reputation survey 30aug2010
Post on 14-Feb-2017
221 Views
Preview:
TRANSCRIPT
Reputation and the Invisible Hand: A Review of Empirical Research
For
The Oxford Academic Symposium on Reputation
September 15-17, 2010
and
The Handbook of Corporate Reputation
Jonathan M. Karpoff
First draft: 30 August 2010
This is a first draft. Please do not quote or cite without the author’s permission.
1
Reputation and the Invisible Hand: A Review of Empirical Research
“[Each person] generally, indeed, neither intends to promote the public interest, nor knows how much he is promoting it … he intends only his own security; and by directing [his] industry in such a manner as its produce may be of the greatest value, he intends only his own gain, and he is in this, as in many other cases, led by an invisible hand to promote an end which was no part of his intention. Nor is it always the worse for the society that it was not part of it. By pursuing his own interest he frequently promotes that of the society more effectually than when he really intends to promote it.” – Adam Smith, The Wealth of Nations (IV.ii.6-9, page 456 of the 1776 Glasgow Edition of Smith’s works; vol. IV, ch. 2, p. 477 of 1776 U. of Chicago Edition.)
I. The economic problem
Adam Smith’s “invisible hand” is perhaps the most important idea in the history of
economic thought. It captures one of the great discoveries of social science, that the pursuit of
self-interest can promote social welfare. But note that, in coining the term, Smith inserts a
qualifier. The pursuit of self-interest does not always promote society’s interest. Rather, it
frequently does so.
In the 234 years since The Wealth of Nations, economic philosophers have worked to
understand the conditions and processes by which the invisible hand works to promote social
good, and also to understand the conditions and processes in which it does not. So far, we have
three broad answers for why it might not: Self-interest can decrease the social pie when
monopoly is involved, when the costs of externalities are sufficiently high, or when market
outcomes are deemed undesirable for equity reasons.
This essay considers another category of economic activity in which the pursuit of self-
interest does not, in general, serve to promote social welfare. This is when economic actors lie,
2
cheat, and steal.1 It is widely acknowledged that opportunistic behavior impedes value creation
and that a culture of corruption tends to be inversely related to economic development.2
Economists also understand how trust is fundamental to exchange. In Akerlof’s (1970) lemons
problem, buyers demand discounts to compensate for their risk of being ripped off by sellers who
have an informational advantage about the quality of the good to be exchanged. Sellers who
stand ready to sell in the face of such discounts signal that their products are relatively inferior,
causing buyers to demand even larger discounts. In the end, trade breaks down, as buyers infer
that only sellers of the lowest quality products remain in the market at low, discounted prices that
buyers are willing to pay.
Akerlof’s problem is pervasive, in that information is costly and sellers typically know
more than buyers about the quality of what they are trying to sell. The problem goes the other
way, too, as buyers can opportunistically take advantage of sellers. In theory, one could get
around the lemons problem by writing contracts that protect opportunistic buyers and sellers
from each other. The problem, of course, is that contracts are costly to write and to enforce, and
therefore will always be incomplete.
Despite the pervasiveness of the lemons problem, buyers and sellers do get together –
millions of times each day. When you think about it, it is a bit of a miracle: millions of trades
occur every day in all parts of the world, despite the ever-present risk that sellers could cheat
buyers or buyers could cheat sellers. How does that happen? What holds the whole economic
process together? Why is fraud not running rampant in the streets? Fraud does occur, but why is
it the exception rather than the norm?
1 I do not take a strong position that this constitutes a fourth category of exceptions to Smith’s invisible hand theory. It is possible to frame misconduct and opportunistic behavior as a type of externality problem. The Coase Theorem implies that the costs of opportunistic behavior arise because the information and transaction costs to fix the problem are prohibitive. 2 A search for “corruption and economic development” on ssrn.com obtains 289 matches.
3
Of course, some people do not cheat others because their personal, moral, or religious
codes discourage cheating. Some types of opportunistic behavior are illegal, and the threat of
legal sanctions deters many people from behaving badly. Personal ethics and the law no doubt
are important. But even in an amoral society with no legal enforcement of contracts, there would
remain a powerful inducement to honest dealing: the prospect that your counterparty will do
business with you again in the future. By performing as promised, individuals and firms can
develop reputations for honest dealing. Such reputations are valuable because they yield
favorable terms of contract with customers, employees, suppliers, and investors.
Klein and Leffler (1981), Shapiro (1983), and others have developed models in which
reputation – and reputation alone – encourages good behavior and disciplines bad behavior. The
upshot is that people and businesses can invest in reputation, just as they might invest in
machinery, R&D, or human capital. Viewed this way, reputation is a valuable asset. It is the
present value of the improvement in net cash flow and lower cost of capital that arises when the
firm’s counterparties trust that the firm will uphold its explicit and implicit contracts, and will
not act opportunistically to their detriment.
Reputational capital is not transparent on a firm’s balance sheet, but circumstantial
evidence suggests that it is important. For example, Beatty, Bunsis, and Hand (1998) find that
investment banks with high reputation obtain higher fees for their services. And Resnick et al.
(2002) find that high-reputation sellers on eBay get higher prices than others, even for the same
goods. Nonetheless, measuring the size of a firm’s reputational capital is difficult, and we have
little direct evidence on its size for most firms. To get around the measurement problem,
researchers have used a different kind of experiment to infer the importance of reputation to
firms. This experiment examines the counterexamples – i.e., instances in which people or firms
4
lose reputation by lying, cheating, or stealing. To the extent that such losses are large, we can
infer whether, and where, reputation matters.
In this essay I survey the empirical research on the costs to firms from misconduct that
affects investors, customers, employees, suppliers, and the community at large. Section II
examines financial misconduct. The prototypical case is Enron, which during 1997-2002
inflated its reported earnings – and its stock price – by underreporting expenses, exaggerating
revenues, and lying about its financial leverage. The results from 32 studies show that firms lose
a significant amount of market capitalization when their financial misconduct is revealed,
although point estimates of the loss range widely.
Evidence of share price losses does not necessarily imply that firms suffer reputational
losses, because share prices can decline for other reasons as well. Section III reports on an
attempt to partition firms’ share value losses into components that reflect regulatory fines, class
action settlements, price deflation due to new information about the firm’s financial position, as
well as reputation. For financial misconduct, reputational loss explains the largest portion of
firms’ total losses. This inference is corroborated by evidence that the cost of equity and debt
financing both increase after investors learn that a firm’s books were in error.
Section IV surveys the empirical evidence regarding firms’ losses for many other types of
misconduct, including frauds of related parties, product recalls, air safety disasters, and
environmental violations. Here the evidence is mixed. Some types of misconduct, such as fraud,
are associated with significant share value and reputational losses. For other types, including
environmental violations and misconduct that affects unrelated parties, there is no systematic
evidence of reputational loss. Even when share values decline because of an environmental
5
violation, the decline reflects the legal penalties imposed on the violating firms, not any
reputational effect.
Section V proposes an explanation for these mixed results. Reputational capital – what
Klein and Leffler (1981) refer to as the quasi-rent stream from delivering the level of quality that
is promised – works to encourage honest dealing among counterparties. It is not a panacea that
discourages all types of misconduct. It has force when the firm’s actions undermine its trust with
its counterparties, but not when it harms unrelated parties. So firms that defraud consumers lose
future sales, and firms that cook their financial books face less favorable terms from investors
and creditors. But, while firms that pollute a river impose harms on downstream users, the
harmed parties typically are not the ones with whom the firm does business.
Thus, the role of reputation is more complex than a bland admonition to “follow the
Golden Rule,” or to “do well by doing right.” The evidence indicates that reputation, and the
threat of losing it, plays a more important role for some types of misconduct than others. As I
discuss in Section V, even firms in the same industry invest in different levels of reputational, so
the size of the reputational capital at stake differs even across otherwise similar firms. This
implies that vendors offer customers, suppliers, and investors different levels of reputational
guarantee, just like they offer different product, price, delivery, payment, and service. Where the
reputation at stake is larger, the likelihood that the firm will act opportunistically is smaller.
There remain many questions about how and in what settings reputation works to
facilitate Adam Smith’s invisible hand. Section VI concludes with a list of ten questions for
further empirical research.
6
II. Market value losses when financial misconduct is revealed
News that a firm’s financial reports are in error can affect the firm’s share values in three
ways. First, the news indicates that the financial information used by many investors to help
value the company is in error. Share values usually are inflated by the misleading financial
information, and they decrease as investors reassess the firm’s value in light of financial data that
provide a more accurate picture of the firm’s earnings. Second, the news suggests that the firm
may face legal penalties in the form of regulatory sanctions, fines, or shareholder lawsuits. And
third, the news can change the company’s real operations or cost of capital. The fact that the
firm’s officers furnished misleading financial information indicates that the company has poor
internal controls, managers who behave opportunistically, or both. As we will see, such
information causes the firm’s investors and other stakeholders to change the terms with which
they are willing to do business with the firm. Lenders, for example, change higher interest rates
to firms whose financial reports or internal controls are suspect.
In this survey I refer to these three effects as (i) share deflation, (ii) legal penalties, and
(iii) impaired operations and reputation. In theory, each of these effects serves to decrease share
values, so news of financial misconduct should be associated with a decrease in share values.
This is exactly what empirical research has documented. Table 1 summarizes the results of 30
studies that examine the share price reactions to news of error in a firm’s financial reports. The
table is organized by the five primary methods that researchers have used to identify financial
misconduct. Panel A reports on four studies that use key word searches from media reports to
identify instances of financial misconduct. Panel B summarizes the results of nine studies that
examine share price reactions for samples of firms that restated earnings. Panel C reports the
results of seven studies that examine the effects of private lawsuits, typically securities class
7
action lawsuits that allege that the firm violated SEC Rule 10b-5. Panel D reports the results of
four papers that use SEC Accounting and Auditing Enforcement Releases (AAERs) to identify
financial misconduct, and Panel E reports the results of two papers that use SEC enforcement
actions to identify such misconduct.
In a few cases I report weighted averages of a combination of smaller subsample results.
But in most cases I report the results as the authors report them. I focus on results from the event
window that the authors emphasize, or absent any such emphasis, the results from a window that
appears most comparable to the other papers in the table. My summary focuses on abnormal
stock returns, which are calculated as the difference between the raw return and a benchmark. In
many studies the benchmark is estimated from a one-factor market model, with estimation
periods that vary across studies. Several papers use alternate benchmarks, such as a simple
market return. A critical review of several individual papers’ empirical procedures could raise
doubts about their particular conclusions. But I believe that these papers’ overall conclusions –
and the inferences I draw from them – do not depend on the particular empirical methods
deployed.
II.A. Media key word searches
Early attempts to examine the share price impacts of business misconduct used data
collected from media sources, usually The Wall Street Journal. Several of these include
subsamples involving financial misreporting. Karpoff and Lott (1993) report an average two-day
abnormal return of -4.66% upon the initial press announcement for a small subsample of
financial reporting violations. Davidson, Worrell, and Lee (1994) report a three-day abnormal
stock return of -2.80% for a slightly larger sample of such violations. Gerety and Lehn (1997)
8
identify 37 instances of financial disclosure violations from SEC, and identify the first public
announcements of the SEC’s actions in The Wall Street Journal. The average three day
abnormal stock return is -3.05% and is marginally significant. To date, few studies investigate
the share price effects of business misconduct outside of North America. An exception is
Tanimura and Okamoto (2010), who examine corporate scandals in Japan, including 40 cases of
financial reporting frauds. Although the events are from a different country and time period than
the other events summarized in Panel A, the average abnormal stock return is similar, -5.99%.
