The Changing Face of the US SEC’s Enforcement Investigations (AAERs) into Financial Statement Fraud in the Post-Enron Environment By Richard Lane# James Cook University Brendan T. O’Connell† James Cook University # School of Business, James Cook University, Townsville, Queensland, Australia 4811 † Professor, School of Business, James Cook University, Townsville, Queensland, Australia 4811 Corresponding Author: Brendan O’Connell 1
78
Embed
Securities and Exchange Commission (SEC). 2002. "SEC Sues ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
The Changing Face of the US SEC’s Enforcement
Investigations (AAERs) into Financial Statement Fraud in the
Post-Enron Environment
By
Richard Lane#James Cook University
Brendan T. O’Connell†James Cook University
# School of Business, James Cook University, Townsville, Queensland, Australia 4811
† Professor, School of Business, James Cook University, Townsville, Queensland, Australia 4811
relate to financial statement reporting. A random sample of 220 cases of financial
statement fraud were subjected to analysis, however, the researchers found that only in 99
cases were they able to obtain the “last clean financial statements” (p.15).
Major findings of the COSO Report were that the companies involved in financial
statement fraud tended to be relatively small; some companies committing the fraud were
experiencing net losses or were close to break-even prior to the fraud; top senior
executives were frequently involved; cumulative amounts of fraud were relatively large
relative to company size and were not isolated to a single period; and, typical fraud
techniques involved overstatement of revenues (COSO Report, 1999: 5-6).
Criticisms of AAERS and the COSO Report (1999)
There have been several studies utilising AAERs which have alluded to their
limitations (see, for example, Feroz et al., 1991; Bonner et al., 1998) and critiques of the
COSO Report (1999) itself (see, for example, Briloff, 2001; O’Connell, 2001). In terms
of the limitations of AAERs themselves, Feroz et al. (1991) report that SEC enforcement
actions differ in their nature and severity. Bonner et al. (1998) recognized this problem
and attempted to control for differences in severity by assigning each action a value for
severity from the external auditor’s perspective. Bonner et al. (1998) also emphasized the
strong possibility of selection bias emanating from reliance on AAERs. They specified
that these enforcement actions might reflect specific SEC agendas prevailing at the time
of the sample selection (p. 505). If this is the case, sole reliance on enforcement actions
might not produce a sample of frauds that is truly representative of the entire population
of financial statement frauds.
5
The SEC admits that its enforcement program is designed to “concentrate on
particular problems areas and to anticipate emerging problems” (SEC, 1989, p. 1). As an
example of this SEC agenda bias, Feroz et al. (1991) reported that 70% of all AAERs
issued between 1982 and 1989 related to alleged overstatement of accounts receivables
and inventories by firms. Furthermore, Feroz et al. (1991, p. 111-2) highlighted that
interviews with current and former SEC officials show that the SEC has more targets for
formal investigations than it can practically pursue and that they need to limit their
investigations to instances where material violations are alleged to have occurred. They
noted that formal investigations are both costly and highly visible and this means that the
SEC is forced to rank candidates for formal investigation “according to the probability of
success and potential message value” (Feroz et al., 1991, p. 112).
Turning to evaluation of the COSO Report (1999) itself, perhaps the most high
profile of these critiques was Briloff (2001). Key aspects of his criticism were that the
selection process produced a group of companies that were not representative of the high
profile cases that, in his view, were “contaminating the accounting and financial reporting
environment” (p.126). Briloff was especially critical of the obvious lack of high profile
cases in the sample given the small size of companies reflected in the demographic data
of the study. Briloff (2001) then produced his own study to support this claim. He
emphasized that the company with the highest total revenue almost equaled the total
revenues of the remaining 98 companies. He concluded that the sampling approach
created a “distorted image of the corporate enterprises in our financial reporting
environment.” (p.126).
6
O’Connell (2001) supported Briloff’s claim stating that “researchers who have
used AAERs also recognize the limitations of relying on AAERs to evaluate financial
statement fraud” (p.168). He highlighted that they differ in their nature and severity and
that they are uneven in their level of disclosure. Further, O’Connell (2001) criticized the
absence of details of individual cases in its sample. He stated that “this lack of financial
data on sample firms is clearly frustrating and a major limitation of the COSO Report”
(p.171).
O’Connell (2001) also evaluated the sampling approach of the COSO Report
(1999). He noted that one of the major conclusions of the COSO Report (1999) is that
“companies committing financial statement fraud were relatively small” (p 2). He argued
that this finding is only valid if one assumes AAERs are truly representative of all
financial statement fraud and if one assumes that the omitted firms are similar in
character to those 99 firms portrayed in the COSO Report (1999). O’Connell (2001)
observed that if one assumes both of these conditions are met then the population will be
skewed towards smaller firms. Moreover, when a population is skewed one way and a
random sample is taken from that population, then the resultant sample is likely to reflect
that particular distribution shape (in this case, a plethora of small firms and not too many
large ones).
O’Connell (2001) concluded that such a sampling outcome is of great concern in
this case as it is the high profile cases of financial statement fraud that make the front
pages of the Wall Street Journal and that if one is concerned about restoring public faith
in the accounting profession, then one must study and enhance understanding of what
7
drives these high profile cases. To enhance the usefulness of the COSO study, he
advocated a sampling approach that includes large, high-profile cases.
The SEC and Institutional Theory
Any study that utilizes AAERs as a primary data source cannot draw conclusions
from them without first considering the institutional environment in which these
enforcement releases are issued. Accordingly, we will draw on institutional theory to
inform our analysis.
