104 of the 214 IPOs being inmanufacturing (49 percent) In contrast, from 1986 to before the 1987 crash, only 360 ofthe 1216 IPOs being in manufacturing (30 percent)

As is often the case in behavioral finance, it should be noted that there is no specificrationale for a change in financial fundamentals given the rather swift marketadjustment in 1987 However, Shiller’s (1989) survey research indicated that investorstraded primarily on price movements and emotions, rather than trading on economicfundamentals

Venture capital and leveraged buyoutsDuring the period prior to the crash from 1982 to 1987, there is a decline in the percent

of IPOs which are venture capital backed The decline in venture backing prior to thecrash reflect IPOs of firms that are of higher risks, lacking certification of venturecapital funds In 1982, 74 percent are venture backed, 73 percent are venture backed in

1983 This level declines to 60 percent in 1984 and further declines to 39 percent in

1985 It stays at a similar level in 1986 at 40 percent, but further declines to 31 percent

in 1987, 23 percent in 1988 and 22 percent in 1989 The decline after the crash is theresult of larger, more established firms that need not have venture backing due to theirlong history in business Within the year of 1987, prior to the crash the venture capitalbacking level was 32 percent and post crash it was 11 percent The level begins toincrease steadily in 1990 at 29 percent, 39 percent in 1991 and 46 percent in 1992 as aresult of quality firms which have gone through the due diligence process and weredeemed worthy of funding by venture capitalists and are seeking to conduct IPOs

In the post 1987 crash years, the data indicates that those firms that receivedventure capital backing were smaller firms in terms of principal in both the near term(1987 crash – 1989) and further out in time (1990-1992) For firms not receiving venturecapital conducting IPOs during the time period of the 1987 crash – 1989, the averagesize in terms of principal was $187 million, whereas those firms that received venturecapital backing were much smaller, with an average size of $56 million (For furtherdiscussion on IPOs as a means of exit for venture capital, see Gilson and Black, 1999.)

The psyche of the stock market

147

The ability of buyout groups (managers and buyout funds) to turn a quick profit

depends on the level of equity in that high stock prices reduce the debt/equity ratio in

addition to debt reduction from paying off debt, and high stock prices enable using

IPOs to cash in The percentage of IPOs having conducted a previous leveraged

buyouts are relatively low during 1982-1984, ranging from 0 percent to 1.30 percent and

0.5 percent In 1985 it increases to 4.65 percent, in 1986 it increases to 3.95 percent and

5.46 percent in 1987 It declines somewhat in 1988 and 1989 to 2.24 percent and 2.91

percent, respectfully After the 1987 crash, with less favorable stock prices, there are

fewer IPOs as lower stock prices means lower IRR to the buyout group, which would

make their payback period longer Although it is not possible to measure privately

held bought out companies, if the market crash is affecting operating business of some

firms adversely, it may cause business failures, explaining part of the smaller pool of

firms to offer IPOs Kosedag and Michayluk (2004) also document that the majority of

the repeated leveraged buyouts (LBOs) were performed following the 1987 stock

market crash

The level of previous leveraged buyout companies increasing to 7.72 percent in

1990, 11.80 percent in 1991 and 15.50 percent in 1992 In these later years following the

crash, the percent increase is due to penned up supply, for these buyout groups to cash

out and reduce the already long payback period

Underwriting and pricing

Differences in pricing and underwriting during the ten years surrounding the 1987

crash can be seen in Table V The offer price is reasonably constant over the entire time

period ranging from a low of $8.91 in 1984 to a high of $12.83 in 1992, within the usual

price range During 1987 up until the crash, the average price was $10.75, but fell to

$8.33 after the crash indicating that stock market prices as a whole fell and that the

market was not willing to pay high prices for IPOs, however, the offer price is based on

the number of shares offered In 1988, the average price was at a similar level to 1987 at

Table V.Pricing

Issue priced relative to filing range

Notes: *Statistically significant at the 0.05 level; **statistically significant at the 0.01 level

Source: Data are from Thomson Financial Securities Data Corporation

We examine whether an issue is priced above, within or below its filing rangeduring the five years prior to 1987, as well as the five years after the 1987 crash During

1987 prior to the crash 4.76 percent were price above the filing range, 77 percent werepriced within the filing range, and 17 percent were priced below the filing range Weobserve the high ratio of above versus below the filing range and the trend Due to theoriginal range being set with the prevailing market condition in mind whether hot orcold market for IPOs, the adjustments prior to offer reflect short term demand orspeculative fervor for the shares It could also reflect perceived under or over pricing.Intentional underpricing after the crash may be necessary, thus, to explain theobservations

