Does Fair Value Accounting Contribute to Market Price Volatility? An Experimental Approach (Digest Summary)

In an experimental setting, the authors investigate the effect of fair value accounting (FVA) and historical cost accounting (HCA), two accounting measurement rules for held-for-trading (HFT) investments, on market price volatility for nonfinancial firms. They find no systematic difference between FVA and HCA for three different measures of market price volatility. But they do notice that FVA tends to be characterized by higher earnings volatility and slightly heavier trading. They also recognize that FVA could affect the judgment of users of financial statements and that unsophisticated investors may not fully understand the implications of FVA.

FVA has been criticized on the grounds that it could contribute to greater price volatility. Some critics have even blamed FVA for exacerbating the global financial crisis.

In an experimental framework, the authors examine the relationship between FVA and market price volatility for nonfinancial corporations. FVA and HCA are two accounting measurement rules that are used for HFT investments. FVA typically reports HFT investments at fair market value, and unrealized gains and losses are reported on the income statement. HCA reports securities at cost and only affects the income statement once gains and losses are realized. After analyzing three well-known measures of market price volatility—sum of absolute returns, squared returns, and variance—the authors find that there is no major difference in price volatility between FVA and HCA. But they do find higher volatility in accounting numbers and marginally greater trading activity under FVA. The judgment of those who rely on financial statements for guiding investment decisions could also be influenced by FVA. Unsophisticated investors may not always correctly assess the effect of fair value movements.

Standard setters, regulators, and general critics of FVA could find the conclusions of this research useful. It would seem that compared with HCA, FVA does not typically result in higher market price volatility.

The authors use a controlled experiment framework for their research, in which 36 graduate students in finance and public accounting at Université Laval participated. These students had prior experience in valuation and trading activities. All experiments were conducted at the Financial Services Authority trading rooms of the Faculté des sciences de l’administration at Université Laval and use the University of Toronto’s Rotman Interactive Trader. Two groups of participants were asked to trade the same nonfinancial corporation in separate markets. Both groups had identical information, except for information pertaining to HFT securities. One group received financial statements that reported HFT investments at fair market value on the balance sheet together with changes in unrealized gains and losses on the income statement. The other group received financial statements that used HCA principles. Securities were reported at cost and only affected the income statement once a gain or loss was realized. The authors then test whether these accounting rules have an impact on market price volatility by analyzing three volatility measures—sum of absolute returns, squared returns, and variance. They find that market price volatility is not higher under FVA. But FVA does tend to be characterized by higher earnings volatility and greater trading activity.

Although there is no systematic difference in price volatility between FVA and HCA, the authors note that FVA may affect the judgment of those who use financial statements in their investment decision making. Unsophisticated investors may also not have a good understanding of the FVA implications.

There have been various debates on whether FVA contributes to the volatility of market prices. The authors find no major difference between FVA and HCA for three measures of market price volatility of nonfinancial firms. But FVA does tend to be characterized by higher earnings volatility and slightly heavier trading activity. The results are interesting. Although I do have concerns that the participants in the experiment all have a similar background and may not be as diverse as I would have expected, the research supports the view that the use of FVA as an accounting rule does not always result in greater share price fluctuations.