II.B. Studies based on earnings restatements
Several researchers have examined how firm value is affected by news that a firm has to
restate earnings. Using a hand-collected dataset on 286 earnings restatements, Palmrose,
Richardson, and Scholz (2004) find that a restatement announcement is associated with a share
value decline of 9.2%, on average. Anderson and Yohn (2002), Akhigbe, Kudla, and Madura
(2005), and Agrawal and Chadha (2005) collect restatement samples using such media-based
databases as Lexis-Nexis. They also find negative stock price reactions to news of a restatement,
although their point estimates are smaller in magnitude. Several teams of researchers use
restatement data compiled by the Government Accountability Office (GAO 2003). Using the
GAO data, Hribar and Jenkins (2004), Arthaud-Day, Certo, Dalton, and Dalton (2006), Burns
and Kedia (2006), and Desai, Hogan, and Wilkins (2006) estimate the loss at 9-11%. Using the
same database, however, Kravet and Shevlin’s (2010) obtain an estimate of -5.4%.
The differences in point estimates illustrate a potential problem in using the GAO data to
identify instances of misconduct. Hennes, Leone, and Miller (2008) point out that the GAO
database includes both intentional and unintentional restatements. By their estimate, fully 73%
9
of the GAO restatements are unintentional or technical restatements, or what they call “errors.”
Hennes et al. measure the average abnormal stock return for 69 such “error” restatements as -
1.93%. In contrast, the average abnormal stock return for events that reflect intentional
misstatements, or what they call “irregularities,” is -13.64%. This implies that the consequences
to intentional misrepresentation are much larger than they are for minor accounting errors.
Researchers who use the GAO database must decide whether a restatement event
corresponds to a material case of financial misconduct. The fact that different researchers obtain
different point estimates of the share value loss reflects, in part, different choices on which
restatements to include in their analyses.
Karpoff, Lee, and Martin (2008a) point out another potential problem from using
restatements to measure the loss from financial misconduct. In some cases investors first learn
of the misconduct when the firm announces a restatement. In many other cases, however, a
restatement is made many weeks or months after the first disclosure of misconduct. For
example, HealthSouth Corp. engaged in a financial fraud from 1996 until March 2003. The firm
restated some previous years’ earnings in 2005. But investors first learned of the misconduct in
late 2002 and early 2003 through a series of press releases and federal regulatory
announcements. Clearly, by the time of the 2005 restatement, the firm’s misconduct was old
news. Karpoff, Lee, and Martin (2008b) report that restatement announcements coincide with
the first public revelation for only 14% (53 out of 371) of the misconduct cases in their sample.
This implies that restatement data will provide low power tests of the hypothesis that news of
financial misconduct affects stock prices. Despite such power issues, the evidence clearly
indicates that share prices decline by a large and statistically significant amount when firms
announce an earnings restatement.
10
II.C. Studies based on securities class action lawsuits
Several studies use lawsuits as indications of financial misconduct. Most examine
securities class action lawsuits filed under SEC Rule 10b-5. Rule 10b-5 prohibits any “untrue
statement or a material fact or to omit to state a material fact … in connection with the purchase
or sale of any security.” Plaintiffs can sue for damages if they can show that managers willfully
released inaccurate financial information. In a 1964 case, the U.S. Supreme Court referred to
private lawsuits using Rule 10b-5 as acting like “private attorneys general,” helping to enforce
securities laws and promote accurate financial reporting.
As reported in Panel B of Table 1, most studies examine the share impacts around either
or both of two key events in a securities class action lawsuit: the public disclosure that prompts
the lawsuit, and the filing of the lawsuit itself. In an early investigation, Kellogg (1984) finds
that the initial discovery of financial misrepresentation that leads to a lawsuit is associated with
an average share price decline of 3.9%. Subsequent investigations, however, find that the public
disclosure is associated with much larger losses. Francis, Philbrick, and Schipper (1994) find
that the initial adverse earnings disclosure that subsequently prompts a class action lawsuit is
associated with a stock price decline of 17.16%. Using the lawsuit complaints to identify the
initial public revelation of financial fraud, Ferris and Pritchard (2001) find that the average three-
day abnormal return is -24.99%. They argue that such a large reaction is consistent with the
notion that plaintiffs’ lawyers select cases with large share price declines and a larger likelihood
of recovery. Using a large sample of 2,133 lawsuits from 1990 – 2001, Griffin, Grundfest, and
Perino (2004) measure a three-day average abnormal return of -16.60%.
11
Although news of the misconduct is associated with the largest share price declines, the
filing of a lawsuit also is associated with a decline in share values. The estimates range from -
2.71% (Bhagat, Bizjak, and Coles 1998, reporting on “SEC-type Suits”) to -4.66% (Gande and
Lewis 2009). These results suggest that, even after the public learns of the misconduct, news of
the filing conveys additional and unfavorable information about the eventual cost of the
misconduct to the firm.
II.D. Studies based on AAERs
A large accounting literature examines the quality of firms’ reported earnings (see
Dechow, Ge, and Schrand (2010) for an excellent survey). Several papers in this literature use
AAERs to identify instances of financial misconduct. As described in the first AAER issued in
1982, an AAER is a designation that the SEC attaches to an enforcement release when it
involves an accountant. The idea is that “…[I]nterested persons will be able to easily distinguish
enforcement releases involving accountants from releases in which the Commission announces
the adoption or revision of rules related to financial reporting or discusses its interpretive views
on financial reporting matters” (page 1).3
Feroz, Park, and Pastena (1991) first introduced the idea of using AAERs to identify
instances of financial misconduct. For 58 firms that subsequently were the subjects of AAERs,
the public disclosure of misconduct is associated with a two-day average abnormal stock return
of -12.9%. Dechow, Hutton, and Sloan (1996) report a one-day average abnormal stock return
of -8.8% for 92 firms that were the subjects of AAERs. Beneish (1999) uses AAERs to identify
50 firms that overstated earnings. Most likely because these firms’ misconduct was particularly
3 Accounting and Auditing Enforcement Release No. AAER-1, 1982 SEC LEXIS 2565, May 17, 1982; Accounting and Auditing Enforcement Release No. AAER-1, 1982 SEC Act Rel No. 6396, Exch Act Rel No. 18649 / April 15, 1982.
12
egregious, the 3-day abnormal stock return for the initial disclosure is relatively large at -20.2%.
Bonner, Palmrose, and Young (1998) also use AAERs to identify a sample of misconduct firms.
They do not report an event study, but report that the average abnormal return over the 12-month
period centered on the end of a firm’s violation period is -24%. Karpoff and Lou (2010) report
that the end of the violation period is on average 2 months before the initial public revelation of
the misconduct, so this result likely reflects, in part, the share price decline that accompanies
such revelations. Finally, Miller (2006) reports an average abnormal return of -6.3% upon
publication of the first press article for a sample of firms that attracted enforcement releases
receiving an AAER designation.
In addition to these results, Feroz, Park, and Pastena (1991) examined share price
reactions to related events. The initiation of an SEC investigation that leads to an AAER is
associated with a share price decline of 7.5%, on average. Feroz et al. find no significant share
price reaction to the announcement of a settlement of an SEC action.
All of these studies conclude that news of financial misconduct reduces share values.
The point estimates of the loss, however, vary widely. As with most studies, some of the
discrepancy can be attributed to differences in the specific measure of abnormal returns and
sample selection. Below, however, I argue that much of the variation is due to differences in the
power of the tests. Some samples yield powerful tests of the hypothesis that news of misconduct
is associated with share value losses because they correctly identify the earliest public release of
information about cases that actually involve misconduct. Other samples, in contrast, tend to
capture neutral events, or events that occur well after the initial release of information about the
misconduct.
13
II.E. Studies based on SEC enforcement actions
Karpoff, Lee, and Martin (2008b) (hereafter, KLM) measure the impact on share values
of SEC enforcement actions for financial misrepresentation from 1978 through September 2006.
They claim that their sample is the universe of federal enforcement actions for violations of
provisions that require firms to keep accurate books and records (15 U.S.C. §§ 78m(b)(2)(A))
and to maintain a system of internal accounting controls (15 U.S.C. §§ 78m(b)(2)(B)).
Figure 1 illustrates the sequence of events that constitute a typical SEC enforcement
action. Most enforcement actions follow a conspicuous trigger event that publicizes the potential
for misconduct and attracts the SEC’s scrutiny. Trigger events include self-disclosures of
malfeasance, restatements, auditor departures, unusual trading, and whistle-blower lawsuits.
KLM report that the average one-day market-adjusted share return on such trigger dates is
–25.24%. Using an updated version of the KLM data, Karpoff and Lou (2010) report an average
abnormal return of -18.2% on the trigger date. These results are similar in magnitude to those
from studies, such as those in Panel C of Table 1, that focus on the initial disclosure of
misconduct.
KLM (2008b) report results that help to clarify the discrepancies in many of the results in
Table 1. When the trigger event is an earnings restatement announcement, the average abnormal
return is -24%. This is much larger than the averages in Panel B, but this is because many of the
restatements in prior studies are preceded by prior public revelation of the misconduct. That is,
most samples of earnings restatement announcements include many events that convey relatively
little new information about the misconduct.
14
KLM also report that the initial filing of a securities class action lawsuit, which typically
occurs after the trigger date, is associated with an average share price decline of -7.0%. This is
larger than the results for filing dates reported in Panel B. This reflects the fact that lawsuits in
the KLM sample are related to relatively egregious or costly types of financial misconduct,
because they are accompanied by SEC sanctions. The lawsuits investigated by the papers in
Panel B, in contrast, include a broader range of events, including relatively small lawsuits that
impose little cost on shareholders.
Following a trigger event, the SEC gathers information through an informal inquiry that
may develop into a formal investigation of financial misconduct. As illustrated in Figure 1, the
announcement of a formal investigation is associated with an average share price decline of
13.7%.
The SEC releases its findings and penalties in its Administrative Releases and Litigation
Releases, some – but not all – of which also receive a designation as an AAER. (KLM (2008a)
report that 63% of the regulatory releases in their sample also were assigned an AAER number.)
If the SEC proceeds and imposes sanctions, the news of its initial regulatory action is associated
with a further 9.6% decline. Subsequent releases that indicate that the matter is resolved are
associated with an average decline of 4.2%. These are the events that are captured in the AAER-
based samples summarized in Panel D. Notice that most researchers who use AAERs to infer
financial misconduct backfill their data to identify the first disclosure of the misconduct. So the
results in Panel D most likely reflect a combination of trigger events and regulatory releases as
illustrated in Figure 1.
The events illustrated in Figure 1 typically take several years to play out. Karpoff and
Lou (2010) report that in their sample of enforcement events, the median length of the violation
15
period is 24 months, and the median length from the beginning of the violation until its initial
public revelation is 26 months. From the initial public revelation until the end of the
enforcement action takes an additional 41 months.
III. Reputational losses for financial misconduct
The research summarized in Table 1 clearly establishes that share values fall upon news
of financial misconduct. Some authors interpret such losses as a type of penalty that deters
misconduct (e.g., see Gunthorpe 1997). This, however, confuses the source with the measure of
firm losses. In reasonably efficient markets, share prices reflect investors’ expected values of
future cash flows to equity. They change when such expectations change. That is, changes in
share values are not the source of penalties, but rather, a measure of investors’ expectations of
the penalties. The sources of penalties include the regulatory fines, class action settlements,
higher cost of capital, and impaired operations that result from the discovery of misconduct. In
addition, share values decline as investors realize they had been relying on incorrect financial
information to forecast the firm’s future cash flows.
