Given the findings presented by Desir et al. 2010, what evidence do they provide that might explain why there were so many people who opposed the proposed standard to enhance disclosures on expected losses?

Accounting Horizons Vol. 24, No. 4 2010 pp. 525-545 American Accounting Association DOI: 10.2308/acch.2010.24.4.525 Are Revisions to SFAS No. 5 Needed? SYNOPSIS: In the Financial Accounting Standards Board's FASB project, \"Disclosure of Certain Loss Contingencies,\" a central issue underlying the debate is whether existing implementation of FASB Accounting Standards Codification Topic 450-20 previously Statement of Financial Accounting Standards No. 5 provides sufficient and timely information to financial statement users. The Exposure Draft explains that constituents' assertions of inadequate disclosures are the primary motive underlying the FASB's re-examination of this issue see, for example, page v of the Exposure Draft. However, little actual data are available to indicate the extent of the alleged problem. This manuscript presents the results of a study undertaken to provide such data. For a sample of litigation-related losses, we find a surprisingly large incidence of non-disclosure of contingent losses that cannot be readily explained. Moreover, even where there is disclosure, we find many cases where firms do not provide estimates of expected losses, presumably under the permitted exception for cases where firms claim to be unable to estimate the magnitude of expected losses. On the other hand, we find relatively frequent disclosure of the items called for in the Exposure Draft, consistent with the conjecture that at least some of these items are being demanded by users. INTRODUCTION etting financial reporting standards, like most regulatory activities, requires decision making under uncertainty. In a typical standard-setting issue, standard setters receive information typically in the form of often-unsupported assertionsthat there is some deficiency in existing Generally Accepted Accounting Principles. For those alleged deficiencies that exceed a threshold determined by the Chair of the Financial Accounting Standards Board FASB in consultation with her/his fellow Board members, the Board and Staff of the FASB initiate a standardsetting project. In their work on projects, Board and Staff members must make difficult judgments about the nature of guidance likely to resolve the issue at hand and must anticipate any indirect consequences of the contemplated new guidance. S Rosemond Desir is an Assistant Professor at Colorado State University, Kirsten Fanning is an Assistant Professor at Villanova University, and Ray J. Pfeiffer is a Professor at Texas Christian University. This project was undertaken while Ray Pfeiffer was the FASB Research Fellow. The authors are grateful to the members of the Financial Accounting Standards Board's Statement of Financial Accounting Standards No. 5 team, David Elsbree in particular, for helpful comments and suggestions. We also acknowledge helpful comments from Lisa Bryant-Kutcher, Rick Cazier, Renee Hall, In-Mu Haw, Suzanne Lowensohn, Elizabeth Plummer, Mary Stanford, Bob Vigeland, Bill Wempe, the two anonymous referees, and Christine Botosan, Associate Editor. Submitted: October 2009 Accepted: May 2010 Published Online: December 2010 Corresponding author: Ray J. Pfeiffer Email: r.pfeiffer@tcu.edu 525 Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 Rosemond Desir, Kirsten Fanning, and Ray J. Pfeiffer 526 Desir, Fanning, and Pfeiffer Investors and other users of financial information have expressed concerns that disclosures about loss contingencies under the existing guidance in FASB Statement No. 5, Accounting for Contingencies, do not provide adequate information to assist users of financial statements in assessing the likelihood, timing, and amount of future cash flows associated with loss contingencies FASB 2008a, page v. In particular, the Exposure Draft cited the following as shortcomings of the existing guidance and application thereof: 1 2 3 See, for example, the recent report of the American Accounting Association Task Force Moehrle et al. 2009. Such work is in the spirit of the call issued by the SEC's Advisory Committee on Improvements to Financial Reporting see SEC 2008. In addition to the SEC's Staff Accounting Bulletin No. 92 that deals specifically with environmental liabilities which we did not focus on in this study, the SEC's guidance regarding the MD&Aspecifically Securities Act of 1933 17, CFR 229.103requires five disclosures, including the name of the court where the litigation is taking place, the date it was instituted, the principal parties involved, the factual basis alleged for the suit, and the relief sought by the plaintiff. CFR 229.303 also requires more generally that, for \"material contingencies,\" the registrant \"shall provide such other information that the registrant believes to be necessary to an understanding of its financial condition, changes in financial condition, and results of operations.\" See also 229.303a4iB. While we do not report on the extent of apparent compliance with these disclosure rules, we code MD&A disclosures as part of our data gathering and are surprised to find relatively little evidence that these disclosure rules are being followed explicitly by firms. Accounting Horizons American Accounting Association December 2010 Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 Empirical academic research has the potential to play a useful role in the standard-setting process to the extent that it can be used to reduce the inherent uncertainty facing Board members.1 Of course, this can happen both before or after new guidance is issued. Work of the latter type is more readily done, as changes in guidance create data that lend themselves naturally to pre-post analyses.2 Ex ante empirical evidence is more desirable because it enhances the likelihood that new guidance is actually warranted based on the data and that any new guidance achieves its intended purpose of improving the information available to financial statement users. This paper reports the findings of an example of such ex ante empirical research. The issue in question is the allegation that firms tend to provide insufficient and insufficiently timely disclosures of contingent legal obligations. This assertion, made to the FASB by various constituents in the form of anecdotal examples, was an explicit part of the rationale underlying the FASB's Exposure Draft, Disclosure of Loss Contingencies FASB 2008b. This paper has two intended contributions: 1 to report the findings of empirical research about the extent of the alleged insufficient disclosure problem; and 2 to serve as an example of how empirical academic research can be informative to the FASB Board and Staff. Background Accounting Standards Codification ASC 450-20-50 previously known as Statement of Financial Accounting Standards No. 5 FASB 2008a, which was originally issued in March 1975 provides guidance regarding the recognition and disclosure of loss contingencies. In brief, it specifies that preparers must evaluate the likelihood that a loss contingency will result in an unfavorable future outcome. The likelihood for a given contingency is classified as either \"probable,\" \"reasonably possible,\" or \"remote.