Institutional theory, according to Scott (1987) and DiMaggio and Powell (1983),
proposes that many aspects of formal organizational structures, policies and procedures
result from prevailing societal attitudes and the views of important constituents (p.53.).
Organisations obey these rules and requirements, not just on efficiency grounds, but also
to enhance their legitimacy, resources, and survival capacities (Kondra & Hinings 1998).
Institutional pressures operate in conjunction with other forces such as competition to
effect ecological dynamics. Organisational behaviour is inextricably ingrained in a
vibrant system of interrelated economic, institutional, and ecological influences
(DiMaggio & Powell 1983).
Of relevance to the present study, Bealing et al. (1996) utilized institutional theory
to examine the historical development of the SEC and, in particular, the form, content and
rhetoric of its early regulatory actions as a case example of an organisation attempting to
justify its existence and role in the financial markets. They found “that in order for the
SEC as an organization to become legitimated…as part of the regulatory arena, it had to
8
take part in building up…a framework of social control applied to the accounting
profession as well as reporting entities” (p.335).
Bealing et al. (1996) found that the SEC’s legitimacy was considerably enhanced
when it began issuing enforcement releases. They concluded that by clearly stating the
appropriate audit and accounting procedures that should have been in place but were not,
the SEC established itself as an exemplar “of professional service within public
accounting” and that it was able to protect the investing public through ensuring “full and
fair disclosure to investors of all material facts” (p.335).
Using institutional theory, it would seem that following the considerable fallout
from the Enron and WorldCom accounting scandals of the early part of this decade, the
SEC would sharpen its focus in pursuing financial statement and accounting fraud.
Certainly the pronouncements of politicians (see, for example, Bush, 2002a;b;c) and the
passage of the Sarbanes-Oxley Act of 2002 reflect the prevailing environment for major
reform of corporate regulation and a tougher stand generally on unscrupulous behavior by
corporate executives. In such an environment, the SEC would not be immune from
pressure to increase the rate and intensity of its surveillance of company reporting. This
pressure is likely to manifest itself in the AAERs issued post-2001.
Research Hypotheses
As noted earlier, a major objective of this study is to identify what key
characteristics, if any, of the AAERs have changed post the period covered by the COSO
Report (1999). We hypothesize that the changed institutional environment following the
accounting scandals may have impacted on the level of AAERs issuance, the size of
9
companies subject to AAERS and the nature of the allegations. Specifically, we test the
following six hypotheses:
H1: There has been an increase in the issuance of AAERs by the SEC in the period after 2001, that is, post the US accounting scandals.
This first hypothesis is based on the hostile environment faced by the SEC and
other regulators in the wake of Enron, WorldCom and other accounting scandals (see, for
example, Rockness & Rockness, 2005; Bush, 2002a,b&c). Using institutional theory, one
would expect to see the SEC react to this changed environment by actively seeking out
new cases of possible fraud or risk losing its institutional legitimacy (Kondra & Hinings
1998). Notwithstanding any possible legitimacy concerns, the post-Enron fallout may act
as a signal to the SEC that its surveillance activities needed to be boosted.
H2: There has been an increase in the size of companies subject to issuance of AAERs by the SEC in the period after 2001, that is, post the US accounting scandals.
This second hypothesis reflects the “high profile” nature of the major accounting
scandals. The COSO Report (1999) concluded that the vast majority of frauds in its
sample involved small companies. Consistent with institutional theory, one would expect
to see that following the accounting scandals of 2001, the SEC would aggressively pursue
“high profile” cases and trumpet any prosecutions as vindication of its pivotal role in
ensuring reliable financial disclosures to investors. As noted by De Fond and Smith
(1992) “the SEC Enforcement Division chooses cases that enhance its stature as an
effective law enforcement agency” (p.144). It follows that larger companies are likely to
have been subject to AAERS since 2001. It may also be plausible that the Enron fallout
10
alerted the SEC to potentially greater instances of fraud in large companies. It follows
that the SEC’s enforcement activities of larger entities may have increased for this reason
post-Enron.
Examples of this increasingly aggressive approach toward “high profile” cases are
found in SEC press releases from 2002 onwards. For example, in a 2002 press release
relating to SEC investigations concerning the well publicised case of Tyco International
(SEC, 2002), the SEC Director of Enforcement is quoted as follows:
This enforcement action is the latest chapter in the Commission's ongoing investigation, together with the Manhattan District Attorney, of corruption and self-dealing at the highest levels of Tyco management … The Commission today, together with the criminal authorities, serves notice that misconduct by outside directors, as well as by company management, will not be countenanced (SEC Press Release Number 2002-177).
Similarly, an SEC Press Release in 2003 relating to another well publicized fraud,
Vivendi Universal, included the following quote from the Deputy Director of the
Commission's Division of Enforcement:
This case shows the Commission's ongoing commitment to enforcing the disclosure obligations of issuers and it shows our successful use of a new enforcement tool provided by the Sarbanes-Oxley Act (SEC Press Release Number 2003-184).
The following four hypotheses are directly derived from the findings of the COSO
Report (1999). Specifically, that report found that many companies committing the fraud
were experiencing net losses or were in close to break-even situations; most frauds were
not isolated to a single fiscal period and large relative to the company size; the frauds
tended to involve improper revenue recognition or overstatement of assets; and, very
senior executives were frequently involved (p.5-6).
11
H3: Companies subject to issuance of AAERs by the SEC are likely to exhibit poor financial condition in the period leading up to the period of alleged accounting fraud.
H4: The alleged accounting frauds pertaining to companies subject to issuance of AAERs by the SEC are likely to have occurred over multiple periods and to involve large amounts relative to the company size.
