臺大管理論叢 NTU Management Review VOL.30 NO.2

Predicting Future Performance Using Fair Value versus Historical Cost: Evidence from Investment Property 312 1. Introduction This study is motivated by the debate over the predictive ability of fair value, and the mixed findings on whether unrealized gains and losses are a predictor of future performance. The International Accounting Standards Board (IASB) and Financial Accounting Standards Board (FASB) in their joint project, the Conceptual Framework for Financial Reporting 2010 (Financial Accounting Standards Boards, 2010), states that financial information needs to possess predictive value to be useful for decision-making. The IASB considers fair value measurement superior to other measurement methods in reflecting current economic situations and future expectations (Barth, 2006). Fair value measures could have better predictive value for future performance because fair value measurements of assets can be realized by settlement or sale (Evans, Hodder, and Hopkins, 2014) and more accurately reflects volatility. Therefore, fair value accounting incorporates more price-relevant information than historical cost information in financial statements. In addition, fair value accounting is considered to be a more accurate measure of an investment property’s selling price than the asset’s historical cost (So and Smith, 2009). A more accurate measure for current performance can align with future performance more closely than a measure with less precision. However, a potential disadvantage of fair value accounting is that, by incorporating unrealized economic gains and losses into financial statements, earnings become more volatile and difficult to predict (Ball, 2006). Recognizing changes in asset fair values in net income will likely increase earnings volatility, which reduce the predictive value of earnings (Singleton-Green, 2007; Liang and Riedl, 2014). The survey evidence of Graham, Harvey, and Rajgopal (2005) suggest that managers believe that earnings volatility is negatively associated with earnings predictability. Dichev and Tang (2009) report that financial analysts do not understand how earnings volatility translates into future earnings. Indeed, the illiquidity of some asset markets (Laux and Leuz, 2009) cause volatility in balance sheets (Schipper, 2007). The extant literature raises concerns and questions about whether measuring and recording assets at fair value is an improvement over historical cost in predicting future firm performance. These concerns constitute important empirical research questions since the objective of financial reporting is to provide firm stakeholders with information that is useful for decision-making, assist in the prediction of returns, as well as assess the amounts, timing, and uncertainty of future performance (FASB, 2010).