\" Contingencies judged as \"probable\" and that are reasonably estimable by preparers are recognized in the financial statements, and key information about the contingency is disclosed in the notes to the financial statements. Contingencies judged as \"reasonably possible\" do not require recognition, but do require disclosure. \"Remote\" contingencies are neither required to be recognized nor disclosed.3 In September 2007, the FASB added a project to its agenda to address certain nonfinancial liabilities and contingencies. Over time, this project was narrowed in scope to focus specifically on litigation-related contingencies. On June 5, 2008, the FASB issued an Exposure Draft containing proposed changes to disclosures in the scope of FASB ASC Section 450-20-50. That document explicitly stated the rationale for the changes as follows: Are Revisions to SFAS No. 5 Needed? 527 EXHIBIT 1 Disclosure Item (by paragraph in the Exposure Draft) % Provided 1. The amount of the total claim or estimate of maximum loss exposure 7a 2. A description of the contingency 7b 3. An explanation of how the contingency arose 7b 4. The legal or contractual basis of the claim 7b 5. The current status of the contingency 7b 6. The anticipated timing of its resolution 7b 7. Identification of factors likely to influence the outcome 7b 8. Significant assumptions made in estimating the outcome 7b 9. Description of insurance or indemnification against any loss 7c 10. Information about changes in recognized contingencies during the period 8 Accounting Horizons 35% 76% 78% 75% 80% 49% 6% 6% 12% 11% December 2010 American Accounting Association Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 1 The initial disclosure of specific information about a loss contingency often does not occur until a material accrual is recognized for that loss contingency. 2 The at least reasonably possible threshold for disclosing loss contingencies has not resulted in the disclosure of the full population of an entity's existing loss contingencies that would be of interest to financial statement users. 3 The option to state that \"an estimate of the possible loss or range of loss cannot be made\" is exercised with such frequency by financial statement preparers that users often have no basis for assessing an entity's possible future cash flows associated with loss contingencies. 4 The amounts recognized in the financial statements related to loss contingencies are not transparent to users FASB 2008b, paragraph A3. To address these perceived shortcomings, the Exposure Draft proposed expanding disclosures about litigation-related loss contingencies. Specific items that would be required to be disclosed under the proposed approach included added details about the litigation, assumptions underlying estimates of the outcome, and other items see Exhibit 1 for a list containing many of the new disclosures included in the proposal. We note that some of these items can be interpreted as already required under current Generally Accepted Accounting Principles GAAP. For example, paragraph 10 of SFAS No. 5 ASC 450-20-50-4a and b and 450-20-50-5 requires that the \"nature of the contingency\" and the \"estimate of the possible loss or range of loss,\" or a statement that the estimate cannot be made be disclosed. While all of the items in Exhibit 1 can be thought of as describing certain aspects of these general requirements, the Exposure Draft's language identifies disclosure items more specifically. In this sense, while many of these items are arguably already required under present GAAP, the existing language of SFAS No. 5, together with evolution in practice, appear to have produced disclosure the FASB felt was insufficient to meet the spirit of the guidance. The Exposure Draft's more precise language can be seen as an attempt to increase the amount of information communicated. We assume that the disclosure decision regarding litigation-related loss contingencies is made based upon perceived costs and benefits. Potential litigation losses are a particularly good example of disclosure that carries potentially significant costs. Given the adversarial process underlying the 528 Desir, Fanning, and Pfeiffer The proposed amendments are a \"solution\" in search of a problem. There is no systemic failure that warrants the proposed change. Investors are not suffering from inadequate disclosure of litigationrelated loss contingencies in financial statements.4 The FASB Staff had little more than a handful of anecdotal cases of alleged insufficient disclosure to use as evidence to evaluate the arguments in the comment letters. To address this deficiency, we provide direct evidence about the alleged shortcomings mentioned above. Namely, we examine firms' disclosures in the quarter immediately prior to the recognition of a loss; the frequency with which companies claim not to be able to estimate possible losses; and the extent to which companies clearly disclose recognized accruals for contingent losses, as well as other attributes of the disclosures that we examine. In real time, it is certainly not possible to observe lawsuits that the firm judged to be less than \"at least reasonably possible\" because indeed such lawsuits would not be disclosed. However, the Exposure Draft's assertion is that \"too often\" firms disclose significant adverse outcomes regarding previously undisclosed lawsuits. If this is indeed true, then one of the following explanations must be true for the