After the crash in 1987, there were no IPOs priced above the filing range, 95 percentwere priced within the filing range and 4.55 percent were priced below range reflectinggreater caution and more effort put into pricing to gauge the market, again consistentwith a perception in a change in the market’s demand for IPOs In the three years prior

to the 1987 crash, the percent priced below the filing range declines each year from 29percent in 1984, to 16 percent below in 1985, to 14 percent in 1986, indicating thatmarkets were growing hot In the years following the 1987 crash, you tend to see moreinvestment banks pricing the issue within the filing range, erring on the side ofconservatism In 1988, the percent priced within range is 82.88 percent up from 78.37percent in 1987 The same conservative pricing trend continues in 1989 at 89.27 percentpriced within range, 1990 with 82.50 percent priced within range, and 1991 with 81.61percent priced within range This contrasts with the number of IPOs priced withinrange in the years leading up to the 1987 crash In the years 1982-1984, 60.18, 66.77 and69.01 percent, respectively were priced within range In hot markets, lesser firms maytry to aggressively price their IPOs, which may explain the higher numbers of IPOsbelow the original filing range in hot markets

Underwriting fees

We also examine the fee structure of IPO to get a sense of change in the investmentbanks’ effort needed to market IPOs due to short term or structural change in themarket (Table VI) The fees we look at are the gross spread, the reallowance fee andunderwriting fee The gross spread is very constant over the time period ranging from

a high of 8.72 percent in 1982 to a low of 7.24 percent in 1992, with the midpoint in time

of 1987 at 8.04 percent This near constant fee centering around 7 percent is welldocumented (Chen and Ritter, 2000)

The slight increase in the gross spread percentage seen in the few months after the

1987 crash of 8.47 percent is clearly a function of lower prices and greater selling effortneeded to market the issues The reallowance fee and underwriting fees are alsoextremely resilient to the market crash of 1987 This may in part be due to the numbersbeing a function of investment bankers rather than market sentiment Reallowanceexpense is that portion of fee the underwriting group decided to share with the sellingbrokers – it reflects the need to give brokers more incentives to sell – a result of lessperceived demand and greater need to market The large increase in reallowance toselling brokers, 36 basis points (2.00-1.64 percent), account for most of the increase, 45

The psyche of the stock market

149

basis points (8.47-8.02 percent), in fees charged by the lead banks between the pre and

post crash period

Another way to measure market receptiveness to new issues is to look at the

overallotment sold numbers Overallotment is a function of both underpricing and

temporary excess demand for the shares and is also a means underwriters receive

compensations In 1982, the amount of overallotment sold as a percent of the total

principal amount was a relatively low 5.39 percent, in 1983 it was similarly 6.88 percent

and 5.80 percent in 1984 During the period 1985-1987, the percent of overallotment

sold increased to 10.30 percent in 1985, 11.68 percent in 1986 and 9.48 percent in 1987

Within 1987 prior to the crash, the overallotment sold was 9.58 percent, but dropped to

6.97 percent after the crash In the years following, the number declines to 8.66 percent

in 1988, 7.01 percent in 1989 and 6.22 percent in 1990 evidencing that the market’s

excess demand for IPOs was less for a few years following the crash

The potential perception of higher gross spread in the pre period is mainly due to

the period’s higher percentage of smaller IPOs There is a possibility that structure of

the market might have changed, such as a reduction in transaction costs that favor

larger issues, but our data indicates otherwise We find between the period 1987 crash –

1989 and 1990-1992 a doubling of the issue size from $13.5 mm to $27 mm This reduces

gross spread by 64 basis points for small issues and by 23 basis points for larger

issues

New industries versus established industries

We also examine the number of IPOs in new industries versus IPOs in established

industries five years prior to the crash and five years following the crash New

industries are defined as those firms conducting an IPO with a new SIC code within

five years of the IPO Established industries are those firms conducting an IPO with an

SIC code that is over five years old Our rationale for looking at the quantity of IPOs in

Table VI.Underwriting fees

Year

Gross spread

as % price

Reallowance fee as %

of principal

Underwriting Fee as %

of principal

Overallot amt sold as %

Notes: *Statistically significant at the 0.05 level; **statistically significant at the 0.01 level

Source: Data are from Thomson Financial Securities Data Corporation

The sample of all IPOs was divided into those in new industries and those inestablished industries The process for dividing the firms is as follows Industryclassification codes for the IPOs were obtained from Thomson Financial SecuritiesData Corporation SIC codes were obtained going back to 1926 A new industry isformed when the first industry classification appears An IPO is categorized asoccurring in a new industry for the first five years after the first IPO appears in theindustry An IPO is categorized as occurring in an established industry five years afterthe first IPO appears in the industry.