H5: The alleged accounting frauds pertaining to companies subject to issuance of AAERs by the SEC are likely to involve improper revenue recognition or overstatement of assets.
H6: The alleged accounting frauds pertaining to companies subject to issuance of AAERs by the SEC are likely to have been orchestrated by the most senior executives of companies, that is, the CEO and the CFO.
RESEARCH METHODOLOGY
The present investigators commenced the study by identifying those AAERs that
involved an alleged breach of Rule 10(b) – 5 of the Securities Exchange Act 1934 or
Section 17(a) of the Securities Act of 1933 or other Federal anti-fraud statutes. These are
the sections that represent the primary antifraud provisions relating to financial reporting.
Excluded from the analysis were restatements of financial reports due to errors or any
activities that did not result in a violation of federal antifraud statutes. This approach was
consistent with that of the COSO Report (1999).
The present investigators reviewed 870 AAERs for the four-year period between
January 2002 and December 2005. The search identified 350 AAERs that related to
fraudulent financial reporting between 2002 and 2005 (a period deliberately chosen to
reflect the post-Enron fallout). However, there were multiple AAERs for some cases
amongst the 350 identified. Where there were multiple AAERs relating to one company,
these were counted as one1. This reduced the number of companies to 330. A random
1 In some cases a separate AAER was issued for the company and for each individual executive involved in the accounting fraud.
12
sample of 55 was taken from this group of companies which compares to a final sample
of 99 cases for the COSO Report. Appendix One lists each of the 55 cases used in our
sample together with key characteristics of each case2. Table 1 below shows a
comparison of the sampling approach used in the present study to that of the COSO
Report (1999).
INSERT TABLE 1 ABOUT HERE
In addition to a random sample of all AAERs for the study period, the present
study has adopted the recommendation of O’Connell (2001) in using a sampling
approach which deliberately adds five large “high profile” cases of financial statement
fraud. The five cases examined are Waste Management, Qwest Communication, Tyco
International, HealthSouth Corporation and Adelphia Communications.
This approach avoids the sampling problem of the COSO Report (1999) which
led Briloff (2001) to conclude that it was not representative of the “really stinking stuff
which is contaminating the accounting and financial reporting environment” (p.126). As a
consequence of this sampling approach, high profile companies make up 8.3% of the total
sample. None of these cases were duplicated in the random sample. A detailed analysis of
the “high profile cases” is found later in this paper.
It should be noted that we deliberately did not include Enron and WorldCom in
the sample as we felt that these scandals had already been subject to a considerable
amount of prior research and analysis (see, for example, Benston & Hartgraves, 2002;
2 Briloff (2001) was also critical of the absence from the COSO Report (1999) of such a list.
13
O’Connell, 2004). Accordingly, we think that the facts and circumstances of these two
cases are quite well understood.
There are six key characteristics of AAERs that were specifically examined in this
study. All of these six were also studied in the COSO Report (1999): the size of the
company committing the alleged financial statement fraud; the financial condition of the
company in the period prior to the alleged fraud; which senior executives were involved
in the alleged fraud; the cumulative amount of alleged fraud compared to the size of the
company; whether the alleged fraud covered more than one fiscal period; and, the typical
financial statement fraud techniques involved.
In addition to this analysis of AAERs, we conducted a qualitative content analysis
of the AAERs to ascertain common key motives behind the financial statement frauds.
This latter analysis we believe adds considerably to the richness of our understanding of
the factors behind these phenomena and was not undertaken in the COSO Report (1999).
RESULTS AND DISCUSSION
SEC Issuance of AAERS
Our findings support hypothesis one which proposes that there has been an
increase in the issuance of AAERs by the SEC in the period after 2001, that is, post the
US accounting scandals. The COSO Report (1999) utilised all AAERS issued between
January 1987 and December 1997. The authors of that report stated the following:
We read over 800 AAERs, beginning with AAER#123 and ending with AAER#1004. From this process, we identified nearly 300 companies involved in alleged instances of fraudulent financial reporting (p.12).
14
The final sample of the COSO Report was then randomly selected from that group
of 300. If we compare these figures to those of the present investigation, there has been a
marked increase in issuance of AAERs in recent years. Our analysis reveals that over the
four year period between 2002 and 2005, a total of 871 AAERs were issued. The
breakdown is as follows: 212 in 2002, 239 in 2003, 220 in 2004 and 200 in 2005. When
we identified which of these AAERs alleged fraud violations related to Rule 10(b)-5 of
the 1934 Securities Exchange Act or Section 17(a) of the 1933 Securities Act, a direct
comparison with the COSO Report approach, we found that 351 had been issued during
the study period. The split is as follows: 95 in 2002, 121 in 2003, 68 in 2004 and 67 in
20053. It follows that the SEC issued AAERs at a far higher rate in the period 2002 to
2005 when compared to 1987-1997. In just a four-year period almost as many AAERs
were issued as for the entire 11-year period studied by the COSO Report (1999).
Furthermore, those AAERs that specifically related to financial statement fraud appeared
to have risen substantially since the 1990s. It should also be noted that many of these
were issued in the period immediately after the scandals i.e. 2002 and 2003.
These findings could be explained by two possibilities. First, the SEC seeking to
legitimize itself to stakeholders through a greater focus on possible fraudulent financial
reporting. While the SEC did receive an injection of funding from 2002 onwards to
combat corporate fraud (see, Bush, 2002b) the prevailing climate in the US following the
accounting scandals of 2000-2001 meant that the SEC needed to be seen to be vigilant in
this area. They may have acted as predicted by institutional theory. Second, the SEC
stepped up its enforcement activities post-Enron following a realization that financial
statement fraud was more prevalent than previously envisaged.