cases we examine: a The firm judged the likelihood of loss from the lawsuit to be less than \"reasonably possible\" based on unbiased interpretations of the definition of \"reasonably possible\" and the facts at the time of disclosure. Facts then changed quite drastically such that in the following approximately 90 days, the probability was revised from less than \"reasonably possible\" to \"certain\"; b The firm believed that the lawsuit was \"at least reasonably possible\" but elected not to disclose it on the grounds that it was immaterial; c The firm's interpretation of the \"reasonably possible\" threshold or of the facts was unintentionally or intentionally biased; d The firm believed the lawsuit was at least \"reasonably possible\" but elected not to disclose it thereby violating GAAP. We exploit hindsight to document the frequency with which a significant adverse outcome was not previously disclosed by the firm. With the benefit of hindsight, scenario a is highly unlikely; it seems quite implausible that the probability of loss from a material lawsuit could change from less than \"reasonably possible\" to certain within the span of a single quarter. Also by assumption, we find scenario b to be implausible, because the ultimate disclosure of the outcome is prima facie evidence that the lawsuit was material. Thus, we believe that evidence about how often we observe cases where non-disclosure is followed in the immediate next quarter by an adverse outcome is insightful about scenarios c and 4 Comment letter No. 16, located at the FASB website: http://www.fasb.org, page 2. Accounting Horizons American Accounting Association December 2010 Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 resolution of lawsuits in the United States, revelation of information to the public including the plaintiff comes with real and direct potential costs. The potential benefitslower perceived information risk on the part of current and potential investors leading to reduced cost of capital are more difficult to discern. The disclosures we observe reflect firms' trade-offs of these uncertain costs and benefits. Opposition to the Exposure Draft was significant. The FASB received over 200 comment letters in response to the Exposure Draft, and according to the FASB Staff, the vast majority of them were critical of the proposal. Parties differed in the extent to which they perceived a problem. Their opinions, though often strongly held, were not accompanied by empirical evidence. The comment letter written by the Association of Corporate Counsel 2008 is typical: Are Revisions to SFAS No. 5 Needed? 529 RELATED LITERATURE SFAS No. 5 was issued in March 1975 and was first applied in financial statements in 1976. Shortly thereafter, the Commission on Auditors' Responsibilities issued a report, \"Report, Conclusions, and Recommendations,\" AICPA 1978 suggesting that disclosure requirements for litigation uncertainties be expanded. In a 1987 report, The Private Companies Practice Section of the AICPA 1987, while opposing the expansion of disclosures, nevertheless expressed dissatisfaction with the state of contingency disclosures. Fesler and Hagler 1989; hereafter FH conducted a study similar to ours intended to shed light on how SFAS No. 5 was being applied in practice.5 In the 126 lawsuits they examined from 1982 and 1983 financial statements, they found that 35 percent of companies did not mention the litigation in the year prior to its resolution. Even for cases that were at relatively high materiality levels 50 percent of net income, non-disclosure prior to resolution was surprisingly common 24 percent. In addition, FH discuss a prior study by Robert Braun 1980 regarding non-litigation contingencies. Braun 1980 found that the incidence of satisfactory forewarnings before SFAS No. 5 was 35 percent, while after SFAS No. 5 was effective, the incidence increased to 64 percent, indicating that SFAS No. 5 appeared to improve the reporting of contingencies. The purpose of our study is to determine whether there is evidence to support the contention that further improvements are required. While our study is similar to FH, there is a need for more recent evidence on the quality of firms' SFAS No. 5 disclosures. Significant changes in the U.S. litigation environment, auditing approaches, markets, and financial reporting practices each potentially affect the financial reporting choices that firms make regarding litigation-related contingencies. Because the evidence in FH is over 25 years old, it is of limited use for current policy analyses and discussions. Also, in contrast to FH, we do not prespecify a materiality level. Instead, we simply document disclosure behavior and leave it to the reader to assess whether any of the cases we document represent material departures from the guidance in SFAS No. 5. Moreover, while FH classify firms' disclosures as satisfactory or not, we instead capture attributes of the disclosure through our coding but do not draw any arbitrary distinctions among different disclosure approaches. In a concurrent study, Hennes 2009 documents that the majority of the firms in her sample do not disclose plaintiffs' claim amounts related to their employment discrimination lawsuits and 5 We note that observed disclosure practices are a joint product of firms' strategic disclosure choices, external auditor oversight and effectiveness, the advice of corporate counsel, and the constraints imposed by SFAS No. 5. Thus any apparent disclosure deficiencies observed by prior work or in our study cannot necessarily be attributed to a shortcoming in the standard. Standard setters, auditors, and other regulators must ultimately make the judgment about the most effective way to bring about improvements in disclosures, and changes in the accounting standards may not be the optimal solution. Accounting Horizons December 2010 American Accounting Association Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 d and thus is potentially useful in understanding the ways in which firms apply the guidance about probability thresholds such as \"at least reasonably possible\" in SFAS No. 5. Taken together, our findings are generally supportive of the assertions underlying the Exposure Draft, as well as some of the specifically recommended disclosure changes. We provided an earlier draft of this manuscript to the FASB, and it was included in the materials provided to the Board by the FASB Staff as part of their preparation for re-deliberation at the August 19, 2009, public Board meeting. It gave Board members \"hard\" and objective data to complement the largely anecdotal evidence reported to them by a variety of constituents. In so doing, we believe that our work was useful in reducing the uncertainty surrounding the Board's decisions regarding proposed changes to SFAS No. 5. 