Looking at Table VII, one can see that the percent of IPOs in new industries versusestablished industries in the five years prior to the crash is 5.68 percent, whereas afterthe crash the percent declines to 1.48 percent New industries present a specialcomplication for the investing public in that less is known about a new industry interms of profitability, strategy and risk The reduction in the supply of IPO funds after

1987, especially for new industries shows the market could affect the supply of funds tofirms for investments (Morck et al., 1990) Whereas with an established industry moreinformation is available and better understood given the longer history In terms ofnumbers, the number of IPOs in new industries during the five years prior to the crashwas 149, compared to 26 IPOs during the five years after the crash For establishedindustries, the number of IPOs during the five years prior to the crash was 2625,compared to 1760 during the five years after the crash Thus, given that fewer IPOs innew industries chose to go public after the 1987 crash leads one to believe that becausethe market was less willing to bear risk, the ability of new industries, which areinherently more uncertain, suffered Thus, we are able to document a case of negativeexternality of market crashes

Underpricing and short term stock returnsAny change in the market psyche will likely alter the level of issue underpricing.Hypothetically, if the investors’ psyche has changed and they demand greatercompensation for risk, underpricing after the 1987 crash will be higher, to compensateinvestors for their higher perception of risks One could call this hypothesis the

‘‘increase risk aversion hypothesis’’ The alternate hypothesis is that due to a change inmarket psyche, instead of demanding greater compensation for risks, the market willonly have demand for quality IPOs, and no demand for low quality IPOs Anotheraspect which may occur is that the low quality IPO are priced at such a steep discount,that the potential issuing firms would withdraw As a result, only quality issues wereoffered and there is an increase in quality but no increase in underpricing One couldcall this the ‘‘flight to quality hypothesis’’ Note this flight to quality hypothesis is alsosupported by our previous results with accounting numbers, volume, amount per

Table VII.

Number of new

industries offering IPOs

in each year around this

period

IPOs in new industries

IPOs in established industries

Percent of IPOs in new/established industries

The psyche of the stock market

151

issue, etc Of course, we should also mention that there is the null hypothesis that the

1987 crash had no effect

Numerous studies document underpricing of IPOs such as Stoll and Curley (1970),

Logue (1973), Reilly (1973) and Ibbotson (1975) Our study looks at the level of

underpricing surrounding the 1987 crash Table VIII illustrates the amount of

underpricing and the short term price changes during one year prior to and one year

following the 1987 crash, as well as the three-year period beyond each one year period

surrounding the crash In terms of underpricing, there is larger underpricing in both

latter periods following the 1987 crash when compared to the periods prior to the crash

For the –48 months to –13 months prior to the crash, the mean first day return is a mere

5.9 percent and similarly, for IPOs one year prior to the crash, the mean first day return

is 5.75 percent After the crash, the amount of underpricing increases to 6.5 percent for

the year after the crash and rises to 10.03 percent for the periodþ13 months to þ48

months after the 1987 crash The result that underpricing after the 1987 crash is higher

lends support for the ‘‘increased risk aversion hypothesis’’ It appears that the market

crash of 1987 did affect the psyche of the market In order to compensate investors for

their higher perception of risks, the market demanded greater compensation for this

risk It should be noted that the pricing of IPOs compared to its filing range (Table V) is

related to the IPO underpricing shown in Table VIII

In terms of short term price changes, the one month return declines steadily over the

periods prior to the crash and continues to decline further after the crash The one

month Nasdaq adjusted return for the period –48 to –13 months prior to the crash was

7.62 percent, and 5.97 percent for those IPOs occurring within one year prior to the

crash The one month returns decline further after the crash with the Nasdaq adjusted

return being 4.68 percent for those firms doing an IPO within one year of the crash and

the return being 3.14 for firms doing an IPO 13 months to 48 months after the crash

Table VIII.Underpricing and short

term returns

Time of

Mean first day return

One month return

Nasdaq adjusted

Six month return

Nasdaq adjusted

Notes: This table is a comparison of underpricing and short term returns looking at the time

period of three-year before last year, last year before, one year after and three years after that, i.e.