III.A. Legal penalties for financial misconduct
One reason share values decrease is that investors anticipate that the firm will face
monetary or non-monetary sanctions from the SEC and other regulators, or lawsuits related to
the misconduct. Consider, for example, the array of legal penalties described by KLM (2009)
imposed on WorldCom related to its 1999-2002 accounting fraud:
On June 27, 2002, the U.S. Securities and Exchange Commission (SEC) filed a complaint alleging that WorldCom Inc. misreported its earnings from January 1999 through March 2002 by a cumulative amount of $11 billion. In subsequent actions, the SEC halted trading in WorldCom stock for two days, barred four WorldCom executives from serving
16
as officers or directors of publicly held companies, suspended five directors and employees who were CPAs from appearing or practicing before the Commission as accountants, and fined the company $2.25 billion (an amount reduced in bankruptcy and district court negotiations to $750 million). The Department of Justice (DOJ) fined the company $27 million, ordered the firm to provide an additional $670,000 in communication services, and prosecuted CEO Bernie Ebbers, CFO Scott Sullivan, and four others for financial fraud, culminating in combined sentences of 32.4 years in jail and $49.2 million in restitution. In a parallel class action lawsuit, the firm and its officers paid a total of $6.8 billion to investors.
KLM (2008b) report that the SEC imposed an average of 1.4 non-monetary sanctions for
each of the 585 firms in their sample. Most (90.6%) were cease-and-desist orders or permanent
injunctions – actions that appear to impose extremely small penalties. Monetary penalties are
less common. Only 47, or 8%, of the 585 firms were assessed monetary penalties by regulatory
agencies. The mean fine was $106.98 million, but this is affected by WorldCom, which is an
outlier. Excluding the WorldCom case, the mean fine is only $59.8 million. In recent years the
SEC has imposed larger fines in high-profile enforcement actions involving such firms as Qwest
Communications International ($250 million), Bristol-Myers Squibb Co. ($100 million), and the
Royal Dutch/Shell Group ($100 million).
Monetary penalties from shareholder class action lawsuits are much more common than
fines from regulators. KLM report that 39% of the firms in their sample paid class action
settlements, with a mean settlement of $37.7 million.
Although legal penalties sometimes are large, they are much smaller than firms’ share
value declines when their misconduct is revealed. Using the KLM (2008b) data, the mean legal
penalty equals only 3.1% of the total loss in the market values of the targeted companies. Class
action settlements account for an additional 5.4% of the loss. Together, these legal penalties
equal only 8.8% of the total dollar loss associated with the enforcement actions.
17
III.B. Share deflation to correct to full-information value
Another reason that news about financial misconduct decreases share values is that
investors learn that the firm’s accounting statements contain bad information. So the stock price
must be revalued using information that the firm previously was cooking its books. This can
affect investors’ view of the firm’s cash flow from operations or the growth in such flows. Here
is an example:
Suppose the Bunda Company is an all-equity firm with a book value of assets of $100 and a market-to-book ratio of 1.5. The market value of the firm's assets, and its stock, is therefore $150. But then assume that Bunda issues a misleading financial statement that overstates its asset values by $10. If the firm's market-to-book ratio stays the same, its share values will increase temporarily by ($10 x 1.5) to $165. But when the financial misrepresentation is discovered, Bunda's book values will adjust back to $100. And if there is no other impact, the share value will fall back to $150. That is, Bunda's shares will drop in value from their inflated value of $165 to their "correct" value of $150.
Things get more complicated if the market-to-book ratio depends on the firm’s reported assets,
as would be the case if investors’ views of the firm’s long-term growth depends on reported
financials.
Figure 2 illustrates the share price inflation for an actual firm. In early 1997 Xerox began
to inflate its reported earnings by accelerating its recognition of revenue on its equipment lease
contracts. Rather than recognizing revenue when lease payments were made, it booked the full
stream of expected lease payments when the lease agreement was made. The effect was to
increase near-term revenue and earnings. As illustrated in Figure 2, this reporting strategy
helped to boost Xerox’s share price for over two years. The only way such revenue-acceleration
schemes work is if the company generates sufficiently high real growth in sales to make up for,
and to cover up, the eventual shortfall in future periods. Xerox was unable to generate such high
18
sales growth, and eventually it had to recognize that it would not be able to cover up its
aggressive reporting practice. On October 8, 1999, the firm announced that its third quarter
earnings would not meet projections. Investors, inferring that the announcement was only the tip
of a larger problem, lowered their values of the company and the stock price fell immediately. In
the ensuing months, Xerox made additional disclosures of earnings problems, the SEC launched
an investigation of its reporting practices, and the firm eventually was penalized for
misrepresenting its financial statements.
Clearly, Xerox’s share price was artificially inflated during the period that its books were
in error. The share price decline, as investors found out about Xerox’s misconduct, was due in
part to the price returning to the level it would have been had Xerox never misrepresented its
financials in the first place. Measuring this share inflation – and deflation – is a central point of
argument in many securities class action lawsuits. I am aware, however, of only one published
attempt to develop a systematic estimate of the share price inflation. KLM (2008b) use a
multiple of the firm’s asset write-offs in the period after its misrepresentation is uncovered. The
multiple, in turn, is the median market-to-book ratio of firms in its 2-digit SIC industry.
Using this method, KLM estimate that, on average, 28.8% of the drop in share values is
due to this correction in firms’ artificially inflated share values. This is the amount by which
firms’ share values decline when their financial misconduct is discovered because investors have
more accurate information about the firm’s financial performance.4
4 In dollar terms, the 28.8% effect equals $54 million for the average firm. This indicates that, on average, share values were falsely inflated $54 million by the misrepresentation. The correction amount is based on the firm’s largest asset write downs during the total period the firm was under suspicion or investigation for financial misrepresentation. Many write downs are legitimate and unrelated to any cooking of the financial books. So the 28.8% figure is an overstatement of the correction amount. Nonetheless, even with such an overstatement our main observation – that the impaired operations and reputation slice of the pie is very large – is apparent.
19
In Figure 3 I apply the KLM method to illustrate the size of the share inflation and
deflation for Xerox. During the period Xerox was misrepresenting its revenues and earnings, its
share values increased. This is illustrated in stylized form by the green line. Immediately before
Xerox announced that third quarter 1999 earnings would be less than previously forecast, the
firm’s market capitalization was $16.864 billion. The cumulated loss in market cap, measured
over the sequence of events by which investors learned of the misconduct, was $5 billion,
leaving the company with a market cap (adjusted for market-wide movements) of $11.864
billion.
Of the $5 billion loss, the KLM method implies that $1.15 billion, or 23%, is the
deflation effect. This represents share values returning the level at which they hypothetically
would have been if no misrepresentation had occurred. An additional $0.523 billion of the loss
can be attributed to amounts Xerox paid in fines and to settle a class action lawsuit. The rest of
the loss – $3.3 billion – is due to something else. The most plausible explanation is that much,
even all, of the $3.3 billion is due to impaired operations and lost reputation – which I abbreviate
to simply “the reputational loss.”
III.C. The reputational loss from financial misconduct
The discovery of financial misconduct can impair the firm’s operations if its managers
are indicted, lose their jobs, or divert time and energy to the investigation rather than attending to
company business. The investigation also could force the firm to adopt new monitoring and
control policies that increase its cost of operations. Such higher costs of operation will lower the
firm's future earnings and result in a lower current value. In addition, the firm suffers a loss in
reputation that also lowers firm value.
20
The term “reputation” has many colloquial meanings. Merriam-Webster.com defines
“reputation” as:
1 a : overall quality or character as seen or judged by people in general
b : recognition by other people of some characteristic or ability
2: a place in public esteem or regard : good name
Here, I define reputation as the present value of the cash flows earned when an individual or firm
eschews opportunism and performs as promised on explicit and implicit contracts. It is the value
of the quasi-rent stream when accrues when counterparties offer favorable terms of contract as
insurance against being cheated.5
It is well established that firms that restate earnings or have financial reporting problems
face a higher cost of capital (e.g., see Hribar and Jenkins 2004, or Kravet and Shevlin 2010).
This occurs as investors charge a premium for their increased risk of investing in firms with poor
internal controls and/or managers who have displayed a willingness to behave opportunistically.
Graham, Li, and Qiu (2008) also show that lenders also charge higher rates to firms that have
restated earnings. And Murphy, Shrieves, and Tibbs (2009) show that misconduct firms also
have lower cash flows from operations.
Collectively, this evidence indicates that firms that misreport their financials lose
investors’ trust, pay a higher cost of capital, and may even experience lower cash flows from
operations. This is because at least some of these firm’s investors and counterparties do not want
to do business with a company that lied in its financial statements. Banks are reluctant to lend
the company money if they cannot trust the firm's financial statements. Existing lenders increase
their scrutiny of the company and become less flexible with their terms of credit. Suppliers tend
5 Klein and Leffler (1981) do not use the term “reputation.” Nonetheless, the reputation loss equals W2 in their model.
21
to be reluctant to offer the firm trade credit and could require that the firm pay cash for its
supplies. The firm's customers could even be affected, if the firm's bad publicity causes them to
doubt the firm's products and guarantees. For example, customers might place less faith in the
firm's guarantees to support warranties or to supply compatible services or parts in the future.
The sum total of all of these effects is what I call the firm’s reputational loss. The
reputational loss is the present value of the higher costs and/or lower revenues when firms are
discovered to have cheated their investors, suppliers, employees, or customers. It occurs because
of direct impairments to the firm’s ongoing operations, and also because counterparties alter the
terms with which they are willing to continue to do business with the firm.
But how costly are such effects? One measure is the portion of the firm’s total loss that
cannot be explained by other factors. By this measure, Xerox’s reputational loss was $3.33
billion, or 67% of its total loss. This is illustrated in Figure 3.
The Xerox example is close to the average from a large sample of firms that were
revealed to have cooked their books. Using a sample of 384 firms that faced SEC sanctions for
financial misrepresentation, KLM (2008b) estimate that 8.8% of these firms’ total losses were
due to legal penalties, 24.5% to share deflation, and 66.6% to lost reputation. This breakdown is
illustrated in Figure 4. Using median rather than mean values, limiting the sample to firms that
survived the enforcement process, or using alternate measures of the share deflation all yielded
similar or larger measures of the reputation loss.
Stated differently, KLM’s results indicate that firms can temporarily increase their share
values by misrepresenting their earnings and assets. When the misrepresentation is detected,
however, firm value decreases by more than the original inflation. For every dollar of inflated
value during the period that the firm’s books are in error, the firm loses that dollar when its
22
misrepresentation is uncovered. In addition, the firm loses an additional $3.08. Some of this
additional loss – 36¢ – is due to the legal penalties these firm’s incur. Most – $2.71 – is due to
lost reputation.
These results support the argument that financial reporting violations carry large
penalties. The largest penalties are not from regulators or private lawsuits. Rather, they are from
the firm’s investors and other counterparties. It is unlikely that investors and firm counterparties
intend, or even are aware, that they impose penalties on the offending firm. Rather, they simply
are protecting their own interests, requiring a premium to do business with firms that are less
trustworthy than they previously believed.
IV. Market value losses for other types of business misconduct
A large amount of research also has investigated whether there are significant share price
reactions to other types of business misconduct. The results are summarized in Table 2.
Here, the results are not as straightforward as for financial misconduct. Panel A reports
on eight studies that examine illegal corporate activities, including fraud, bribery, breach of
contract, securities fraud, patent infringement, environmental violations, price fixing, illegal
campaign contributions, tax evasion, and OSHA violations. Some find evidence of negative
stock price effects. Davidson and Worrell (1988), for example, report an average abnormal
return of -3.48% for a sample of 96 announcements of “corporate illegalities.” Reichart, Lockett,
and Rao (1996) and Bhagat, Bizjak, and Coles (1998) report statistically negative stock returns to
announcements of indictments or lawsuits for firm misbehavior. Overall, however, the size of
the share value losses are much smaller than reported in Table 1 for financial misconduct. Two
studies even report that the average share price reactions are not statistically significant.