530 Desir, Fanning, and Pfeiffer that very few firms provide estimates of probable losses. The firms in our sample exhibit similar behavior. Interestingly, however, she finds the textual disclosures, though seemingly deficient, appear to contain information that is useful in predicting the ultimate outcomes of the lawsuits and that investors appear to use this information in setting share prices.6 6 7 8 9 Note that Hennes' evidence about the usefulness of the disclosures does not contradict the claim that the disclosures are insufficient relative to investors' needs. Presumably, more informative disclosures could be even more useful to investors in determining the valuation implications of firms' ongoing litigation. The year 2007 was selected arbitrarily. We have no reason to believe that there was anything systematically different about lawsuit resolutions disclosed in 2007 as compared with other years from which we might have otherwise selected our sample. The 200 cases was an arbitrary cutoff point, selected based on practical grounds. Coding each case required reading several pages of descriptions of firms' litigation-related activities across multiple quarterly SEC filings. We processed cases until we reached the point where we felt we had sufficient representative data to characterize the state of existing disclosure practice. Identifying and coding disclosures related to each lawsuit was extremely time-consuming, and thus we were forced to limit the size of our sample on practicality grounds. Nevertheless, the final sample is, as described below, relatively representative on several dimensions. Accounting Horizons American Accounting Association December 2010 Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 EMPIRICAL APPROACH We assemble a random sample of lawsuit resolutions reported in companies' financial statements. We first identify all unfavorable settlements and adjudications of lawsuits mentioned in companies' forms 10-Q and 10K, filed during the first half of 2007 using keyword searches of Westlaw Business' LivEdgar. We select from SEC filings dated between January 1 and June 30, 2007, of the type 10K, 10K/A, 10-KT, or 10-K405.7 Within those filings, we search the 10K document sections \"Item 3. Legal Proceedings\" and \"Notes to Financial Statements\" using the following search rule to identify potential cases: \"settled or settlement or judgment or verdict or decision or decided or arbitrat* w/30 litigation or lawsuit or defendant or plaintiff or class action or court or claim.\" This search yields 425 hits. We randomly select from that set of 425 filings by sorting them based on filing date. We examine each of the first 200 cases and read all of the disclosures in the indicated document sections.8 Despite the intended precision of the search term, for approximately three-quarters of these cases the resolution happened in an earlier quarter, or the set of terms appeared together but did not describe a relevant situation. The remaining quarter represents the 51 cases in our sample.9 For each case, we code disclosures related to the lawsuit, capturing information relevant to our analyses from both financial statement notes and required SEC disclosures included in the 10K or 10-Q. In tracing a given lawsuit, we occasionally identify additional cases for a given firm resolved in other years. When this occurred, we add such cases to our sample. Our sample includes lawsuits from 26 companies and 25 different industries, with lawsuits filed between 1998 and 2007, the majority originating during 2005 and 2006. These two years account for almost 50 percent of the total sample. It is important to note that our paper does not explicitly test hypotheses; its evidence is descriptive in nature. As such the normal concerns about sample sizepower of the testare less applicable. However, we do indeed draw conclusions from sample statistics, and in so doing, sample size is relevant, especially in judging the extent to which our sample is representative of the full population of lawsuits. On that score, we are confident our sample is representative for two reasons: 1 we select the sample randomly from the population; and 2 the characteristics of the sample clearly indicate variation across industry 25 different industries are represented and size we have firms with assets as small as $8.1 million and as large as $63.2 billionsee Table 2. A second possible concern is that we are necessarily limited in the information we can obtain about a given lawsuit, in that we can only discover what is disclosed by the firm. For example, Are Revisions to SFAS No. 5 Needed? 531 That requirement the requirement in 450-20-25-2b to permit non-recognition of a loss due to an inability to reasonably estimate the loss shall not delay accrual of a loss until only a single amount can be reasonably estimated. To the contrary, when the condition in paragraph 450-20-25-2a is met and information available indicates that the estimated amount of loss is within a range of amounts, it follows that some amount of loss has occurred and can be reasonably estimated. Thus, when the condition in paragraph 450-20-25-2a is met with respect to a particular loss contingency and the reasonable estimate of the loss is a range, the condition in paragraph 450-20-25-2b is met and an amount shall be accrued for the loss. FINDINGS Table 1 details sample composition by industry, based on four-digit SIC codes. \"Functions Related to Deposit Banking, Not Elsewhere Classified\" SIC 6099 account for the greatest proportion almost 12 percent of the sample. Table 2 shows that the median mean loss was $1,175,000 $56,423,543, representing 5.2 percent 20.9 percent of pretax income.10 As a percentage of total assets, the median mean loss was 0.7 percent 3.8 percent. Table 3 describes various features of firms' disclosures and recognition of loss contingencies prior to the unfavorable outcomes. Despite the fact that all of these lawsuits ultimately