Fair Value Accounting in the Banking Industry

Abstract

This paper studies the application of fair value accounting in bank holding companies in the United States with the purpose of evaluating the effects of expanding fair value accounting in the banking industry. The paper documents the current application of fair value accounting in the industry, showing what proportions of recognized assets and liabilities of bank holding companies are at or close to fair value on the balance sheet, have related unrealized gains and losses in income, or have fair values disclosed in the footnotes. In each case, it evaluates the advantages and disadvantages of the current treatment and makes an assessment of the magnitudes of economic assets and liabilities that currently are omitted from the balance sheet. Turning to the issue of the likely impact of expanded application of fair values, the paper asks how large are the actual and potential differences between fair values and book values of various assets and liabilities and how credit, interest rate and prepayment risks are likely to determine those differences. It considers the availability of market-based information for measuring fair value and evaluates the correlations between fair values of economic assets and liabilities due to natural hedges, asset-liability management, and changes in asset values affecting the risk and value of liabilities. It also addresses the issue of how a switch to fair value accounting would mitigate or, alternatively, facilitate earnings management activities. The primary conclusion of the analysis is that expanding fair value accounting is not likely to significantly improve the information in bank financial statements and, in some cases, may introduce distortions that reduce accounting quality. The analysis is pertinent to bank analysts and investors for it not only documents the impact of fair value accounting on banks but also informs about the drivers of value and risk on which accounting should report.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 36 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.