3 It should be noted that the COSO Report (1999) did not report a year-by-year dissection of AAERs.
15
Size of Companies Committing Financial Fraud
The COSO Report (1999) stated that the sample companies are relatively small in
size. The company size (total assets and revenue) comparisons between the COSO Report
(1999) and the present study are shown in Table 2.
INSERT TABLE 2 ABOUT HERE
The COSO Report (1999) stated that the “companies committing financial
statement fraud were relatively small” and that “most of the sample companies ranged
well below $100 million in total assets” (p.5). As depicted in Table 2, the COSO Report
(1999) found mean total assets of companies subject to AAERs were $533 million
(median of $15.7 million). This compares with $3.3 billion (median of $187 million).
These figures represent a 522% increase in the mean total assets and a 1,092% rise in the
median. Turning to total revenues, the COSO Report (1999) portrayed a mean of $233
million (median of $13 million). This compares with the present study’s equivalents of
$3.1 billion and $62.6 million respectively. These figures depict a 1,232% increase in the
mean total revenue and a 380% rise in the median.
It should be noted that we did not conduct statistical comparisons of the two
groups (present study sample versus COSO Report sample) relating to company size
because the authors of the COSO Report (1999) did not identify which specific AAERs
were included in their analysis. Their report provided only aggregated findings and hence
a valid statistical comparison such as a chi-square test is impossible4. Standard deviations
are also not reported because our sample (together with that of the COSO Report, 1999) 4 It should be noted that both Briloff and O’Connell personally contacted the authors of the COSO Report (1999) in 2001 in an effort to obtain details of the AAERs included in their sample but were informed that due to confidentiality issues with regard to the Treadway Commission they were not prepared to release this information to us.
16
is heavily positively skewed. Reliance on a standard deviation would be misleading in
such a skewed distribution and consequently, median and quartile results are provided
instead to provide a more meaningful picture of the empirical data (Tabachnick & Fidell,
2001).
Clearly, the companies within the COSO Report (1999) sample are considerably
smaller, on average, than those in the present study regardless of whether the samples are
compared on the basis of their means, medians or quartiles. Hence, hypothesis two which
states that there has been an increase in the size of companies subject to issuance of
AAERs by the SEC in the period after 2001 is supported by our results notwithstanding
the inability to conduct a statistical comparison as noted in the previous paragraph.
There are several possible conclusions that can be drawn from these findings.
First, that the COSO Report (1999) sampling approach was skewed towards smaller
companies. However, this seems unlikely as the COSO Report (1999) claims to have
randomly selected its sample so this would suggest that it was representative of the
population of the time. Second, that the SEC deliberately selected larger, more “high
profile” companies post-2001 for prosecution consistent with an institutional theory
argument or because they came to the realization that financial statement fraud was more
prevalent in large companies than previously thought. This may be quite plausible in light
of the evidence presented here and the pronouncements of the SEC mentioned earlier.
Third, that the frauds committed post-1998 were generally much larger than pre 1998.
This would seem unlikely. Fourth, potential frauds were less likely to be detected and/or
prevented prior to 2001 and were less likely to be prosecuted by the SEC especially
where larger companies were concerned. Again, this would seem improbable.
17
Profitability of Companies Committing Financial Fraud
Consistent with the COSO Report (1999), we examined the net income for the
year prior to commencement of the fraud. Our findings are displayed in Table 3 and show
a far higher net income figure for our sample (mean of $48.3 million, median of $1.8
million) when compared to the COSO Report (mean of $8.6 million, median of
$175,000). This represents a 463% increase in the mean and a 929% increase in the
median.
INSERT TABLE 3 ABOUT HERE
These findings contrast somewhat with the COSO Report (1999) which concluded
that “pressures of financial strain or distress may have provided incentives for fraudulent
activities for some fraud companies” (p.5). Companies in the present investigation were
generally profitable in the lead up period to the fraud so the incentive to commit fraud
appears to have come from other factors. In fact, only 14 out of 55 of our sample (25.5%)
reported a loss in the period immediately prior to the alleged fraud. These findings do not
support hypothesis three which stated that companies subject to issuance of AAERs by
the SEC are likely to exhibit poor financial condition in the period leading up to the
period of alleged accounting fraud.
Our qualitative analysis of AAERs provided valuable insights into why executives
in the sample resorted to financial statement fraud despite many of the sample being quite
profitable entities. It is apparent that pressure to achieve Wall Street analysts’ profit
18
forecasts was a recurring theme at the heart of many of these frauds and was commonly
mentioned in the AAERs:
RSA stated that it had achieved analysts’ earnings expectations for the first quarter of 2001 … without the accounting change [a new, aggressive method of recognizing sales for shipments to distributors] RSA would have failed to meet analysts’ earnings expectations by approximately $0.02 per share or 12.5%, and its operating income would have been 17.3% lower (AAER No. 1817, RSA Security, July 23, 2003).
In the Spring of 1999, Miller [former CEO] devised and implemented a scheme to fraudulently overstate the company’s net income to meet analysts’ expectations. Pursuant to the plan, the company fraudulently reclassified rent and salary expenses that Master Graphics had already paid to its division presidents in the first quarter to assets on the company’s balance sheet thus reducing expenses and increasing income (AAER No. 2035, Master Graphics, June 14, 2004).