resulted in 10 While we took note of instances where the company specifically mentioned insurance coverage for losses, we are not typically able to discern what part of the loss is insured and thus these are gross losses. Accounting Horizons December 2010 American Accounting Association Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 some relevant information we would like to know for our cases is whether the lawsuit in question is covered by insurance and thus might make an otherwise large potential loss into a relatively smaller net loss and whether the company was actually able to estimate the magnitude of the loss prior to the outcome. With respect to insurance coverage, while we do not know with certainty whether a given lawsuit is covered by a relevant insurance policy for all of our cases, when firms disclose the presence of relevant insurance, we do indeed capture that information, and thus we do know about the presence of insurance for at least some approximately 12 percent of our cases. Further, this low frequency of disclosure of insurance coverage indicates that the existence of insurance does not explain the decision not to disclose. Moreover, even if a firm has relevant insurance coverage for a given lawsuit, many times the insurance coverage itself becomes the subject of further costly litigation to the firm. Also, insurance recoveries might help to compensate for the direct costs of the litigation but do not compensate for other costs, such as loss of market share, or damage to the firm's reputation in product and/or equity markets. Thus, in our view, the presence of insurance does not fully insulate the firm from loss, nor does it reduce investors' need for the information. Therefore, it does not remove the firm's obligation to disclose contingent losses. Finally, unless the insurance recovery is certain to be coincident with the loss, the firm actually has a liability the payable to the plaintiff in the litigation and a receivable the claim filed with the insurance company. In this sense, there isn't actually a netting of the two, which would render the presence of insurance in the assessment of the measurement of and probability of loss moot. The issue of our ability to determine whether the company was able to estimate the loss is related to the discussion in the \"background\" section above: Given the unfavorable resolution occurred in quarter t, we find it highly unlikely that it was impossible for the firm to estimate its expected loss or range of possible loss as of the issue date of its quarter t1 financial statements. So, while the estimability of the loss is indeed unobservable, we argue that the circumstantial evidence that we bring to bear does shed useful light on firms' strategic reporting choices. Indeed ASC Topic 450-20-25-5 says on this point: 532 Desir, Fanning, and Pfeiffer TABLE 1 Number of Litigation Cases Investigated by Industry Classification Industry % 6 4 4 4 4 3 3 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 51 11.76% 7.84% 7.84% 7.84% 7.84% 5.88% 5.88% 3.92% 3.92% 3.92% 3.92% 3.92% 1.96% 1.96% 1.96% 1.96% 1.96% 1.96% 1.96% 1.96% 1.96% 1.96% 1.96% 1.96% 1.96% 100.00% losses to the sample companies and thus might reasonably have been judged as \"probable\" or \"at least reasonably possible\" in the quarter prior to resolution; Panel B shows that firms provided estimates of the loss or range of possible lossor explicitly stated that an estimate could not be made as required by FASB ASC Topic 450-20-50-4in only 24 of the 51 lawsuits 47.1 percent. In four of the 51 cases 7.8 percent, the existence of the lawsuit is not mentioned in the immediately prior financial report Table 3, Panel A.11 In the next section, we analyze these four cases in more detail; such analyses are more informative than the overall summary statistic. We note that this represents a substantial improvement in the rate of disclosure relative to that reported by FH, who found no mention of the litigation in 35 percent of their sample cases. 11 We analyze disclosures in the financial report immediately preceding the period of resolution. While it is possible that a company may have made a disclosure in a quarterly or annual period earlier than the report we analyze, we searched prior quarterly reports to confirm any case where there was not a disclosure immediately preceding the resolution. Note that APB 28, paragraph 22 ASC Topic 270-10-50-6 requires that once a disclosure has been made in any prior period, the disclosures would be repeated in every filing through resolution, and thus we did not anticipate finding gaps in disclosure. Accounting Horizons American Accounting Association December 2010 Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 Functions Rel to Dep Bke, NEC Computer Communication Equipment Computer Processing, Data Prep Service Eating Places Prepackaged Software Business Services, NEC Drugs and ProprietaryWhsl Calculate, Acct Mach, Ex Comp Elect Meas and Test Instruments Groceries and Related ProductsWhsl Newspaper: Publishing and Printing Semiconductors and Related Devices Apparel and Accessory Stores Chemicals and Allied ProductsWhsl Computer Integrated System Design Computer Storage Devices Electric Housewares and Fans Electronic Components, NEC Lab Analytical Instruments Metalworking, Machinery and Equipment Misc Electric Machinery, Equipment, Supplies Misc Manufacturing Industries Patent Owners and Lessors Pharmaceutical Preparations Photographic Equipment and Supplies Total # of Lawsuits Total Assets % Loss/Total Assets Pretax Income % Loss/Pretax Income Fiscal Year AFP Imaging Corp. Aspen Technology, Inc. Aura Systems, Inc. Authentidate Holding Corp. Bally Technologies, Inc. Buffets Holdings, Inc. Buffets Holdings, Inc. Cardinal Health, Inc. Cardinal Health, Inc. Cardinal Health, Inc. Children's Place Retail Stores Collectors Universe, Inc. Collectors Universe, Inc. Dollar Financial Corp. Dollar Financial Corp. Dollar Financial Corp. Dollar Financial Corp. Dyntek, Inc. Hyperdynamics Corp. Hyperdynamics Corp. Kennametal, Inc. LeCroy Corp. LeCroy Corp. Maxim Integrated Maxim Integrated Microsoft Corp. MIPS Technologies, Inc. Misonix, Inc. 325,000 5,600,000 2,200,000 343,000 1,250,000 14,400,000 7,600,000 600,000,000 40,000,000 35,000,000 2,100,000 600,000 73,000 4,000,000 1,100,000 700,000 20,000 47,500 8,870,000 280,000 900,000 1,000,000 2,800,000 19,000,000 9,100,000 