23
A drawback of the results in Panel A is that they are drawn from samples that combine
different types of misconduct. Several studies, in contrast, examine only one type of misconduct,
or provide results on isolated types of misconduct. Panel B, for example, reports on four studies
that examine track the share price reaction to news that a firm defrauded or failed to perform as
contracted with a counterparty. As an example, during the 1980s BeechNut Corp. sold bogus
apple juice for babies(!). Labeled "100%" apple juice for infants, the product actually was a
concoction of water, sugar, artificial flavors, and food coloring. Eventually the fraud was
discovered and the company's President and V-P for operations were sentenced to jail. All four
of the studies in Panel B find that news of such misconduct is associated with share price
declines that range from 1-5%.
Panel C reports on studies that examine frauds of government. As an example, in the
1990s Rohr Inc. falsified test reports for the parts and adhesive materials used to attach the wings
onto two military aircraft, the Air Force C-5 transport plane and the Navy F-14 Tomcat fighter.
The prospect of wings falling off planes was too much for the Department of Defense, and Rohr
Inc. eventually pleaded guilty to eight counts of making false statements.6 Again, the results of
these studies indicate that news of such misconduct is associated with small but statistically
significant share price declines.
Several papers have examined product recalls. While not necessarily the result of
intentional misconduct, a product recall represents at least some degree of failure by the firm to
deliver on its explicit or implicit guarantee to provide a safe and reliable product. Recalls in the
automobile industry are associated with small stock price declines. The point estimate from
Barber and Darrough (1996), who examine the largest sample of the papers in this group, is -
6 In a similar case in 1991, General Motors' Hughes Aircraft division falsified tests of the electronic circuits used in aircraft and missile guidance systems. This raises horrible images of missiles going to the wrong targets because of an error in the guidance system!
24
0.36%. The effects of a drug recall announcement is larger, averaging from 6%-8% (see Jarrell
and Peltzman 1985; Ahmed, Gardella, and Nanda 2002).
Three papers in Panel E examine a variety of product recalls. Pruitt and Peterson (1986)
and Davidson and Worrell (1992) conclude that the average stock price reaction is negative but
less than 1% in magnitude. Using a long event window, Rubin, Murphy, and Jarrell (1988)
report an average stock price reaction of -6.9%.
An airplane crash represents an extreme abrogation of the firm’s guarantee to provide a
safe product or service. Three studies listed in Panel F measure airlines’ loss in value when a
plane crashes. Chalk (1987) and Borenstein and Zimmerman (1988) find that share values
decline. Mitchell and Maloney (1989) show that the loss depends on the cause. When pilot error
is to blame, the airline suffers a statistically significant loss of 1.68%. When external factors are
blamed for the crash, such as the weather, the stock price change is statistically insignificant.
Panel G summarizes the results of other types of product failures. Mitchell (1989) and
Dowdell et al. (1992) examine a 1982 case of product tampering, in which capsules of Tylenol
were laced with cyanide and seven people died. This case is noteworthy because both customers
and the manufacturer, Johnson & Johnson, were victimized by the perpetrator. Even though it
was not directly responsible, however, Johnson & Johnson experienced a significant decline in
share values.
Bhagat, Bizjak, and Coles (1998) and Prince and Rubin (2002) measure the stock price
reactions to announcements that lawsuits have been filed seeking damages for product liability.
The reactions are negative and small in magnitude, and are statistically significant only in the
Bhagat et al. sample.
25
Another example of a product failure is a breach of security among software and database
providers. Such breaches can result in the public dissemination of clients’ and customers’
personal information. As shown in Panel I, most studies find that security breaches result in
small or statistically insignificant stock price reactions. Breaches that expose confidential
information, in contrast, are more likely to result in significant share price declines for the
offending firm.
Panels K through O of Table 2 report findings for employee-related misconduct, false
advertising, investigations of IPO underwriters, punitive damage-seeking lawsuits, antitrust
violations, and price fixing. All are associated with statistically significant declines in share
values.
The results in Panels H and I, however, are mixed. Panel H reports the results of seven
investigations of environmental violations. A typical example of such violations is a case
involving Tyco International, Inc. From 1999 to 2001, employees of a Tyco subsidiary dumped
wastewater containing high levels of lead, copper, oil, and other chemicals into the public water
treatment systems of two cities in Connecticut. In an attempt to cover up its pollution, company
employees altered wastewater samples and falsified reports to the U.S. Department of
Environment Protection. In several papers, news of such violations is associated with negative
share price reactions. But other researchers conclude that the average share price reaction is
statistically insignificant.
Panel I reports on investigations into violations or frauds that affect unrelated parties, i.e.,
not parties with whom the firm does business. Examples include violations for check kiting or
failure to report currency transactions.
26
V. Reputational losses for other types of business misconduct
The results in Table 2 indicate that, frequently, news that a firm had engaged in some
type of misconduct is associated with a decline in its share values. This indicates that investors
regard such news as costly for most firms. But what is the source of the cost? Firms that
defraud customers, for example, are likely to face regulatory penalties and class action lawsuits.
In fact, the sizes of such legal penalties can be substantial. Murphy, Shrieves, and Tibbs (2009)
report that the mean legal fine for firms that commit frauds that affect related parties (as in Panel
B of Table 2) is $18.79 million. For violations that affect third parties (Panel K), the mean fine
is $32.87 million. Karpoff, Lee, and Vendrzyk (1999) report that defense procurement frauds
triggered fines that average $8.0 million plus required restitution of $4.4 million. Not all legal
penalties are as large. Cohen (1992) reports that the penalties for antitrust violations, for
example, average $0.3 million in fines plus an additional $0.1 million in restitution.
To tease out the importance of reputational losses, it is essential to examine whether the
share value losses can be explained by such legal penalties, or other losses that occur when a
firm’s misconduct is revealed. Other losses include any losses from having to abandon the
activities that were revealed. For example, Van den Broek, Kemp, Verschoor, and de Vries
(2010) consider the effect on the firms’ conspiracy-related profits when estimating the
reputational loss from antitrust actions.
Many of the researchers cited in Table 2 have made attempts to distinguish between share
value losses and the different types of penalties that they reflect. Table 3 summarizes the
qualitative assessments made by authors who have sought to shed light on the size of any
reputational loss. The estimates of reputational loss are positive for frauds of related parties,
some frauds of government, false advertising, product recalls, air safety disasters, other product
27
failures, employee discrimination or safety problems, punitive damage awards, and antitrust
violations. Researchers have concluded that reputational losses do not arise, however, for
environmental violations, security breaches, and regulatory violations or frauds affecting
unrelated parties.
To illustrate the nature of such conclusions, consider environmental violations. Jones and
Rubin (2001) find that, among a sample of public utility companies, news of an environmental
violation is not associated with a decline in share values, on average. They conclude that there
must be very little reputational loss for these companies. Karpoff, Lott, and Wehrly (2005), in
contrast, find that the average share price reaction is negative in their broader sample of
environmental violations. However, they also find that the legal penalties for the firms in their
sample are of similar magnitude to the share value losses. Violations affecting air quality during
the 1980s and 1990s, for example, resulted in an average fine of $31.7 million (in constant 2002
dollars). In addition, the guilty companies were required to incur costs averaging $123 million to
comply with air quality rules or to remediate the damage of their pollution. Firms that were
responsible for contaminated sites faced an average penalty of $11.0 million and a clean-up cost
of $108 million. While firms that are caught contaminating air, water, or land resources face
significant costs, these costs are all those imposed by regulators and the courts. Since these costs
fully explain the defendant firms’ losses in share values, they conclude that the reputational loss
from violating environmental rules is negligible, on average.
A number of researchers have made quantitative assessments of the importance of
reputational losses. These results are summarized in Figure 5. When more than one set of
researchers provides information from which a reputational loss can be estimated, I present the
average of the loss estimates.
28
Averaging over the estimates from three studies, 70% of firms’ losses from financial
misrepresentation can be attributed to lost reputation. Reputational losses also explain large
portions of the losses that accrue from several other types of misconduct. The largest
reputational loss component is for punitive damage lawsuits. I note, however, that such lawsuits
include many different types of misconduct, from fraud to business negligence to defective
products. So there is some overlap between the events in the punitive lawsuit category and other
types of misconduct. For these types of wrongdoing, the penalties imposed through legal fines,
penalties, and damage awards are dwarfed by the huge reputational losses these firms
experience. For security breaches, environmental violations, and unrelated party violations, in
contrast, reputation plays a negligible role.
VI. How reputation encourages greater quality assurance
VI.A. How reputational capital discourages opportunism
To understand why reputational losses are large for some types of misconduct and not for
others, it is important to understand how and why reputational losses occur at all. Consider the
following questions:
• When you travel, do you dine at local eateries or well-known chain restaurants?
• When did you last buy a car? Did you go to a new-car dealer or a used-car dealer? Did
you buy through a newspaper ad?
• When you buy over-the-counter medicines (say, acetaminophen), do you buy brand name
products like Tylenol™, or generic products?
• When you buy acetaminophen for an infant child, do you buy brand name products or
generic products?
29
The fact that peoples’ answers differ illustrates how reputation works and why reputational
losses can occur when companies behave opportunistically.
Consider consumers’ choices of restaurants when they travel. Adventurous travelers who
eat at a local Joe's Diner might experience a gastronomic delight – or they might get Ptomaine
poisoning. That is, without local knowledge the traveler faces a high-variance outcome. To
avoid such uncertainty, many travelers prefer to eat at a chain restaurant – say, McDonalds.
McDonalds and other chain restaurants sell not only hamburgers and fries, but also a high level
of quality assurance. Note that high quality assurance is not the same thing as high quality.
Rather, it is a high level of certainty about what the buyer is going to receive. People who eat at
McDonalds know with a high degree of certainty what they are going to get, even if they are not
in their hometowns.
Not surprisingly, McDonalds has invested heavily to provide its customers a high level of
certainty in their products. Through its training and franchising procedures, the company works
to assure that its dining experience meets a certain standard and that its customers will not be
surprised. In return, the company is rewarded with customers who return again and again.
Notice that McDonalds could decide to cut corners and sell its customers poorer fare than it has
promised. But to do so would be to alienate its customers and see its sales dry up. In other
words, McDonalds has high reputational capital, just like other firms have high physical capital.
If McDonalds were to cheat its customers its reputational capital would depreciate. That is, the
firm would suffer a reputational loss, as it would lose future sales and the profits that come with
them.7
7 The literature on franchising emphasizes the role that the franchise contract plays in discouraging franchisees from free-riding on the brand name, and in encouraging a standardized product (e.g., see Klein 1995).
30
Similar concerns explain why some people buy automobiles through established car
dealers, while others purchase through newspaper ads. One difference between a car dealer and
a private seller is the degree of reputation at stake. Car dealers who sell lemons risk losing future
sales. For a private seller using a newspaper ad, in contrast, little reputation is at risk. People
who buy from a private person through a newspaper ad prefer not to pay the price premium that
accompanies a dealer's higher reputation – the quasi-rent that Klein and Leffler (1981) show
serves as incentive to not cheat. Such buyers might have automotive skills that allow them to
assess the car's quality without having to rely on a dealer's reputation. Or, since quality
assurance is an income-normal good, consumers who buy from low-reputation outlets might
simply have low incomes and low demands for quality assurance.
We observe similar differences in the sale and purchase of acetaminophen. Some
consumers like the higher level of certainty that comes from buying a brand name like
Tylenol™. They know, or at least intuit, that Tylenol™ has more to lose if it screws up and
delivers poor quality acetaminophen. Sellers of generic acetaminophen, in contrast, face smaller
reputational losses if they do not deliver the promised quality.