48 to 13, 12 to 1, þ1 to þ12, þ13 to þ48 months; Nasdaq adjusted return is the one

month return for the IPO subtracting off the corresponding Nasdaq return for the given time

period; *statistically significant at the 0.05 level; **statistically significant at the 0.01 level

Sources: Data are from The University of Chicago’s Center for Research in Securities Prices

(CRSP) as well as Thomson Financial Securities Data Corporation

is 10.03 percent and for firms doing an IPO 13 months to 48 months beyond the crash,the Nasdaq adjusted return is 8.45 percent The latter is an indication of underpricing,

or compensation for risk when discount for risk in the after crash period went up.Long term stock returns

In addition to looking at short term pricing effects, this paper seeks to explain how the

1987 crash affected the stock market with respect to IPOs in the longer term Therehave been several studies documenting the long-run underperformance of IPOs such

as Loughran and Ritter (1995), and Brav and Gompers (1997) We examine how thisphenomenon changes surrounding the 1987 crash by looking at 1 year, 3 year and 5year returns, both raw and Nasdaq market adjusted returns

There are statistically significant differences between the long term returns of IPOsprior to and after the crash The one year market adjusted return average for the fiveyears and ten months prior to the crash is –10.57 percent compared to –0.05 percentafter the crash This difference is statistically significant with a t-statistic of 3.35.Similarly, the three-year market adjusted return average prior to the crash is –34.32percent compared to –3.24 percent after the crash (see Table IX) This difference isstatistically significant with a t-statistic of 4.09 The five year market adjusted returnaverage prior to the crash is –22.64 percent compared to –2.36 percent after the crash.This difference is statistically significant with a t-statistic of 3.83 These long termreturns indicate that investors demand greater compensation for risk in the form of

Table IX.

Long term returns

Year

One year raw return

One year market adjusted

Three-year raw return

Three-year market adjusted

Five-year raw return

Five-year market adjusted

t-test for prior and

Notes: *Statistically significant at the 0.05 level; **statistically significant at the 0.01 level Sources: Data are from The University of Chicago’s Center for Research in Securities Prices (CRSP) as well as Thomson Financial Securities Data Corporation

The psyche of the stock market

153

higher returns In the process, they avoided the long term decline in IPOs due to over

pricing at IPOs, or market exuberance It is important to note that some of the period in

the before crash did not cover the year of the crash, and thus, the long term decline is

not due to the crash alone The results could also be explained by greater speculation

when the IPO is initially traded on the stock exchange during the pre-period IPO

market In the post-period, higher quality IPOs and less speculation in that early

trading period could lead to superior long-term returns This possibility is supported

by the one-month return returns of Table VIII showing lower one month returns after

the crash implying less speculation Finally, the result is also consistent with having

higher average quality issuers after the crash, and thus, avoiding unmet expectations

in previous years

Conclusion

We find evidence that the stock market crash of 1987 did change the market psyche, as

demonstrated by this detailed study into one market, the IPO market We show

evidence in support of a change in greater risk aversion in the post crash period as

evidenced by fewer IPOs from riskier firms (small issue size, more debt, lower revenue,

less profit, etc.) as they were being shut out of the market We find pricing to be more

rational – less one day run-up, less upward adjustment from offering range, and less

likely to be overpriced in intermediate and longer terms

However, we find the change in market psyche is short to medium term, in this case

extending from 1987 post-crash to 1990 After these 3-4 years, the market recovers, and

once again exhibits more optimisms with the return of less risk averse investors, as

seven years later, in the mid 1990 till 2000 when the tech market bubble burst again

Thus, the effect of the crash was felt, but could not be said with lasting memory

However, it does demonstrate the role of market sentiment, and how it changes

determines stock market prices, and provide some explanation of the market cycles –

alternating periods of exuberance and cautious aversion

Note

1 An alternative explanation is that investors myopic loss aversion may also explain why

investors remember what happened recently but slowly forget what happened in the

past For further reading, (see Bernartzi and Thaler, 1995)

References

Bernanke, B., Gertler, M and Gilchrist, S (1996), ‘‘The financial accelerator and the flight to

quality’’, Review of Economics and Statistics, Vol 78 No 1, pp 1-15

Bernartzi, S and Thaler, R (1995), ‘‘Myopic loss aversion and the equity premium puzzle’’,

Quarterly Journal of Economics, Vol 110 No 1, pp 73-92

Brav, A and Gompers, P (1997), ‘‘Myth or reality? the long-run underperformance of initial

public offerings; evidence from venture and nonventure capital-backed companies’’,

Journal of Finance, Vol 52 No 5, pp.1791-1803

Chen, H.C and Ritter, J (2000), ‘‘The seven percent solution’’, Journal of Finance, Vol 55 No 3,

pp.1105-31

Dunbar, C (2000), ‘‘Factors affecting investment bank initial public offering market share’’,

Journal of Financial Economics, Vol 55 No 1, pp 3-41

Grossman, S (1990), ‘‘Introduction to NBER Symposium on the October 1987 Crash’’, Review of

Financial Studies, Vol 3 No 1, pp 1-4

Shiller, R (1989), Market Volatility, MIT Press, Cambridge, MA.