1,200,000,000 3,300,000 419,000 8,153,000 274,238,000 56,122,000 48,704,000 824,895,000 538,496,000 952,299,000 23,153,800,000 23,153,800,000 23,153,800,000 997,537,000 75,534,000 78,101,000 551,825,000 551,825,000 551,825,000 833,619,000 33,968,000 11,480,000 11,480,000 2,435,272,000 193,168,000 201,974,000 3,286,537,000 3,606,784,000 63,171,000,000 127,546,000 38,086,000 3.99% 2.04% 3.92% 0.70% 0.15% 2.67% 0.80% 2.59% 0.17% 0.15% 0.21% 0.79% 0.09% 0.72% 0.20% 0.13% 0.00% 0.14% 77.26% 2.44% 0.04% 0.52% 1.39% 0.58% 0.25% 1.90% 2.59% 1.10% 1,229,000 25,947,000 24,403,000 16,118,000 38,106,000 5,514,000 100,533,000 1,252,300,000 1,252,300,000 1,252,300,000 88,751,000 7,899,000 782,000 34,479,000 34,479,000 34,479,000 5,532,000 10,785,000 23,199,000 23,199,000 447,719,000 3,015,000 10,036,000 571,854,000 413,680,000 20,101,000,000 16,309,000 1,242,000 26.44% 21.58% 9.02% 2.13% 3.28% 261.15% 7.56% 47.91% 3.19% 2.79% 2.37% 7.60% 9.34% 11.60% 3.19% 2.03% 0.36% 0.44% 38.23% 1.21% 0.20% 33.17% 27.90% 3.32% 2.20% 5.97% 20.23% 33.74% 2005 2006 1999 2007 2007 2006 2007 2007 2007 2007 2007 2005 2007 2006 2006 2006 2007 2007 2007 2007 2006 2005 2006 2006 2007 2007 2005 2005 (continued on next page) Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 Loss Amount 533 December 2010 American Accounting Association Company Are Revisions to SFAS No. 5 Needed? Accounting Horizons TABLE 2 Magnitude of Litigation Losses Relative to Reported Total Assets and Pretax Income Loss Amount Total Assets % Loss/Total Assets Pretax Income % Loss/Pretax Income Fiscal Year Nuco2, Inc. Perrigo Co. Pizza Inn, Inc./Mo. Pizza Inn, Inc./Mo. Salton, Inc. Sparton Corp. Super Micro Computer, Inc. Super Micro Computer, Inc. Super Micro Computer, Inc. TSR, Inc. USA Technologies, Inc. USA Technologies, Inc. 3,000,000 1,000,000 2,800,000 410,000 3,000,000 11,675,000 150,000 125,000 179,000 900,000 270,000 18,000 199,007,000 1,750,624,000 19,001,000 19,001,000 553,532,000 108,976,000 131,001,000 131,001,000 131,001,000 17,642,000 23,419,000 34,491,000 1.51% 0.06% 14.74% 2.16% 0.54% 10.71% 0.11% 0.10% 0.14% 5.10% 1.15% 0.05% 17,689,000 105,935,000 7,018,000 7,018,000 65,805,000 12,273,000 26,622,000 26,622,000 26,622,000 2,332,000 14,847,000 17,782,000 16.96% 0.94% 39.90% 5.84% 4.56% 95.13% 0.56% 0.47% 0.67% 38.59% 1.82% 0.10% 2006 2006 2006 2006 2006 2000 2006 2006 2006 2007 2006 2007 Mean Median 56,423,543 1,175,000 4,309,384,343 196,087,500 3.80% 0.68% 721,912,543 6,715,500 20.89% 5.20% Company Desir, Fanning, and Pfeiffer December 2010 Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 534 Accounting Horizons American Accounting Association TABLE 2 (continued) Are Revisions to SFAS No. 5 Needed? 535 TABLE 3 Evidence on the Disclosure of Loss Contingencies Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 (continued on next page) Accounting Horizons December 2010 American Accounting Association 536 Desir, Fanning, and Pfeiffer TABLE 3 (continued) Detailed Analysis of the Four Non-Disclosure Cases For each of the four sample observations where we find no disclosure in the quarter immediately preceding the quarter where the litigation was resolved, we further analyze the disclosures as well as additional preceding periods' disclosures. These analyses provide some additional nuance to help understand the nature of companies' disclosure behavior beyond the summary statistics presented below: Company Misonix, Inc. Nuco2, Inc. Loss Amount Loss as % of Pretax Income Fiscal Year $419,000 $3,000,000 33.7% 17.0% 2005 2006 (continued on next page) 12 Note that on the surface, it would appear impossible for there to be more cases where there was a loss accrual Panel D than there were cases with estimates of the loss or range of loss Panel B. However, for some cases, even though there had been a prior accrual of the loss, the first mention of the accrual did not occur until the period of the resolution. Accounting Horizons American Accounting Association December 2010 Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 In the quarter of resolution, 10 companies 19.6 percent do not report the amount of the realized loss Table 3, Panel C and 20 companies 39.2 percent do not accrue for the loss or at least do not disclose the existence of a prior accrual prior to resolution of the lawsuit Panel D.12 Are Revisions to SFAS No. 5 Needed? Company LeCroy Corp. Collectors Universe, Inc. 537 Loss Amount Loss as % of Pretax Income Fiscal Year $2,800,000 $73,000 27.9% 9.3% 2006 2007 Subsample AnalysisCases with Larger Magnitudes or That Are Related to Core Operations We conjecture that cases involving the company's core operations are more likely to be disclosed than other cases because failing to prevail in the lawsuit might have continuing impacts on the profitability of the company. For instance, there might be constraints on the behavior of the company going forward, such as in the case of patent infringement lawsuits, the loss of rights to produce a product. Also, we expect that cases involving more material amounts will be more likely to be disclosed, regardless of whether they are related to core operations. To investigate these two issues, we analyze two subsamples of our data: those cases where the loss is greater than 5 percent of pretax net income an arbitrary cutoff to identify and analyze relatively more significant losses,13 and those cases where the lawsuit is related to core operations of the company. Note that because ten of our sample companies Table 3, Panel C do not provide 13 SEC SAB No. 99, Topic 1M, \"Materiality\" SEC 1999, ASC Topic 250-10-S99-1 states, \"Quantifying, in percentage terms, the magnitude of a misstatement is only the beginning of an analysis of materiality; it cannot appropriately be used as a substitute for a full analysis of all relevant considerations.