From sales figures, we can see that some people buy generic acetaminophen for
themselves but brand-name acetaminophen for their children. These consumers apparently place
a lower value on certainty when buying acetaminophen for themselves than when they buy
acetaminophen for their children. They know, at least implicitly, that brand-name sellers like
Tylenol™ would suffer large reputational losses if there are problems with their products, so
they are likely to control their production and distribution procedures with more care. Some
buyers are willing to pay a bit more for the extra quality assurance.
31
VI.B. Cross-sectional differences in reputational losses
These examples illustrate why different sellers face different reputational losses if they do
not deliver the products they promise to deliver. Similarly, some types of misconduct expose
companies to greater reputational losses than others. Lying to investors by misrepresenting
financial statements triggers large reputational losses. So does defrauding customers, as with the
incident in which BeechNut sold fake juice that was labeled "100% pure" apple juice. In such
incidents, the perpetrator reveals itself to be untrustworthy. Companies that defraud customers
therefore tend to lose sales. Those that cheat employees or other suppliers face higher input
costs or lost trade credit. And those that lie to their investors face higher financing costs.
Reputational losses are not uniformly high, however. As indicated in Figure 5, they are
negligible, on average, for firms that violate environmental laws. For these types of misconduct,
firms' losses in share values are attributable wholly to their legal penalties.
This is most apparent from the empirical results regarding environmental violations. In
theory, firms that violate environmental rules could suffer reputational losses, if consumers and
suppliers refuse to do business with environmentally unfriendly firms. After the 1989 Exxon
Valdez oil spill, for example, some consumers cut up their Exxon credit cards and vowed to buy
gasoline from other vendors. The data, however, show that, on average, any reputational loss
from harming the environment is negligible. Jones and Rubin (2001) and Karpoff, Lott, and
Wehrly (2005) argue that this is because firms who violate environmental rules do not impose
costs on parties with whom they do business. As an example, downstream fishermen are
damaged if an electroplating company dumps toxic chemicals into a municipal storm sewer. But
the fishermen do no business with the firm, and the firm’s customers have no direct incentive to
32
lower their demands for the firm’s products if the dumping does not affect the quality of those
products. As a result, the polluting electroplating company experiences no reputational costs.
A similar argument holds for frauds of unrelated parties. Although such actions as check
kiting are against the law, it is not evident that any of the parties with whom the firm does
business are harmed by the activity. As a result, when firms are caught violating these types of
rules they may face legal penalties. But since they do not directly harm their customers,
investors, or suppliers, they do not suffer a reputational loss. That is, investors do not expect
them to lose sales or face higher operating costs.
V.C. How reputational losses show up in firms' operations
The evidence implies that firms engaging in many types of misconduct incur large
reputational losses. This evidence is based, however, on observations that share prices decline
when investors find out about the misconduct. But are investors correct? That is, do reputational
losses actually show up in firms' subsequent performance? Do these firms subsequently lose
business, incur higher costs, or experience a higher cost of capital?
The research on this question is still developing. But a number of findings are consistent
with the event study results. For example, Karpoff and Lott (1993) find that firms charged with
defrauding customers and other stakeholders do in fact have lower operating earnings over the
following five years. Alexander (1999) reports that 57% of such firms experience termination or
suspension of specific contracts. However, such business losses occur only following frauds of
parties with whom the firm does business. Offenses against other parties with whom the firm
does not do business, in contrast, do not lead to a high rate of lost sales. Similarly, Murphy,
33
Shrieves, and Tibbs (2009) find that allegations of illegal acts are accompanied by a significant
decrease in firms' earnings and an increase in uncertainty over future earnings.
There is fairly strong evidence that financial misconduct results in a higher cost of capital
for the firm. Hribar and Jenkins (2004), Kravet and Shevlin (2010), and others show that the
cost of equity capital increases for firms that restate earnings. And Graham, Li, and Qiu (2008)
show that bank lender rates increase for restating firms.
Several papers document direct evidence of reputational losses. Beatty, Bunsis, and
Hand (1998) find direct evidence of operating losses for investment bankers that are investigated
by the SEC for problems in bringing IPO firms to the public market. These firms experience
sharp decreases in their shares of the IPO underwriting market after they are targeted by an SEC
investigation. The share prices of these firms' client companies also decline, indicating that the
decrease in the investment banker's reputation affects its clients as well.
One place in which we would expect to see important reputational effects is in online
auctions. Here, buyers deal with sellers they do not know and cannot even see. So we would
expect that sellers with high reputations would charge higher prices than others. Such price
premiums are the amounts that some buyers willingly pay for the guarantee that they will not be
ripped off. And the chance to earn such premiums is sufficient to encourage most high-
reputational sellers to deliver on their promise not to cheat buyers.
Consistent with such expectations, evidence indicates that high-reputation sellers on eBay
do in fact sell at higher prices than others, even for the same items. For example, Resnick,
Zeckhauser, Swanson, and Lockwood (2002) find that the prices of vintage postcards on eBay
average 7.6% higher for sellers with high reputation than for other sellers – even for the exact
same postcards. Dewally and Ederington (2002) find similar results for comic books sold on
34
eBay. The impact of a seller's reputation on price is particularly great when the quality of the
comic book has not been certified by a third party.
Finally, several findings indicate that managers who involve their companies in financial
misconduct end up losing their jobs. Jayaraman, Mulford, and Wedge (2004) find that managers
of firms that are subjects of SEC Accounting and Auditing Enforcement Releases tend to be
displaced at an unusually high rate. Desai, Hogan, and Wilkins (2006) and Agrawal and Cooper
(2007) find that managers of firms that have to restate their earnings share similar fates. KLM
(2008a) report that 92% of managers who the SEC identifies as involved in financial
misrepresentation lose their jobs; 81% lose their jobs even before the SEC imposes any
sanctions.
VII. Ten questions for further research
Opportunistic behavior by buyers and sellers, or its potential, can impede exchange. At
the extreme, it has the potential to cause markets to break down, impeding trade and increasing
poverty. At the beginning of this survey I noted that, when self-interested behavior includes
lying, cheating, stealing, or opportunistic gaming, Adam Smith’s invisible hand might not work
to channel private motivation into social gain.
Economists have long appreciated that repeat contracting has the potential to ameliorate
this problem. In theory, people and firms invest in and develop reputations that encourage others
to trust that they will not be cheated. It is difficult to measure reputation, however, and the
importance of reputation as a guarantor of honest dealing is an empirical matter.
As a way to gain insight into the importance of reputation, a number of researchers have
examined instances in which firms might be expected to lose reputation – that is, when they are
35
discovered to have engaged in illegal or opportunistic activities. The evidence indicates that
firms lose market value when they are caught in many types of misconduct, including financial
misrepresentation and frauds of related parties. A subset of researchers in this area have
attempted to understand the nature of such losses by partitioning them into portions that can be
explained by legal penalties, share price deflation, or lost cheating profits. The portion that
remains unexplained is used as a measure of reputational losses.
Using this approach, the evidence indicates that firms experience significant reputational
losses when their misconduct imposes costs on their counterparties. Firms that cook their
financial books face a higher cost of capital, and firms that cheat customers lose sales. For some
types of misconduct, however, there appear to be few reputational losses. A notable example is
environmental violations. Firms that violate, say Clean Air Act rules about emissions, appear
not to lose sales or face higher financing costs. This implies that reputation and market forces
play a small role in disciplining environmental violations, and that regulations and legal penalties
play a more important role.
To the extent that reputation plays an important role in encouraging honest dealing, it
offers a relatively low-cost solution to the lemons problem and facilitates the operation of Adam
Smith’s invisible hand. As such, understanding whether and how reputation helps to enforce
contractual performance is a vital part of understanding the extent to which markets work or do
not work.
The earliest papers cited in this review are over 20 years old. Nonetheless, the empirical
research on reputational losses has only begun to provide a picture of reputation’s role in
facilitating the development of markets and economic growth. I conclude by offering a list of
ten questions for future research. Work on these questions can begin to fill out the gap between
36
our theoretical understanding of reputation and the way it actually plays different roles in
different markets.
1. How important are reputational penalties around the world? This can help to explain the
extent to which market forces encourage honest dealing and facilitate the working of the
invisible hand.
2. What is the probability of getting caught for different types of misconduct? This can help
to explain the trade-off managers face when considering opportunistic actions. It also can help
guide public policy toward optimal penalties.
3. Given that reputational losses are important for some but not all types of misconduct, do
legal penalties conform to optimality conditions?
4. How do reputational penalties interact with public and private (lawsuit) enforcement of
securities and other laws?
5. How and when do firms rebuild damaged reputation?
6. What explains cross-sectional differences in reputational penalties?
7. Who detects corporate misconduct?
8. How do event-study measures of reputational penalties correspond to lower cash flows or
higher costs?
9. Given the legal and reputational losses from getting caught, why do managers engage in
illegal activity?
10. How does corporate governance affect the likelihood and cost of opportunistic behavior
by corporations?
37
References
Acquisti, Alessandro, Allan Friedman, Rahul Telang, 2006, Is there a cost to privacy breaches? An event study, Proceedings of the International Conference of Information Systems (ICIS) 2006. Agrawal, Anup, and Sahiba Chadha, 2005, Corporate governance and accounting scandals, The Journal of Law and Economics, 48, 371–406. Ahmed, Parvez, John Gardella, and Sudhir Nanda, 2002, Wealth effect of drug withdrawals on firms and their competitors, Financial Management, 31, 21-41. Akhigbe, Aigbe, Ronald J.Kudla, and Jeff Madura, 2005, Why are some corporate earnings restatements more damaging, Applied Financial Economics, 15, 327–336. Alexander, Cindy, 1999, On the nature of the reputational penalty for corporate crime: evidence, The Journal of Law and Economics, 42, 489-526. Anderson, Kirsten L., and Teri Lombardi Yohn, 2002, The effect of 10-K restatements on firm value, information asymmetries, and investors’ reliance on earnings, Working paper, Georgetown University. Arthaud-Day, Marne L.., S. Trevis Certo, Catherine M. Dalton, and Dan R. Dalton, 2006, A changing of the guard: Executive and director turnover following corporate financial restatements, Academy of Management Journal, 49, 1119–1136. Barber, Brad M., and Masako N. Darrough, 1996, Product reliability and firm value: The experience of American and Japanese automakers, 1973-1992, The Journal of Political Economy, 104, 1084-1099. Baucus, Melissa S., and David A. Baucus, 1997, Paying the piper: An empirical examination of longer-term financial consequences of illegal corporate behavior, Academy of Management Journal, 40, 129-151. Beatty, Randolph P., Howard Bunsis, and John R.M. Hand, 1998, The indirect economic penalties in SEC investigations of underwriters, Journal of Financial Economics, 50, 151-186. Beneish, Messod D., 1999, Incentives and Penalties Related to Earnings Overstatements that Violate GAAP, The Accounting Review, 74, 425-457. Bernile, Gennaro, and Gregg A. Jarrell, 2009, The impact of the options backdating scandal on shareholders, Journal of Accounting and Economics, 47, 2-26. Bhagat, Sanjai, and Roberta Romano, 2002, Event studies and the law: Part II: Empirical studies of corporate law, American Law and Economics Review, 4, 380-423. Bhagat, Sanjai, James A. Brickley and Jeffrey L. Coles, 1994, The Costs of Inefficient Bargaining and Financial Distress, Journal of Financial Economics, 35, 221-247. Bhagat, Sanjai, John M. Bizjak, and Jeffrey L. Coles, 1998, The shareholder wealth implications of corporate lawsuits, Financial Management, 27, 5-27. Block, Michael K., 1991, Optimal penalties, criminal law and the control of corporate behavior, Boston University Law Review, 71, 395, 414-415. Bonner, Sarah E., Zoe-Vonna Palmrose, and Susan M. Young, 1998, Fraud type and auditor litigation: an analysis of SEC accounting and auditing enforcement releases, The Accounting Review, 73, 503-532. Borenstein, Severin, and Martin B. Zimmerman, 1988, Market incentives for safe commercial airline operation, American Economic Review, 78, 913-935.