Stockholders’ Equity (Deficit) of Companies Committing Financial Fraud
Our analysis of stockholders’ equity in the period preceding the financial fraud
reveals a similar picture to that of the profitability analysis. As shown in Table 4, the
COSO Report (1999) reported a mean Stockholders’ Equity of $86.1 million (median of
$5 million). Our investigation shows a mean stockholders’ equity of $735.1 million
(median of $58.2 million). These findings reflect a 753% higher mean and a 1,062%
larger median for the sample in the present study. Again, hypothesis three is not
supported by these findings.
INSERT TABLE 4 ABOUT HERE
19
Executives Named in the AAERs
Table 5 depicts that there is a marked difference in the results obtained in the
present study and those found in the COSO Report (1999). The COSO Report (1999)
found that the Chief Executive Officer (141 companies or 72% of the sample) was the
highest ranking executive involved in most fraud cases5. In the present investigation, the
senior executive most frequently named was the Chief Financial Officer (33 companies
representing 60% of the sample). There also seems to be a greater spread of individuals
cited in this study. The Controller (13% of sample), Chief Operating Officer (16%) and
other Vice President Positions (20%) were commonly mentioned in the AAERs.
However, it should be noted that the companies in our sample are, on average, much
larger than that of the COSO Report (1999). Hence, this greater spread of accused may
simply reflect the greater size and complexity of the companies in our study. Our findings
generally support hypothesis six which states that the frauds are likely to have been
orchestrated by the most senior executives of companies, that is, the CEO and the CFO.
As Table 5 shows this is clearly the situation in most cases.
INSERT TABLE 5 ABOUT HERE
Our qualitative analysis of AAERs offers important insights into the direct
involvement of unscrupulous senior executives in many of the frauds:
Signal Tech’s former management created a corporate atmosphere which encouraged managers to engage in improper accounting in an effort to improve Signal Tech’s bottom line. Signal Tech’s former chairman and other senior officers expressed scorn for accounting principles. … As a
5 It should be noted that in determining Table 5, the highest managerial title for an individual was used.
20
result of the pressure for profits at all costs and disdain for accounting principles by former senior officers … Keltec’s controller knowingly allowed improper accounting practices … Moreover, Signal Tech’s senior officers personally directed specific improper accounting practices and dictated misleading entries to be made on Keltec’s books to overstate revenue and understate costs (AAER No. 1534, Signal Technology, March 27, 2002).
During the first and second quarters of 2000, the CFO placed substantial pressure on the Controller to continue to meet analysts’ earnings expectations in reported results. As a result of that pressure, the Controller intentionally failed to record certain operating expenses which were material to both quarters (AAER No. 1691, Mercator Software, December 16, 2002).
Cumulative Dollar Amount of Fraud for a Company
As shown in the Table 6 the average fraud amount in our sample involved $137
million of cumulative misstatement or misappropriation (median of $10.8 million)6. This
is much higher than those found in the COSO Report (1999) where a mean and median of
$25 million and $4.1 million respectively were reported. This comparison indicates a
448% increase in the mean size of the fraud and a 163% rise in the median. Moreover, the
cumulative amounts of the fraud were quite high relative to the size of the company. The
median fraud of $4.1 million for the COSO Report (1999) sample represented 25% of
median total assets ($15.7 million). In our sample, the median fraud of $10.8 million
represented 6% of median total assets ($186.9 million). In both samples, the median fraud
amount considerably exceeded the median net income. It follows that hypothesis four is
supported as the frauds do, on average, involve large amounts relative to company size.
6 It should be noted that on a few occasions the AAERs did not fully disclose the quantum of dollars relating to the fraud. Accordingly, we conducted additional research through the SEC’s Litigation Releases et al, [found at http://www.sec.gov/litigation/litreleases.shtml] to ascertain the precise fraud amount.
21
INSERT TABLE 6 ABOUT HERE
Qualitative analysis of the AAERs shows a recurring theme that executives
perceived that their manipulations of reported earnings would have a significant impact
on the share price and their associated bonuses:
The compliant alleges that Gless [former CFO] and other Peregrine senior officers engaged in deceptive practices to artificially inflate Peregrine’s revenue and stock price, and that Gless then took fraudulent action to conceal the scheme (AAER No. 1759, Peregrine Systems, April 16, 2003).
In 2001 and 2002 Huntington reported inflated earnings in its financial statements, enabling Huntington to meet or exceed Wall Street analyst earnings per share expectations and internal EPS targets that determined bonuses for senior management (AAER No. 2251, Huntington Bancshares, June 2, 2005).
Length of Fraud Period
The COSO Report (1999) found that the average fraud was for two years with a
median of 21 months (p.30). Table 7 indicates that 36% of frauds in the current study
related to one year or less with 36% between one and two years, and 18% lasting for
three years. In all, 64% of all frauds in our sample were for greater than one year. These
results are consistent with hypothesis four in indicating that many frauds occur across
multiple periods.
INSERT TABLE 7 ABOUT HERE
Common Financial Fraud Techniques
As can be seen in Table 8, by far the most common technique used to fraudulently
misstate financial statement information in our sample involved deliberate overstatement
22
of revenue. In fact, 96% of alleged fraud involved this method. In the COSO Report
(1999), 50% of the sample employed this method suggesting either the SEC has
deliberately increased its surveillance activities in this area or that this type of fraud has
become far more prevalent since the 1990s. The second most common technique to
fraudulently misstate financial statement information was the overstatement of assets
(53%). However, the use overstatement of assets and understatement of
expenses/liabilities have remained relatively consistent across the two periods.
Misappropriation of assets has declined somewhat across the two periods (down from
40% in the COSO Report to 11% in the present study). The above findings support
hypothesis five in that the sample frauds commonly involved improper revenue
recognition or overstatement of assets.