\" In this paper, we do not purport to identify \"material\" losses because the many additional factors that are used by auditors to determine materiality for a given firm and financial statement are unobservable to us. Instead, our classification should be viewed as a loose approximation of those cases that might be material in a financial reporting sense. Accounting Horizons December 2010 American Accounting Association Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 We provide the actual disclosures for these four companies above in the Appendix. Our comments regarding each case are noted in italics within the disclosures in the Appendix. In each of these four cases, we find no prior disclosure of the loss, even in periods prior to the period we analyze. It can be argued that there are extenuating circumstances for Nuco2, Inc. and Collectors Universe, Inc. Nuco2, Inc. indicated that the loss was covered by insurance policies, and thus management might have judged the case to be immaterial in previous periods. However, as noted earlier, we assume that disclosure of the outcome of the lawsuit is prima facie evidence that the case was material to the company's financial statements. Moreover, despite the existence of insurance, financial statement users might reasonably argue that information about even insured losses is relevant to their evaluations of the companies, especially given that insurance coverage itself can sometimes be subject to limitations, interpretation, and even follow-up litigation. The Collectors Universe, Inc. litigation loss, while a substantial proportion of the company's 2007 pretax loss, would not have been large relative to its income loss in surrounding years. In 2006 and 2008, Collectors Universe, Inc. had pretax income loss of $6,137,000 and $16,646,000, respectively, and the $73,000 loss for the lawsuit in question would have been just 1.2 percent and 0.4 percent of those numbers, respectively. This might explain the non-disclosure in the period immediately prior to resolution. Whether the apparent exceptions to the spirit or letter of disclosure rules represent cause for concern sufficient to warrant changes in GAAP is, of course, a matter of subjective judgment. Before conducting the study, we expected to find relatively few exceptions, given the sensitivity of information about litigation and the relatively clear guidance provided by ASC Topic 450-20-50. In our view, the exception rate of 4 out of 51 cases is surprisingly large. 538 Desir, Fanning, and Pfeiffer the loss amount, we cannot classify those cases based on the magnitude of the loss. See Table 2 for detailed data for all cases for which we had loss information. Our analyses reveal the following See Table 3, Panel J, for a summary of the following list of comparisons: Additional Analysis: Extent of Provision of Proposed Disclosures One view of the market for financial reporting information would argue that if disclosures are useful to investors, then investors would demand such disclosures both explicitly through their communications with the companies, and indirectly by rewarding those companies who provide more extensive disclosures with lower costs of capital. Such demand provides incentives for companies to disclose additional information about contingent legal liabilities. If companies provide disclosures such as those proposed by the FASB in the Exposure Draft, this could be interpreted as indirect evidence of user demand for such information and, in turn, Accounting Horizons American Accounting Association December 2010 Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 20 of the 41 cases that provide loss amounts have losses that exceed 5 percent of pretax income see Table 2. Indeed, as mentioned above, the median case has a loss of 5.2 percent, indicating that fully half of our sample consists of substantial losses. 16 of those cases 80.0 percent mention the existence of the lawsuit in the prior financial report, but the remaining 4 20.0 percent do not Table 3, Panel E. Although we do not perform statistical analysis of the difference because of the small sample, this finding contradicts the notion that non-disclosure in the broad sample can be explained by relatively smaller dollar amounts, as the full-sample non-disclosure rate is just 7.8 percent Panel A. Lawsuits exceeding 5 percent of pretax income are accrued in advance of resolution or at least disclosure of an accrual is made in 55.0 percent of the sample cases Panel F, which is similar to the sample-wide rate of 60.8 percent Panel D, indicating that more material cases are not necessarily accrued more often. 22 of our sample lawsuits relate to the companies' core operations. For those lawsuits related to core operations, companies appear to be no more likely to mention the existence of the lawsuit in prior financial reports no mention in 3 of the 22 core-operations-related lawsuits, Panel G, as compared with 4 out of the full sample of 51 lawsuits Panel A. The likelihood that companies accrue for such losses or at least disclose that an accrual has been made does not differ when litigation is related to core operations 59.1 percent of such cases disclosed accruals Panel H, as compared with 60.8 percent in Panel D. We investigate whether prior mention of the lawsuit is associated with the likelihood of accruals in advance of resolution or at least disclosure thereof. While the sample-wide rate of accrual disclosure is 60.8 percent Panel D, the rate for companies who previously mention the existence of the lawsuit is similar, 63.8 percent Panel I. Overall, we find evidence supportive of the assertions made in the Exposure Draft. Specifically, we find that in an unexpectedly large proportion of the sample of cases we investigate, companies do not disclose the existence of the lawsuit until the loss occurs. We also find a relatively high incidence of companies not providing estimated losses or ranges of losses prior to the resolution of litigation. And accruals for estimated loss amounts are not made or at least are not disclosed prior to the realization of the loss for a substantial number of cases. In our opinion, these findings validate the concerns expressed by financial statement users that there is often insufficient timely information about litigation to enable estimation of its impact on future cash flows. Are Revisions to SFAS No. 5 Needed? 