38
Bosch, Jean-Claude, E Woodrow Eckard.Jr, 1991, The profitability of price fixing: Evidence from stock market reaction to federal indictments, The Review of Economics and Statistics, 73, 309-317. Bosch, Jean-Claude, Insup Lee, 1994, Wealth effects of food and drug administration (FDA) decisions, Managerial and Decision Economics, 15, 589-599. Bromiley, Philip and Alfred Marcus, 1989, The deterrent to dubious corporate behavior: profitability, probability and safety recalls, Strategic Management Journal, 10, 233-250. Burns, Natasha, and Simi Kedia, 2006, The impact of performance-based compensation on misreporting, Journal of Financial Economics, 79, 35–67. Campbell, Katherine, Lawrence A.Gordon, Martin P. Loeb and Lei Zhou, 2003, The economic cost of publicly announced information security breaches: Empirical evidence from the stock market, Journal of Computer Security, 11, 431–448. Cavusoglu, Huseyin, Birendra Mishra, and Srinivasan Raghunathan, 2004, The effect of internet security breach announcements on market value of breached firms and internet security developers, International Journal of Electronic Commerce, 9, 70-104. Chalk, Andrew J., 1987, Market forces and commercial aircraft safety, The Journal of Industrial Economics, 36, 61-81. Cloninger, Dale, 1982, Moral and systematic risk: A rationale for unfair business practice, Journal of Behavioral Economics 7, 33-49. Cloninger, Dale, 1985, An analysis of the effect of illegal corporate activity on share value, Journal of Behavioral Economics 14, 3-31. Cohen, Mark A., 1992, Environmental crime and punishment: Legal/Economic theory and empirical evidence on enforcement of federal environmental statutes, The Journal of Criminal Law and Criminology, 82, 1054-1108. Davidson III, Wallace N. and Dan L.Worrell, 1988, The impact of announcements of corporate illegalities on shareholder returns, The Academy of Management Journal, 31, 195-200. Davidson III, Wallace N. and Dan L.Worrell, 1992, The effect of product recall announcements on shareholder wealth source, Strategic Management Journal, 13, 467-473. Davidson III, Wallace N., Dan L. Worrell and Chun I. Lee, 1994, Stock Market Reactions to Announced Corporate Illegalities, Journal of Business Ethics, 13, 979-987. Davidson III, Wallace N., Dan L.Worrell, and Louis T. W. Cheng, 1994, The effectiveness of OSHA penalties: A stock market-based test, Environment Lawsuit or case settlement involving toxic or hazardous waste materials mismanagement, Industrial Relations: A Journal of Economy and Society, 33, 283-296. Dechow, Patricia M., Richard Sloan, and Amy Sweeney, 1996, Causes and consequences of earnings manipulation: An analysis of firms subject to enforcement actions by the SEC, Contemporary Accounting Research 13, 1-36. Desai, Hemang, Chris E. Hogan, and Michael S. Wilkins, 2006, Reputational consequences of aggressive accounting: Earnings restatements and management turnover, The Accounting Review, 81, 83-112. Dewally, Michael, and Louis Ederington, 2006, Reputation, certification, warranties, and information as remedies for seller-buyer information asymmetries: Lessons from the on-line comic book market, Journal of Business, 79, 693-730.
39
Dewally, Michael, and Louis Ederington, 2002, A comparison of reputation, certification, warranties, and disclosure as remedies for information asymmetries: Lessons from the on-line comic book market, Working paper, University of Oklahoma . Dowdell, Thomas D., Suresh Govindaraj, and Prem C. Jain, 1992, The Tylenol Incident, ensuing regulation, and stock prices, The Journal of Financial and Quantitative Analysis, 27, 283-301. DuCharme, Larry L., Paul Malatesta and Stephan S. Sefcik, 2004, Earnings management, stock issues, and shareholder lawsuits, Journal of Financial Economics, 71, 27-49. Farber, David B., 2005, Restoring trust after fraud: Does corporate governance matter?, The Accounting Review 80, 539–561. Feroz, Ehsan H., Kyung Joo Park, and Victor S. Pastena, 1991, The financial and market effects of the SEC’s Accounting and Auditing Enforcement Releases, Journal of Accounting Research, 29, 107–142. Francis, Jennifer, Donna Philbrick, and Katherine Schipper, 1994, Shareholder Litigation and Corporate Disclosures, Journal of Accounting Research, 32, 137-164. Frooman, Jeff, 1997, Socially irresponsible and illegal behavior and shareholder wealth., Business and Society, 36, 221–249. Gande, Amar, and Craig M.Lewis, 2009, Shareholder initiated class action lawsuits: Shareholder wealth effects and industry spillovers, Journal of Financial and Quantitative Analysis, 44, 823–850. Gleason, Christi A., Nicole Thorne Jenkins, and W. Bruce Johnson, 2008, Financial statement credibility: The contagion effects of accounting restatements, The Accounting Review 83, 83-110. Golub, Alla., Joshua Detre, and John M. Connor, 2005, The profitability of price fixing: Have stronger antitrust sanctions deterred?, Working paper, Pudue University. Grady, Duane B., and Thomas H.Strickland, 2007, Public information as a deterrent to environmental infractions, Applied Economics, 39, 1961-1972. Graham, John R., Si Li, and Jiaping Qiu, 2008, Corporate Misreporting and Bank Loan Contracting, Journal of Financial Economics, 89, 44-61. Griffin, Paul A., Joseph A. Grundfest, and Michael A. Perino, 2004, Stock price response to news of securities fraud litigation: An analysis of sequential and conditional information, Abacus, 40, 21-48. Gunthorpe, Deborah L., 1997, Business ethics: A quantitative analysis of the impact of unethical behavior on publicly traded corporations, Journal of Business Ethics, 16, 537–543. Hamilton, James T., 1995, Pollution as news: media and stock market reactions to the Toxics Release Inventory data, Journal of Environmental Economics and Management, 28, 98-113. Harper, Richard K., and Stephen C. Adams, 1996, CERCLA and deep pockets: Market response to the Superfund program, Contemporary Economic Policy, 14, 107–115. Hendricks, Kevin B., and Vinod R. Singhal, 1997, Delays in new product introductions and the market value of the firm: The consequences of being late to the market, Management Science, 43, 422–436. Hersch, Joni, 1991, Equal employment opportunity law and firm profitability, Journal of Human Resources, 26, 139–153.
40
Hoffer, George E., Stephen W. Pruitt, and Robert J. Reilly, 1988, The Impact of Product Recalls on the Wealth of Sellers: A Reexamination, The Journal of Political Economy, 96, 663-670. Houser, Daniel, and John Wooders, 2006, Reputation in auctions:Theory and evidence from ebay, Journal of Economics and Management Strategy, 15, 353-369. Hovav, Anat, and John D'Arcy, 2003, The impact of denial-of-service attack announcements of the market value of firms, Risk Management and Insurance Review, 6, 97–121. Hribar, Paul, and Nicole T. Jenkins, 2004, The effect of accounting restatements on earnings revisions and the estimated cost of capital, Review of Accounting Studies, 9, 337–356. Jarrell, Gregg, and Sam Peltzman, 1985, The impact of product recalls on the wealth of sellers, The Journal of Political Economy, 93, 512-536. Jayaraman, Narayanan, Charles Mulford, and Lei Wedge, 2004, Accounting Fraud and Management Turnover, Working paper, Georgia Tech University. Jin, Ginger Z., and Andrew Kato, 2006, Price, Quality, and Reputation: Evidence from an Online Field Experiment Author, The RAND Journal of Economics, 37, 983-1004. Jones, Kari,and Paul Rubin, 2001, Effects of harmful environmental events on reputations of firms, Advances in Financial Economics, 6, 161-182. Kannan, Karthik, Jackie Rees, and Sanjay Sridhar, 2007, Market reactions to information security breach announcements: An empirical analysis, Journal of Electronic Commerce, 12, 69-91. Karpoff, Jonathan M., D. Scott Lee, and Valaria P. Vendrzyk, 1999, Defense procurement fraud, penalties, and contractor influence, Journal of Political Economy, 107, 809-842. Karpoff, Jonathan M., and John R.Lott, 1993, The reputational penalty firms bear from committing criminal fraud, Journal of Law & Economics, 2, 757-802. Karpoff, Jonathan M., D. Scott Lee, and Gerald S. Martin, 2008, The cost to firms of cooking the books, Journal of Financial and Quantitative Analysis, 43, 581-611. Karpoff, Jonathan M., John R. Lott Jr., and Graeme Rankine, 1998, Environmental violations, legal penalties, and reputation costs, Working paper, University of Chicago Law School. Karpoff, Jonathan M., John R. Lott, Eric W. Wehrly, 2005, The reputational penalties for environmental violations: Empirical evidence, Journal of Law and Economics, 48, 653-675. Kellogg, Robert L., 1984, Accounting activities, security prices, and class action lawsuits, Journal of Accounting and Economics, 6, 185-204. Klassen, Robert D., and Curtis P. McLaughlin, 1996, The impact of environmental management of firm performance, Management Science, 42, 1199–1214. Klein, B., 1995, The Economics of Franchise Contracts, Journal of Corporate Finance, 2, 9-37. Klein, B., and K. Leffler, 1981, The Role of Market Forces in Assuring Contractual Performance, Journal of Political Economy 89, 615. Kravet, Todd D., Terry J. Shevlin, 2010, Accounting restatements and information risk, Review of Accounting Studies, forthcoming.
41
Langus, Gregor, and Massimo Motta, 2007, The effect of EU antitrust investigations and fines on a firm's valuation, Centre for Economic Policy Research Discussion Papers, DP6176. Laplante, Benoit, and Paul Lanoie, 1994, The market response to environmental incidents in Canada: A theoretical and empirical analysis, Southern Economic Journal, 60, 657–672. Lei, Qin, 2010, Financial value of reputation: Evidence from the eBay auctions of Gmail invitations, Journal of Industrial Economics, forthcoming. Long, D. Michael, and Spuma Rao, 1995, The wealth effects of unethical business behavior, Journal of Economics and Finance, 9, 65-73. Margolis, Joshua D., Hilary A. Elfenbein, and James P. Walsh, 2007, Does it pay to be good? A meta-analysis and redirection of research on the relationship between corporate social and financial performance, Working Paper, Harvard Business School. Miller, Gregory S., 2006, The press as a watchdog for accounting fraud, Journal of Accounting Research 44, 1-33. Mitchell, Mark L., and Michael T. Maloney, 1989, Crisis in the cockpit-the role of market forces in promoting air travel safety, Journal of Law and Economics, 32, 329-355. Moore, Celia, H. Colleen Stuart, and Jo-Ellen Pozner, 2010, Avoiding the consequences of repeated misconduct: Stigma’s licence and stigma’s transferability, Working paper, London Business School. Muoghalu, Michael I., H. David Robison, and John L. Glascock, 1990, Hazardous waste lawsuits, stockholder returns, and deterrence., Southern Economic Journal, 57, 357–370. Murphy, Deborah L., Ronald E. Shrieves, and Samuel L. Tibbs, 2004, Determinants of the stock price reaction to allegations of corporate misconduct: Earnings, risk, and firm size effects, Working paper, University of Tennessee. Murphy, Deborah L., Ronald E. Shrieves, and Samuel L. Tibbs, 2009, Understanding the penalties associated with corporate misconduct: An empirical examination of earnings and risk, Journal of Financial and Quantitative Analysis, 44, 55-83. Ozbas, Oguzhan, 2007, Corporate fraud and real investment, Working paper, University of Southern California Palmrose, Zoe-Vonna, and Susan Scholz, 2004, The circumstances and legal consequences of non-GAAP reporting: Evidence from restatements, Contemporary Accounting Research, 21, 139-180. Palmrose, Zoe-Vonna, Vernon Richardson, and Susan Scholz, 2004, Determinants of market reactions to earnings restatements, Journal of Accounting and Economics, 37, 59-89. Peltzman, Sam, 1981, The effects of FTC advertising regulation, Journal of Law and Economics, 24, 403-448. Prince, David W., and Paul H. Rubin, 2002, The effects of product liability litigation on the value of firms, American Law and Economics Review, 4, 44-87. Pritchard, A.C., Stephen P. Ferris, 2001, Stock price reactions to securities fraud class actions under the private securities litigation reform act, John M. Olin Center for Law and Economics Research Paper 01-009. Pruitt, Stephen W., and David R. Peterson,, 1986, Security prices reactions around product recall announcements, Journal of Financial Research, 9, 113–122. Reichert, Alan K., Michael Lockett, Michael, and Ramesh P. Rao, 1996, The impact of illegal business practice on shareholder returns, The Financial Review, 31, 67-85.