INSERT TABLE 8 ABOUT HERE
Our qualitative analysis of AAERs provided detailed descriptions of the types of
techniques employed by executives to misrepresent the company’s financial performance.
The deliberate overstatement of revenue stands out as the preferred option for accounting
fraud:
[Senior executives] falsely reported millions of dollars of non-existent sales, including sales to a fictitious customer, and used other fraudulent techniques to inflate Anicom’s revenues” (AAER No. 1554, Anicom, May 6, 2002).
During the period from January 1996 through June 1998, Signal Tech’s Keltec division prematurely recognized revenue, failed to record contract losses and failed to write down excess inventory. Signal Tech’s former Chairman and CEO, former CFO, and other former senior corporate officers, knew of and in multiple instances personally directed improper
23
accounting practices. On several occasions the former chairman and other senior officers even dictated specific misleading entries to be made on Keltec’s books (AAER No. 1534, Signal Technology, December 16, 2002).
ANALYSIS OF FIVE HIGH PROFILE COMPANIES
Two of the main criticisms of the COSO Report (1999) by Briloff (2001) and
O’Connell (2001) are a paucity of “high profile” cases in that sample and the lack of
disclosure of the specific AAERs scrutinized by those investigators. In contrast, the
present study provides details of all AAERs examined (see Appendix One) and uses a
sampling approach that specifically includes an analysis of five high profile cases during
the study period. The five frauds examined are Waste Management, Qwest
Communication, Tyco International, HealthSouth Corporation and Adelphia
Communications.
Table 9 displays a summary of the key characteristics of the five major frauds.
Appendix Two provides an overview of each of these five frauds.
INSERT TABLE 9 ABOUT HERE
Table 9 shows that these five companies were large varying from a minimum of
$1.6 billion in total assets through to a maximum of $8.1 billion (mean of $4.6 billion).
The accumulated amount of the frauds was also sizeable varying from a minimum of
$567 million through to a maximum of $3.5 billion (mean of $1.9 billion). It should also
be noted that these fraud amounts were significant when compared to each company’s
revenue or profit situation. For example, in 60% of the cases the accumulated fraud
amount exceeded the company’s total revenue from the previous year’s financial result.
In all cases the frauds covered several reporting periods with a minimum of 4 years and a
24
maximum of 7.5 years (mean of 5.5 years). The CEO was implicated in the fraud in all
cases with the CFO prosecuted in 80% of them. In 40% of the cases, the company was
quite profitable in the year preceding the fraud (as measured by net income). In another
40%, the company reported a large loss the previous year and in the remaining case the
company reported a small loss. Artificial inflation of reported earnings was at the centre
of the prosecution in all but one of the cases.
Our qualitative analyses of the AAERs show that an inappropriate “tone at the
top” was a key issue in most of these cases:
The business unit executives made it clear that subordinates had to meet or exceed these aggressive earnings targets at all costs (AAER No 2209, Qwest Communications International, issued March 15, 2005).
This is a looting case …it involves egregious, self-serving and clandestine misconduct by the three most senior executives at Tyco (AAER No 1839, Tyco International, issued August 13, 2003).
The ubiquitous issue of pressure to appease analysts’ forecasts also was a
common theme:
The complaint was that these eight officers artificially accelerated Qwest’s recognition of revenue in two equipment sale transactions for its Global Business Markets Unit It is alleged that that when the Qwest financial management indicated the company would miss its quarterly targets, the Global Business Markets officers would bridge the gap by fraudulently mischaracterizing these transactions (AAER No 1726, Qwest Communications, issued February 25, 2003).
In summary, the findings from these five high-profile cases largely support those
of the 55 randomly selected cases. The extent and nature of the frauds was significant and
the companies involved were substantial in size.
CONCLUSION
25
This study set out to examine the U.S. SEC’s investigations into financial
statement fraud through an analysis of AAERs from 2002 to 2005. Findings were
compared to the COSO Report which examined AAERs from 1987 to 1997. Using an
institutional theory framework it was hypothesized that the post-Enron environment may
have brought about changes in the activities of the SEC in an effort for this agency to
legitimize itself before major stakeholders. Our study did find evidence of changes when
compared to the COSO Report (1999). Specifically, we found that there far more AAERs
issued post-Enron and the companies involved were, on average, much larger, more
profitable and the frauds more substantial than those exhibited in the COSO Report
(1999). These findings hold even before considering the additional five high profile cases
that we added to accommodate the recommendation of Briloff (2001). Our results suggest
that the SEC became more aggressive at pursuing larger companies for financial
statement fraud in the post-Enron environment than perhaps was the case in the 1990s.
Our study has also addressed a major criticism of the COSO Report (1999) by
Briloff (2001) and O’Connell (2001), namely, that there were an insufficient number of
“high profile” companies included in the sample of the COSO Report (1999) . We believe
that the addition of case studies on five major accounting frauds to 55 randomly selected
cases addresses those criticisms of the COSO Report (1999).
There are several important implications of this research. First, while the COSO
Report (1999: 5) concluded companies committing financial statement fraud tend to be
small this certainly does not appear to be the case based on our more recent sample. In
fact, it would seem that accounting frauds often involve large companies with the most
senior executives implicated. These frauds often cross several reporting periods. A
26
second key implication is that, rather than being motivated to hide losses, the frauds
appear to be a mechanism for boosting reported earnings to appease Wall Street analysts.