539 SUMMARY AND CONCLUSIONS This paper provides evidence from a sample of unfavorable lawsuit outcomes about the disclosure behavior of companies in periods prior to the resolution of litigation. Overall, we find evidence of a surprisingly large incidence of non-disclosure of contingent losses stemming from litigation that cannot be fully explained by the magnitude of the items in question. We also find a relatively high degree of disclosure of the items called for in the FASB's Exposure Draft, consistent with the conjecture that at least some of these items are presently being demanded by users. In sum, these findings provide the basis for a fact-based argument about present disclosure behavior related to litigation contingencies, as well as some initial evidence about the supply of and demand for disclosures incremental to those currently required under existing GAAP. As such, they represent an example of how academic research can play a useful role in debates about standard setting, specifically by identifying and verifying assumptions upon which proposed standard-setting projects and judgments are based. We believe that additional opportunities to contribute exist and should be undertaken to the mutual benefit of both accounting academics and the financial reporting environment. We note that at the time of writing this manuscript, the FASB's deliberations regarding this issue are ongoing. In the August 2009 meeting, the Board reflected on the 240 comment letters they had received on the Exposure Draft and tentatively decided to focus required disclosures on publicly available information, such as the contentions of the parties to the lawsuit and other facts that would not affect the outcome of the litigation. Deliberations about the exact content of any final rule are expected to continue into 2010. We believe there could be benefits to extending this work, including expanding the sample to investigate a larger number of cases; identifying characteristics of the firms and circumstances associated with relatively more or less disclosure about litigation-related contingencies; and investigating the extent to which there might be trends in disclosure across recent time periods in response to changes in the legal, economic, and regulatory environment. Accounting Horizons December 2010 American Accounting Association Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 would validate those provisions of the Exposure Draft. On the other hand, widespread provision of the information might suggest that mandating such disclosures is unnecessary. Our final set of analyses focuses on this question: To what extent do our sample companies provide elements of the disclosures called for in the Exposure Draft? For each of the lawsuits in our sample, we identify whether the company provides the disclosure items listed in Exhibit 1. We also indicate the percentage of our sample companies who provide information on each item in the period immediately preceding the resolution of the case. As shown in Exhibit 1, a substantial proportion of our sample companies appear to disclose four of the items called for in the Exposure Draft items 2-5. Approximately half our sample of companies also provide information about anticipated timing of the resolution of the contingency item 6, and more than one-third provide either the total claim amount or an estimate of the maximum loss exposure item 1. Our finding that items 2-5 required by the Exposure Draft are provided more than 50 percent of the time suggests that there is investor demand for this information and/or firms generally interpret the existing guidance as requiring the disclosure of these particular items. In contrast, our finding that items 1 and 6-10 are disclosed less than half of the time suggests that there is potentially less investor demand for this information and/or managers perceive the cost of disclosing these items to exceed the benefits. Whether those items or the others that have been proposed should become mandatory disclosures is, of course, a policy-making decision that must be made in light of the perceived costs and benefits to various parties of such a change in the accounting standard. 540 Desir, Fanning, and Pfeiffer APPENDIX Excerpts from Financial Reports of Companies with No Prior Disclosure of Material Lawsuits Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 Accounting Horizons American Accounting Association December 2010 Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 December 2010 American Accounting Association Accounting Horizons 541 Are Revisions to SFAS No. 5 Needed? Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 December 2010 Accounting Horizons American Accounting Association Desir, Fanning, and Pfeiffer 542 Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 December 2010 American Accounting Association Accounting Horizons 543 Are Revisions to SFAS No. 5 Needed? 544 Desir, Fanning, and Pfeiffer American Institute of Certified Public Accountants AICPA. 1978. Commission on Auditors' Responsibilities: Report, Conclusions, and Recommendations. New York, NY: AICPA. -. 1987. Private Companies Practice Section Reporter. New York, NY: AICPA. Association of Corporate Counsel. 2008. Comment letter to the FASB on the SFAS No. 5 Exposure Draft. August 8. Available at: http://www.fasb.org/cs/BlobServer?blobcolurldata&blobtable MungoBlobs&blobkeyid&blobwhere1175818505475&blobheaderapplication%2Fpdf. Braun, R. M. 1980. Concerning accounting for uncertainties. Doctoral dissertation, New York University. Fesler, R. D., and J. L. Hagler. 1989. Litigation disclosures under SFAS No. 5: A study of actual cases. Accounting Horizons 3 1: 10-20. Financial Accounting Standards Board FASB. 2008a. Accounting for Contingencies. Norwalk, CT: FASB. Accounting Horizons American Accounting Association December 2010 Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 REFERENCES Are Revisions to SFAS No. 5 Needed? 545 Accounting Horizons December 2010 American Accounting Association Downloaded from http://meridian.allenpress.com/accounting-horizons/article-pdf/24/4/525/1346354/acch_2010_24_4_525.pdf by Julia Higgs on 23 May 2020 -. 2008b. Disclosure of Certain Loss ContingenciesAn Amendment of FASB Statements No. 5 and 141(R). Exposure Draft. Norwalk, CT: FASB. Hennes, K. 2009. The reporting of contingent legal liabilities: Disclosure of employment discrimination lawsuits. Working paper, University of Oklahoma. Moehrle, S., K. Anderson, F. Ayres, C. Bolt-Lee, R. Debreceny, M. Dugan, C. Hogan, M. Maher, and E. Plummer. 2009. The impact of academic accounting research on professional practice: An analysis by the AAA Research Impact Task Force. Working paper, American Accounting Association. U.S. Securities and Exchange Commission SEC. 1999. Materiality. Staff Accounting Bulletin No. 99. Washington, D.C.: SEC. -. 2008. Advisory Committee on Improvements to Financial Reporting to the United States Securities and Exchange Commission. Final Report. Washington, D.C.: SEC