42
Resnick, Paul , Richard Zeckhauser, John Swanson, and Kate Lockwood, 2006, The value of reputation on eBay: a controlled experiment, Experimental Economics, 9, 79-101. Richardson, Scott A., A. Irem Tuna, and Min Wu, 2002, Predicting earnings management: The case of earnings restatements, Working paper, University of Pennsylvania. Romano, Roberta, 1991, The shareholder suit: Litigation without foundation?, Journal of Law, Economics, & Organization, 7, 55-87. Rubin, Paul H., R.Dennis Murphy, and Gregg Jarrell, 1988, Risky products, risky stocks, Regulation, 12, 35-39. Shapiro, Carl, 1983, Premiums for High Quality Products as Returns to Reputations, Quarterly Journal of Economics, 98(4), 659-79. Skantz, Terrance R., Dale O.Cloninger, and Thomas H. Strickland, 1990, Price-fixing and shareholder returns: An empirical study, The Financial Review, 25, 153-163. Song, Wei-Ling, and Hatice Uzun, 2004, The Impact of Clients' Alleged Financial Reporting Fraud on Underwriter Reputation, Working paper #04-16, University of Pennsylvania. Strachan, James L., David B. Smith, and William L.Beedles, 1983, The price reaction to (alleged) corporate crime, The Financial Review, 18, 121-132. Tanimura, Joseph K. and M. Gary Okamoto , 2010, Reputational Losses in Japan Evidence from Corporate Scandals, Working paper, San Diego State University. Tanimura, Joseph K., and Eric W. Wehrly, 2009, The Market Value and Reputational Effects from Lost Confidential Information , Working paper, San Diego State University. Telang, Rahul and Sunil Wattal, 2004, Impact of software vulnerability announcements on the market value of software vendors -A n empirical investigation, Working Paper, Carnegie Mellon University. U.S. General Accounting Office, 2002, Financial Statement Restatements. Report to the U.S. Senate. U.S. General Accounting Office, 2003, Financial Statement Restatements Database. GAO-03-395R. U.S. General Accounting Office, 2006, Financial Statement Restatements: Updates of Public Company Trends, Market Impacts, and Regulatory Enforcement Actions. GAO-06-678. U.S. General Accounting Office, 2007, Securities Exchange Commission: Additional Actions Needed to Ensure Planned Improvements Address Limitations in Enforcement Division Operations. GAO-07-830. U.S. General Accounting Office, 2002, Financial Statement Restatements: Trends, Market Impacts, Regulatory Responses, and Remaining Challenges, GAO-03-138. van den Broek, Stijn, Ron Kemp, William F.C. Verschoor, and Anne-Clair de Vries, 2010, Reputational Penalties to Firms in Antitrust Investigations, Working paper, Erasmus School of Economics. Wu, Min, 2002, Earnings restatements: A capital market perspective, unpublished doctoral thesis, New York University. Wu, Min, 2004, Restatements by Year, 1977-2Q2002, Working paper, New York University.
StudySample
sizeTime
periodPrimary data
source* Event dateEvent
window
Average abnormal
stock return Significance
Karpoff and Lott (1993) 11 1978-87 WSJ Initial revelation 2 days -4.66% < .05
Davidson, Worrell, and Lee (1994) 34 1965-90 WSJ Initial revelation 3 days -2.80% < .01
Gerety and Lehn (1997) 37 1981-87 WSJ Initial revelation 3 days -3.05% < .10
Bernile and Jarrell (2009) 129 2004-07 WSJ Initial revelation 3 days -2.38% < .01
Tanimura and Okamoto (2010) 40 2000-08 Factiva Initial revelation 2 days -5.99% < .01
Anderson and Yohn (2002) 161 1997-99 10-K Wizard Initial revelation 7 days -3.49% < .05
Richardson, Tuna, and Wu (2002) 225 1971-90 Lexis-Nexis Restatement 251 days 25.00% no test
Palmrose, Richardson, and Scholz (2004) 403 1995-99 Media, SEC Restatement 2 days -9.20% < .01
Hribar and Jenkins (2004) 286 1997-02 GAO Restatement 5 days 9.00% no test
Akhigbe, Kudla, and Madura (2005) 542 1991-01 Lexis/Nexis Restatement 3 days -4.10% < .01
Agrawal and Chadha (2005) 119 2000-01 Lexis/Nexis Restatement 3 days -5.65% < .01
Desai, Hogan, and Wilkins (2006) 146 1997-02 GAO Restatement 3 days -11.07% no test
Burns and Kedia (2006) 215 1995-02 GAO Restatement 3 days -8.78% < .01
Arthaud-Day, Certo, Dalton, and Dalton (2006) 116 1997-02 GAO Restatement 3 days -11.00% < .01
Kravet and Shevlin (2010) 299 1997-02 GAO Restatement 3 days -5.40% < .01
Hennes, Leone, and Miller (2008) 69 1997-05 GAO Restatement - errors 3 days -1.93% < .01
Hennes, Leone, and Miller (2008) 101 1997-05 GAO Restatement - irregularities 3 days -13.64% < .01
Moore, Stuart, and Pozner (2010) 437 1995-06 GAO Restatement 3 days -4.11% < .01
Kellogg (1984) 42 1967-79 various Initial revelation 1 day -3.90% < .01
Francis, Philbrick, and Schipper (1994) 45 1988-92 SCAA Initial revelation 1 day -17.16% < .01
Ferris and Pritchard (2001) 89 1995-99 Stanford Initial revelation 3 days -24.99% < .01
Griffin, Grundfest, and Perino (2004) 2133 1990-01 ISS Initial revelation 3 days -16.60% < .01
Share price reactions to lawsuit filings:
Romano (1991) 13 1966-87 various Filing date 2 days -3.16% < .05
Bhagat, Bizjak, and Coles (1998) 46 1981-83 WSJ Filing date 2 days -2.71% < .01
Griffin, Grundfest, and Perino (2004) 2194 1990-01 ISS Filing date 3 days -4.10% < .01
Ferris and Pritchard (2001) 89 1995-99 Stanford Filing date 3 days -3.47% < .01
Gande and Lewis (2009) 605 1996-03 Stanford Filing date 3 days -4.66% < .01
Table 1 is continued on the following page.
Table 1: Financial misrepresentation, event study results
Share price reactions to public disclosures that prompt lawsuits:
Panel C: Studies based on securities class action lawsuits
This table summarizes the main event study results from 32 papers that investigate the share price reactions to announcements of financial misconduct. The results are organized by the source of data used to identify samples of misconduct.
Panel A: Studies that use media key word searches
Panel B: Studies based on earnings restatements
StudySample
sizeTime
periodPrimary data
source* Event dateEvent
window
Average abnormal
stock return Significance
Feroz, Park, and Pastena (1991) 58 1982-89 WSJ Initial revelation 2 days -12.90% < .01
Dechow, Sloan, and Sweeney (1996) 92 1982-92 AAER Initial revelation 1 day -8.80% < .01
Beneish (1999) 50 1987-93 AAER, media Initial revelation 3 days -20.80% < .01
Bonner, Palmrose, and Young (1998) 114 1982-95 AAER End of violation period 12 month -24.00% No test
Miller (2006) 60 1987-02 AAER, media First press mention 1 day -6.30% Not reported
Ozbas (2007) 75 1999-03 AAER First press mention 3 days -22.49% < .01
Ozbas (2007) 75 1999-03 AAER First AAER 3 days -2.88% < .05
Initiation of an SEC investigation: 45 1982-89 AAER SEC investigation 2 days -7.50% < .01
Settlement of an SEC action 33 1982-89 AAER Settlement 2 days -1.20% n.s.
Karpoff, Lee, and Martin (2008b) 371 1978-02 SEC, media Initial revelation 1 day -25.24% < .01
Karpoff and Lou (2010) 454 1988-05 SEC, media Initial revelation 1 day -18.20% < .01
*Abbreviations: WSJ - The Wall Street Journal; GAO - Government Accountability Office; SEC - Securities and Exchange Commission; AAER - Accounting and Auditing Enforcement Release (from the SEC); SCAA - Securities Class Action Alert; Stanford - Stanford Class Action Database; ISS - Institutional Shareholders Services, Inc.
Panel E: Studies based on SEC enforcement actions (KLM database)
Share price reactions to public disclosures that prompt AAERs:
Share price reactions to related regulatory events (from Feroz, Park, and Pastena, 1991):
Panel D: Studies based on Accounting and Auditing Enforcement Releases (AAERs)
Table 1: Financial misrepresentation, event study results, continued This table summarizes the main event study results from 25 papers that investigate the share price reactions to announcements of financial misconduct. The results are organized by the source of data used to identify samples of misconduct.
StudySample
sizeTime
period
Primary data
source*Event date or sample
informationEvent
window
Average abnormal
stock return Significance
Davidson and Worrell (1988) 96 1970-80 Fortune Initial revelation 1 day -3.48 < .01
Davidson, Worrell, and Lee (1994) 501 1965-90 WSJ Initial revelation 3 day 0.21% n.s.
Long and Rao (1995) 54 1989-92 WSJ Initial revelation 1 day -0.12% n.s.
Reichart, Lockett, and Rao (1996) 83 1980-90 WSJ Indictment date 2 days -1.91 < .01
Gunthorpe (1997) 69 1988-92 WSJ 1 day -1.38% < .01
Bhagat, Bizjak, and Coles (1998) 135 1981-83 WSJ Lawuit filing date 2 days -1.26% < .01
Alexander (1999) 17 1977-90 WSJ Initial revelation 2 days -0.04% ??
Block (1991) 32 1985-89 WSJ Initial revelation 2 days -2.23% < .05
Karpoff and Lott (1993) 57 1978-87 WSJ Initial revelation 2 days -1.22% < .05
Murphy, Shrieves, and Tibbs (2009) 175 1982-96 WSJ Initial revelation 2 days -2.30% < .01
Tanimura and Okamoto (2010) 52 2000-08 Factiva Initial revelation 2 days -5.40% < .01
Karpoff and Lott (1993) 40 1978-87 WSJ Initial revelation 2 days -1.67% < .05
Alexander (1999) 24 1977-90 WSJ Initial revelation 2 days -3.06% < .05
Karpoff, Lee, and. Vendrzyk (1999) 63 1983-95 MediaDefense fraud-
unranked contractors 2 days -4.46% < .01
Karpoff, Lee, and. Vendrzyk (1999) 186 1983-95 MediaDefense fraud -
ranked contractors 2 days -0.39% < .05
Tanimura and Okamoto (2010) 17 2000-08 Factiva Japanese firms 2 days -8.50% < .01
Table 2 is continued on the following page.