The motivation for this appears to be ensuring a rising stock price and thus increased
executive remuneration. This evidence from the AAERs would seem to show that devices
designed to reduce agency costs such as stock options have provided a perverse incentive
for some executives to do whatever it takes, including fraud, to boost the stock price in
the short-term. A third implication is that improper revenue recognition through
recording fictitious revenues or recording revenues prematurely continues to be the
preferred methods for deception. Auditors and analysts alike need to remain vigilant in
searching for any signs of such reprehensible actions by management.
There are some limitations to the present study that should be recognized. First,
AAERs tend to be uneven in their level of disclosure and format thus making some more
useful for analysis purposes than others. It is apparent from both the COSO Report
(1999), and the present study, that to enhance the usefulness of AAERs for analysis
purposes, the SEC needs to improve their comparability in relation to language, structure
and content. Second, there is a possibility of selection bias by relying on AAERs as the
primary data source for financial statement fraud. This is because AAERs may reflect the
prevailing agenda of the SEC (Bonner et al., 1998) or differ in their nature or severity
(Feroz et al., 1991). Notwithstanding these limitations AAERs have been widely applied
by researchers as a reasonably objective and reliable data source for studying fraud (see,
for example, De Fond & Smith, 1991; Bonner, Palmrose, & Young, 1998). Third, the
observed rise in enforcement actions and the size of firms prosecuted may be explained
by a variety of factors other than legitimization such as a greater awareness of the
27
possibility of fraud post-Enron. Fourth, if one accepts that the COSO Report (1999)
possessed weaknesses in its sampling approach then it may not be an accurate source of
comparison for studying changes in the SEC’s enforcement activities over the period.
Avenues for future research include more case-based research into financial
statement fraud using a variety of sources such as company documents, court cases,
judgments and interviews. It would also be useful to conduct interviews with surveillance
personnel of the SEC to ascertain information about prevailing SEC agendas and
processes and their specific impact on AAERs.
28
APPENDIX ONESUMMARY OF THE ACCOUNTING AND AUDITING ENFORCEMENT
RELEASES (AAERs) n=55 INCLUDED IN THE SAMPLEAAER No.
Date Issued
Company name Fraud Technique Fraud Amount
Fraud Period
Executives Involved
1489 02.01.02 InaCom Corp Inflating income by various methods
No Title Given 30 15% 1 2%Other Titles 24 12% 6 11%
42
Table 6: Cumulative Fraud Amount for Sample Firms in the COSO Report (1999) and the Present Study
Cumulative Fraud Amount(in $ 000’s)
% Change
COSO Report (1999)
This Study
n=99 n=55Mean $25.0 $137.0 448.0Median $4.1 $10.8 163.41st Quartile $1.6 million $3.95 million 146.93rd Quartile $11.76 million $53.6 million 355.8Smallest Fraud $20 $50 150Largest Fraud $910,000 $3,000,000 229.7Note: Figures were taken from AAERs.
Table 7: Number of Financial Periods (Years) Covered by Fraud for Sample Firms in the Present Study
Financial Periods (Years) Covered by Fraud Number of Sample
% of Total Sample
One Year 20 36Two Years 20 36Three Years 10 18Four Years 2 4Five Years 3 6Note: Figures were taken from AAERs. n=55 companies. The COSO Report (1999) did not provide data on this variable broken down as above. However, The COSO Report (1999) reported that 14% of the frauds were for one year or less with an average period of approximately two years (p.30).
43
Table 8: Common Financial Statement Fraud Techniques used by Sample Firms in the COSO Report (1999) and the Present
Study
COSO Report (1999) Present StudyMethods Used to Misstate
Financial StatementsPercentage of Sample Using Fraud Method
Overstatement of Assets (excluding accounts receivable overstatements due to revenue fraud)
50% 53%
- Overstating Existing Assets 37% 35%- Recording fictitious Assets or Assets not owned
12% 7%
- Capitalizing items that should be expenses
6% 11%
Understatement of Expenses/Liabilities
18% 24%
Misappropriation of Assets andOther Miscellaneous Techniques
40% 11%
Note: Figures were taken from AAERs. COSO Report (1999) n =204; Present Study n = 55.
Subcategories such as “Recording Fictitious Revenues” and “Capitalizing Expenses” do not sum to the overall category totals due to multiple types of fraud involved at a single company. So, for example, a company could have recorded revenues both fictitiously and recorded prematurely.
44
Table 9: Summary of Key Characteristics of Five Major Frauds during the Study PeriodCompany Size Profitability S/holders
2,300 4 years Hiding of liabilities off-balance sheet, inflating earnings. theft of company funds.
45
REFERENCES & BIBLIOGRAPHY
Bailey,J. 2000. “Waste Management Settles SEC Charges That It Misled Investors in Spring of ’99,” Wall Street Journal. (Eastern Edition) New York, N.Y., June 22, pg.B.20, Retrieved on 21 June, 2006 from http://proquest.uni.com/pqdweb?index=105&sid=1&srchmode=1&vinst=PROD&fm
Bealing, W.E. JR, Dirsmith, M.W and Fogarty, J., 1996. “Early Regulatory Actions by the SEC: Institutional Theory Perspective On The Dramaturgy of Political Exchanges,” Accounting Organizations and Society, Vol 21, No 4, pp 317-338.
Beasley, M.S., Carcello , J.V., Hermanson, D.R., Lapides, P.D. 2000. “Fraudulent financial reporting: Consideration of industry traits and corporate governance mechanisms”, Accounting Horizons, December,Vol.14, Iss. 4: 441-55.
Benston, G. J. and Hartgraves, A. L. 2002. “Enron: what happened and what we can learn from it”, Journal of Accounting and Public Policy, Vol. 21, Iss. 2, Summer: 105-127.
