Print
E-MailThis post comes to us from Chang Song, Assistant Professor of Accounting at Sungkyunkwan University, Wayne Thomas, Professor of Accounting at the University of Oklahoma, and Han Yi, Assistant Professor of Accounting at the University of Oklahoma.
In our paper, Value Relevance of FAS 157 Fair Value Hierarchy Information and the Impact of Corporate Governance Mechanisms, forthcoming in The Accounting Review, we use banking firm data from the first three quarters of 2008 to examine two important research questions related to fair value information provided by banks under FAS 157. First, we compare the value relevance of Level 1 and Level 2 fair values to the value relevance of Level 3 fair values. Second, we consider whether the impact of corporate governance on the value relevance of fair values is greater for Level 3 assets compared to Level 1 and Level 2 assets. Under FAS 157, firms are required to disclose fair values of asset and liability types by levels, where levels are based on inputs used to generate fair values: (1) Level 1 (observable inputs from quoted prices in active markets), (2) Level 2 (indirectly observable inputs from quoted prices of comparable items in active markets, identical items in inactive markets, or other market-related information), and (3) Level 3 (unobservable, firm-generated inputs). Thus, fair values are disclosed from most reliable (Level 1) to least reliable (Level 3), with a classification that potentially falls somewhere in the middle (Level 2).
We find the following. First, all Level information is value relevant. Level 1 and Level 2 assets (liabilities) have valuation coefficients close to their theoretically predicted value of one (negative one). However, Level 3 assets are valued less than one and less than Level 1 and Level 2 assets. Level 3 liabilities have a valuation coefficient less than negative one (i.e., absolute value greater than one) and less than those of Level 1 and Level 2 liabilities. These results are robust to potential confounding firm characteristics such as firm size and Tier 1 capital ratios. These results suggest that investors are likely to decrease the weight they place on less reliable Level 3 fair value measurements in their equity pricing decisions due to the information risk, inherent estimation errors, and possible reporting bias. Second, we find evidence that the value relevance of fair value assets varies with the strength of a firm’s corporate governance. For firms with low corporate governance, valuations of Level 1 and Level 2 assets are below one (0.81 and 0.83). These valuations are marginally different from one at approximately the 0.07 level (one-tailed). We find that the valuation of Level 3 assets of low governance firms is close to zero (0.06) and not significant, suggesting no value relevance. For high governance firms, we find that the valuations of Level 1 and Level 2 assets increase to near one (1.01 and 0.97), and valuation of Level 3 assets increases to 0.82. The increase in asset valuations for each of the levels is consistent with strong governance reducing information asymmetry and mitigating estimation errors or reporting biases, and this is especially apparent for Level 3 assets (i.e., unobservable, firm-generated amounts) where information asymmetry is expected to be the highest.
As an additional analysis, we find that FAS 157 fair value hierarchy disclosure disaggregating Type information to Level information (i.e., matrix format for reporting fair values based on asset/liability type and level of input) adds some incremental value relevance to existing Type information. This finding is consistent with the prediction that FAS 157 provides useful incremental information for investors’ equity valuations.
Although we recognize that fair values may affect financial statement users in contexts other than the pricing of equity securities (Holthausen and Watts 2001), we believe that our findings should be important to U.S. and international standard setters for understanding not only the effects of FAS 157 disclosures, but also how future standards (e.g., the joint IASB/FASB project on financial statement presentation) can enhance existing fair value disclosures. For example, we find that investors discount less reliable fair values possibly due to information asymmetry and moral hazard problems. However, to the extent firms have strong governance, these problems appear less severe. The impact of corporate governance on the value relevance of fair values has received limited attention in academic research, especially in the U.S.
The full paper is available for download here.