Siegel, J (1992), ‘‘Equity risk premia, corporate profit forecasts, and investor sentiment aroundthe stock crash of October 1987’’, Journal of Business, Vol 65 No 4, pp 557-71

Stoll, H and Curley, A (1970), ‘‘Small business and the new issues market for equities’’, Journal ofFinancial and Quantitative Analysis, Vol 5 pp 309-22

Wolfson, M (1996), ‘‘Irving Fisher’s debt-inflation theory: its relevance to current conditions’’,Cambridge Journal of Economics, Vol 20 No 3, p 315-21

Further readingSeyhun, H (1990), ‘‘Overreaction or fundamentals: some lessons from insiders’ response to themarket crash of 1987’’, The Journal of Finance, Vol 45 No 5, pp 1363-89

About the authorsJames Ang is the Bank of America Eminent Scholar, Professor of Finance at Florida StateUniversity in Tallahassee, Florida He received his Ph.D from Purdue University Some of hispublications include the Journal of Financial Economics, Journal of Finance and Journal ofFinancial and Quantitative Analysis He has also served as Editor of Financial Management from1987-1993 and has served as President of Financial Management Association International from1996-1997 James Ang is the corresponding author and can be contacted at: jang@cob.fsu.eduCarol Boyer is an Assistant Professor of Finance at Long Island University – CW Post campus

in Brookville, NY She received her Ph.D from Florida State University Some of her publicationsinclude The Cambridge Journal of Economics, Advances in Investment Analysis and PortfolioManagement and Managerial Finance She has presented research papers at the European FMA,EFA, Decision Sciences Institute, FMA and the ABN AMRO International IPO Conference

To purchase reprints of this article please e-mail: reprints@emeraldinsight.com

Or visit our web site for further details: www.emeraldinsight.com/reprints

Capital market developments

Capital market developments in

the post-October 1987 period:

Purpose – The purpose of this paper is to review the evolution of the Canadian financial

environment since the stock market ‘‘crash’’ of 1987.

Design/methodology/approach – The paper provides a chronological account of significant

events in the Canadian economic environment and capital markets, and how they have transformed

the financial climate.

Findings – The late 1980s was a turbulent period with many changes in government and economic

policies which were initiated at a time when governments were wracked with fiscal deficits, and just as the

central bank appointed a dedicated inflation fighter These changes worked their way through the system

to contribute to one of the worst recessions in Canadian history One of the symbols of disparity during this

era was the Stock Market ‘‘Crash’’ of 1987, which was felt in Canada, as well as around the globe However,

for the last decade, the federal government has reported a surplus every year, and Canadians have

benefitted from falling tax rates, declining interest rates, a strong stock market, and a rising currency In

fact, until September of 2008, all of these developments had contributed to unprecedented profitability in

the financial services industry, until the recent widespread economic crisis in the US spread to Canadian

and global economies However, the Canadian economy seems much better poised to deal with such

adversity than it was in October 1987 If the fall of 2008 is any indication, we will find out soon enough.

Originality/value – The paper demonstrates how fallout from the crash of 1987, as well as other

subsequent developments, has contributed to significant changes in the financial environment.

Keywords Canada, Stock markets, Stock prices, Capital markets, National economy

Paper type Research paper

1 Background

Recent market turmoil including a 15 percent decline in the S&P/TSX Composite Index

for the month of September, 2008, a 16 percent decline during the week ending October

10, 2008 and a 40 percent decline since the highs of July have evoked memories of the

crash of October 1987 However, as far as Canada is concerned, it is a lifetime away in

terms of both economic and capital market conditions In this paper, we will first

discuss the political and economic conditions at the time of the crash since much of

what happens in the capital market stems from developments here We then discuss

capital market conditions as we conclude our discussion of what has changed since the

time of the 1987 crash

2 Political developments in the late 1980s

For Canadians, the time of the crash was the period of the Conservative Government of

Brian Mulroney which lasted from 1984 until 1993 Although the Conservatives were

then out of office until 2006, many of the major changes they made have never been

The current issue and full text archive of this journal is available atwww.emeraldinsight.com/1475-7702.htm

The authors thank the Social Sciences and Humanities Research Council of Canada (SSHRC) for

financial support provided for this project