Berman, D.K. and Solomon, D., 2003. “Qwest May Settle SEC Swaps Case – Potential Charges of Fraud Relate to Fiber-Optic Deals With Other Telecom Firms” Wall Street Journal (Eastern edition) New York, N.Y., May 9, pg.A.3. Retrieved on 21 June, 2006 from: http://proquest,uni.com.elibrary.jcu.edu.au/pqdweb?index=1&sid=4&srchmode=1&v
Bonner,S.E., Palmrose,Z.V.& Young,S.M., 1998. “Fraud Type and Auditor Litigation: An Analysis of SEC Accounting and Auditing Enforcement Releases,” The Accounting Review, Vol 73, No.4, October: pp 503-532.
Briloff, A.J., 2001. “Garbage In/Garbage Out: A Critique of Fraudulent Financial Reporting: 1987-1997 (The COSO Report, 1999) The SEC Accounting Regulatory Process (AAERs) Critical Perspectives on Accounting, pp 125-148.
Bush, G.W. 2002a. “President Outlines Plan to Improve Corporate Responsibility” 7 March, White House Press Release: retrieved on 12 June 2006 from http://www.whitehouse.gov/news/releases/2002/03/print/20020307-3.html
Bush, G.W. 2002b. “President Continues Fight Against Corporate Fraud and Abuse,” 26 September, White House Press Release: retrieved on 12 June 2006 from: http://www.whitehouse.gov/news/releases/2006/09/print/20020926-10.html
Bush, G.W. 2002c. “President Bush Signs Corporate Corruption Bill,” 30 July, White House Press Release: retrieved on 12 June 2006 from: http://www.whitehouse.gov/news/releases/2006/09/print/20020926-10.html
COSO Report (Research Commissioned by the Committee of Sponsoring Organizations of the Threadway Commission). 1999. in M.S. Beasley, J.V.Carcello & D.R. Hermanson (eds), Fraudulent Financial Reporting: 1987-1997 An Analysis of US Public Companies (Jersey City, N.J.: Treadway Commission).
De Fond,M., and Smith, D.B., 1991. “Discussion of the Financial Effects of the SEC Accounting & Auditing Enforcement Releases” Journal of Accounting Research Vol 29 Supplement: pp143-148.
DiMaggio, P. J. and Powell, W.W., 1983, “The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields,” American Sociological Review 48: 113-123.
Feroz, E.H., Park,K. & Pastena,V.S., 1991. “The Financial and Market Effects of the SEC’s Accounting and Enforcement Releases, “Journal of Accounting Research, Vol 29, Supplement: pp 107-142.
Kondra, A. Z. and Hinings, C.R., 1998. “Organizational Diversity and Change in Institutional Theory”, Organization Studies, 19, 5: 743-767.
O’Connell,B., 2001. “An Analysis of Key Issues Arising From Professor A.J. Briloff’s Paper Entitled, “Garbage In/Garbage Out: A Critique of Fraudulent Financial Reporting: 1987-1997 (The COSO Report, 1999) The SEC Accounting Regulatory Process (AAERs)” Critical Perspectives on Accounting (2001) 12, 167-186.
O’Connell,B., 2004 “Enron Inc: He That Filches From Me My Good Name…Makes me Poor Indeed,” Critical Perspectives on Accounting, 15,6-7: 733-49.
O’Connell.B., Webb,L., and Schwarzback,H., 2005. “Batten Down the Hatches! U.S. Accounting Scandals and Lessons for Australia,” Australian Accounting Review Vol 15, No 2, 2005.
Perell, P. M. 2006. “Deceived with Ornament: Law, Lawyers and Shakespeare’s The Merchant of Venice” Downloaded on 17 January, 2007 from: http://www.lsuc.on.ca/media/third_colloquium_paul_perell.pdf
Press Release, 8 July 2005. “Federal Court Approves Dynergy Securities Fraud Settlement,” retrieved on 25 May 2006 from http://www.universityofcalifornia.edu/news/2005/jul08C.html
Rockness, H. and Rockness, J. 2005. “Legislated Ethics: From Enron to Sarbanes-Oxley, the Impact on Corporate America”, Journal of Business Ethics, Vol. 57: 31-54.
Scott, W. R., 1987. “The Adolescence of Institutional Theory,” Administrative Science Quarterly 32: 493-511.
Securities and Exchange Commission (SEC). 1989. Fifty-fifth Annual Report of the SEC (Washington DC: Government Printing Office).
Securities and Exchange Commission (SEC). 2002. “SEC Sues Former Tyco Director and Chairman of Compensation Committee Frank E. Walsh Jr. for Hiding $20 Million PaymentFrom Shareholders”, Press Release No. 2002-177, Retrieved on 18 January 2007 from: http://www.sec.gov/news/press/pressarchive/2002press.shtml
Securities and Exchange Commission (SEC). 2003. “Commission Settles Civil Fraud Action Against Vivendi Universal, S.A., Its Former CEO, Jean-Marie Messier, and Its Former CFO, Guillaume Hannezo”, Press Release No. 2003-184, Retrieved on 18 January 2007 from: http://www.sec.gov/news/press/pressarchive/2003press.shtml
Weld, L.G., Bergevin,P.M., and Magrath,L. 2002. “Anatomy of a Financial Fraud, a Forensic Examination of HealthSouth,” The CPA Journal, August 2002, retrieved on 25 May 2006 http://www.nysscpa.org/printversions/cpaj/2004/1004/p44.htm
Wells, J. T. 2001. “Follow fraud to the likely perpetrator”, Journal of Accountancy, New York: March, Vol. 191 ,Iss. 3: 91-4.