*Abbreviations: WSJ - The Wall Street Journal (including the Dow Jones News Retrieval System and Factiva); GAO - Government Accountability Office; SEC - Securities and Exchange Commission; AAER - Accounting and Auditing Enforcement Release (from the SEC); SCAA - Securities Class Action Alert; Stanford - Stanford Class Action Database; ISS - Institutional Shareholders Services, Inc.; CCH - Commerce Clearing House.
**These papers include, without distinguishing between, such events as fraud, bribery, breach of contract, securities fraud, patent infringement, environmental violations, price fixing, illegal campaign contributions, tax evasion, and OSHA violations
Table 2: Non-financial reporting types of business misconduct, event study results This table summarizes the results of research papers that measure the stock price reactions to news of different types of non-financial misconduct. Panel A includes studies that cluster different types of misconduct together. Panels B – O report on results for specific types of business misconduct.
Panel A: Illegal corporate activities (mixed categories)**
Panel B: Frauds of related parties
Panel C: Frauds of government
StudySample
sizeTime
period
Primary data
sourceEvent date or sample
informationEvent
window
Average abnormal
stock return Significance
Jarrell and Peltzman (1985) 116 1967-81 WSJ Auto recalls 11 days -1.60% < .01
Hoffer, Pruitt, and Reilly (1988) 66 1975-81 WSJ Auto recalls 11 days -1.29% < .05
Bromiley and Marcus (1989) 91 1967-83 WSJ Auto recalls 2 days -0.64% < .10
Barber and Darrough (1996) 573 1973-92 WSJ Auto recalls 2 days -0.36% < .01
Jarrell and Peltzman (1985) 26 1974-82 WSJ Drug recalls 10 days -6.13% < .01
Ahmed, Gardella, and Nanda (2002) 70 1966-98 WSJ Drug recalls 2 days -7.85 < .01
Pruitt and Peterson (1986) Product recalls -0.92% < .05
Davidson and Worrell (1992) 133 1968-87 WSJ Product recalls 3 days -0.68% < .01
Rubin, Murphy, and Jarrell (1988) 48 1977-81 Product recalls 81 days -6.90% not reported
Chalk (1987) 19 1966-81 NYTimes First news 5 days -3.77% < .01
Borenstein and Zimmerman (1988) 74 1960-85 WSJ First news 1 day -0.94% < .01
Mitchell and Maloney (1989) 56 1964-87 WSJ First news- pilot error 1 day -1.68% < .01
Mitchell and Maloney (1989) 56 1964-87 WSJFirst news - other
causes 1 day -0.41% n.s.
Mitchell (1989) 1 1982 WSJ Tylenol case 2 days -14.30% < .01
Dowdell, Govindaraj, and Jain (1992) 29 1982 WSJ Tylenol case 4 days -11.92% < .01
Bhagat, Bizjak, and Coles (1998) 38 1981-83 WSJLawsuit filing date-
product liability 2 days -1.46% < .01
Prince and Rubin (2002) 44 1985-95 WSJLawsuit filing date-
product liability 3 days -0.57% n.s.
Table 2 is continued on the following page.
Table 2: Non-financial reporting types of business misconduct, event study results, continued This table summarizes the results of research papers that measure the stock price reactions to news of different types of non-financial misconduct. Panel A includes studies that cluster different types of misconduct together. Panels B – O report on results for specific types of business misconduct.
Panel E: Air safety disasters
Panel F: Other product failures
Panel D: Product recalls
StudySample
sizeTime
period
Primary data
sourceEvent date or sample
informationEvent
window
Average abnormal
stock return Significance
Campbell, Gordon, Loeb, and Zhou (2003) 43 1995-00 Media All breaches 3 days -1.88% n.s.
Garg, Curtis, and Halper (2003) 22 1999-02 N/A All breaches 2 days -3.80% n.s.
Hovav & D'Arcy (2003) 23 1998-02 MediaDenial of service
attacks 2 days not reported n.s.
Cavusoglu, Mishra, and Raghunathan (2004) 66 1996-01CNet, others All breaches 2 days -2.09 < .01
Acquisti, Friedman, and Telang (2006) 79 2000-05Not
reported All breaches 2 days -0.58% < .05
Kannan, Ress, and Sridhar (2007) 72 1997-03 Media All breaches 4 days -0.65% n.s.
Campbell, Gordon, Loeb, and Zhou (2003) 11 1995-00 Media Confidential info only 3 days -5.46% < .01
Garg, Curtis, and Halper (2003) 4 1999-02 N/A Credit card info 2 days -9.00% yes
Kannan, Ress, and Sridhar (2007) 22 1997-03 Media Confidential info only 4 days 1.40% n.s.
Tanimura and Wehrly (2009) 152 2000-07 Attrition Lost confidential info 2 days -0.23 < .05
Muoghalu, Robison and Glascock (1990) 128 1977-86 WSJ Lawsuit filings 2 days -1.23% < .01
Muoghalu, Robison and Glascock (1990) 74 1977-87 WSJ Lawsuit filings 2 days -0.19% n.s.
Laplante and Lanoie (1994) 12 1982-91 Media Incident dates 1 day 0.03% n.s.
Laplante and Lanoie (1994) 9 1982-91 Media Lawsuit filings 1 day not reported n.s.
Laplante and Lanoie (1994) 13 1982-91 Media Settlements 1 day -0.94% n.s.
Hamilton (1995) 436 1989 EPAToxic release inventory
released 1 day -0.28% < .01
Klassen and McLaughlin (1996) 22 1985-91 Media Environmental crises 1 day -0.82% < .05
Jones and Rubin (2001) 98 1970-92 WSJ Initial revelation 2 days 0.17% n.s.
Bhagat, Bizjak, and Coles (1998) 27 1981-83 WSJ Lawsuit filings 2 days -3.08% < .01
Karpoff, Lott, Wehrly (2005) 478 1980-00 Factiva Initial revelation 2 days -1.00% < .01
Table 2 is continued on the following page.
Table 2: Non-financial reporting types of business misconduct, event study results, continued This table summarizes the results of research papers that measure the stock price reactions to news of different types of non-financial misconduct. Panel A includes studies that cluster different types of misconduct together. Panels B – O report on results for specific types of business misconduct.
Panel G: Private security breaches
Panel H: Environmental violations
StudySample
sizeTime
period
Primary data
sourceEvent date or sample
informationEvent
window
Average abnormal
stock return Significance
Block (1991) 32 1985-89 WSJ Initial revelation 2 days -0.32% n.s.
Karpoff and Lott (1993) 24 1978-87 WSJ Initial revelation 2 days -0.93% n.s.
Alexander (1999) 18 1977-90 WSJ Initial revelation 2 days 0.44% n.s.
Murphy, Shrieves, and Tibbs (2009) 217 1982-96 WSJ Initial revelation 2 days -0.80% < .01
Tanimura and Okamoto (2010) 34 2000-08 Factiva Initial revelation 2 days 3.80% < .01
Hersch (1991) 123 1964-86 WSJ EEO lawsuit decisions 4 days -1.56% < .01
Davidson, Worrell and Cheng (1994) 63 1979-89 MediaOSHA penalties -
announcement 3 days -0.56% < .05
Peltzman (1981) 23 1960-75 FTC False advertising 8 days -2.42% < .01
Beatty, Bunsis, and Hand (1998) 21 1980-93 WSJIPO underwriting
violations 1 day -16.50% < .01
Karpoff and Lott (1999) 351 1979-95 MediaInitial press
announcements 2 days -0.45% < .01
Bhagat, Bizjak, and Coles (1998) 62 1981-83 WSJ Lawsuit filing date 2 days -0.81% n.s.
Langus and Motta (2006) 41 1969-05 Public docs Dawn raid 11 days -2.40% < .05Van den Broek, Kemp, Verschoor, and de Vries (2010) 66 1998-08
NMA decisions Regulatory action date 3 days -2.30% < .01
Skantz, Cloninger, and Strickland (1990) 41 N/A N/A N/A 2 days -1.90% "yes"
Bosch and Eckard (1991) 127 1962-80 CCH indictment 2 days -1% <.01
Detre, Golub, and Connor (2005) 24 1981-01 CCH indictment 3 days -3.40% <.05
Table 2: Non-financial reporting types of business misconduct, event study results, continued This table summarizes the results of research papers that measure the stock price reactions to news of different types of non-financial misconduct. Panel A includes studies that cluster different types of misconduct together. Panels B – O report on results for specific types of business misconduct.
Panel K: False advertising
Panel L: IPO underwriting violations
(This result is for the primary clients of underwriters that are targeted by the SEC for investigation of misconduct.)
Panel M: Punitive damage awards
Panel N: Antitrust violations (not including price-fixing)
Panel O: Price-fixing
Panel I: Regulatory violations or frauds affecting unrelated parties
Panel J: Employee-related events (e.g., health and safety, discrimination)
Table 3 Reputational losses for different types of business misconduct
As inferred by the authors of papers listed in Table 2
Reputational losses are large for: Reputational losses are insignificant for:
• Financial misrepresentation • Private security breaches
• Frauds of related parties • Environmental violations
• Frauds of government (if the firm is not politically influential)
• Frauds of unrelated parties
• Product recalls
• Air safety disasters
• Other product failures
• Employee-related violations
• False advertising
• IPO underwriting violations
• Punitive damage awards
• Antitrust violations
Figure 1: Average market-adjusted share price effects for SEC enforcement actions for financial misrepresentation
(From Karpoff, Lee, and Martin 2008b)
SEC Investigation Initial SEC regulatory action
Resolution Filings and Reports with errors
Trigger (initial
revelation
DOWN 25.2%
DOWN 4.2% DOWN
9.6%
DOWN 13.7%
Initial public revelation
Figure2:Xerox’scumulatedmarketadjustedstockreturn,19972006This graph illustrates the sequence of events in a typical SEC enforcement action for financial misrepresentation. News of potential reporting problems (the trigger event) first became public on October 8, 1999, when Xerox announced that its third quarter 1999 earnings would fall short of projections. After a preliminary inquiry, the SEC initiated a formal investigation of Xerox’s financial reports on July 3, 2000. This was followed by several releases in which the SEC announced sanctions, ending with a concluding release on March 26, 2007. In its releases, the SEC identified the violation period – the period during which Xerox’s financial reports were in error, as beginning at the beginning of 1997.
Figure 4: Sources of loss for financial misrepresentation
This chart illustrates the breakdown of the total loss in market capitalization into components for legal penalties, share price deflation, and reputational loss, based on data for 384 cases of financial misrepresentation as reported in KLM (2008b).
Reputatationalloss67%
Sharede4lation24%
Legalpenalties9%
Figure 5: Numerical estimates of reputational losses as a fraction of the total loss in market capitalization for different types of misconduct
Sources: For financial misrepresentation – Beneish (1999), KLM (2008b), and Tanimura and Okamoto (2009); for frauds of related
parties – Karpoff and Lott (1993), Murphy, Shrieves, and Tibbs (2009); for product recalls – Jarrell and Peltzman (1985); for air
safety disasters – Mitchell and Maloney (1989); for other product failures – Prince and Rubin (2002); for private security breaches –
Tanimura and Wehrly (2009); for environmental violations – Karpoff, Lott, and Wehrly (2005), and Jones and Rubin (2001); for
unrelated party violations – Murphy, Shrieves, and Tibbs (2009); for employee-related violations – Hersch (1991); for punitive
damage awards – Karpoff and Lott (1999); for antitrust violations – Van den Brock et al. (2010).
70%56%
77%67%
46%
3% 0% 0%
67%
84%
46%
Financial
misrepresentation
Fraudsofrelated
parties
Productrecalls
Airplanecrashes
Otherproduct
failures
Privatesecurity
breaches
Environm
ental
violations
Unrelatedparty
violations
Employee‐related
violations
Punitivedamage
awards
Antitrustviolations
top related