Discussion—The Reliability of Fair Value versus Historical Cost Information: Evidence from Closed-End Mutual Funds

Abstract

Carroll, Linsmeier, and Petroni examine whether fair value financial statements reported by closed-end mutual funds are reliable and relevant. They also examine whether their inferences are sensitive to fund type, where they classify funds into six categories by their majority investment type. They find a significant association between stock market metrics (price and return) and reported fair values and their changes in all cases, controlling for the relation between stock metrics and historical cost values. They conclude that the fair value financial statements issued by closed-end mutual funds are incrementally value-relevant.I argue that the closed-end fund setting is well suited to investigating the valuation role of accounting, because the contracting role of accounting is relatively small and also highly aligned with the valuation role. The closed-end fund setting thus serves as a best-case benchmark for fair value accounting, and I assess the results from this vantage point. I extend the authors’ cross-sectional analysis to examine whether value-relevance has changed over time for closed-end funds, and find that it has remained quite stable. I evaluate whether the authors’ inferences hold under more powerful tests of relevance and reliability, and find that they do. While the cross-fund tests are consistent with fair value estimates from liquid markets being more reliable, I suggest alternative explanations. Finally, I discuss what we can learn about the desirability of fair value accounting from this paper, and urge caution in generalizing the results to more complex firms.