Are Revisions to SFAS No. 5 Needed?

Abstract

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. In undertaking this project, the FASB heard from a wide array of constituents about their impressions of the quality of information being generated about litigation-related loss contingencies. The Exposure Draft (“Proposed Statement of Financial Accounting Standards: Disclosure of Certain Loss Contingencies”, FASB, June 5, 2008) 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, at the time of the Exposure Draft, there was very little actual data about 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 document the characteristics of the disclosures that firms made prior to the reporting of the unfavorable outcome. We find a surprisingly large incidence of non-disclosure of contingent losses that cannot be readily explained. Moreover, even for instances where there is disclosure, we find a relatively high frequency of 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 loss. On the other hand, we find that firms relatively frequently provide disclosure items called for in the FASB’s Exposure Draft but not yet required, consistent with the conjecture that at least some of these items are presently being demanded by users. Our study was part of the materials provided by the FASB Staff to the Board in their re-deliberations of this issue. We believe it serves as an example of how academic research can contribute ex ante to policy-making, specifically by indentifying and verifying assumptions upon which standard-setting projects and judgments are based.