Separating the Wheat from the Chaff: Identifying the Signal and Predictable Error Components of Target Price Forecasts
A target price forecast scaled by current price provides the analyst’s assessment of the implied return to that stock over the stated forecast horizon. Using a simple valuation framework, we show that implied returns are a function of the expected dividend distribution, analysts’ private information that is not yet reflected in prices, and errors with respect to forecasting cash flows and the discount rate. We provide empirical evidence consistent with implied returns being correlated with proxies for these factors and that analysts make systematic errors with respect to valuation inputs. We further show that job-related incentives impact analysts’ valuations over and above the theoretical inputs. We investigate whether investors distinguish the signal and predictable bias components of implied returns. We find that the signal component is positively related to future returns over short windows (up to two months), while the bias component is negatively related to future stock returns over the six and twelve-month horizon. These results suggest that investors incorporate the signal component fairly quickly into prices, but take time to correct prices for the predictable bias in implied returns.
Conditionally Conservative Fair Value Measurements
We investigate whether firms’ measurements of the fair values of financial instruments based on Level 2 and 3 inputs other than active market prices for those instruments exhibit conditional conservatism (i.e., timelier recognition of unrealized losses than gains). We estimate conditional conservatism using an expansion of Basu’s (1997) reverse regression model that decomposes the dependent variable comprehensive income into the parts attributable to fair value measurements versus other measurement bases and that interacts the returns-related explanatory variables with the proportions of assets that are financial instruments with fair values measured using Level 2 and 3 versus Level 1 inputs. We find firms with more Level 2 and 3, but not Level 1, fair value measurements exhibit more conditionally conservative comprehensive income attributable to those measurements but not to other measurement bases. We further find that this increase is larger for firms with poorer information environments and for years following rather than during the financial crisis and following rather than preceding the effective date of SFAS 157. To explore the generalizability of these findings to non-financial assets, we conduct similar analysis on oil and gas firms’ measurements akin to the fair value of their reserves, obtaining consistent results. Overall, our results indicate that firms report conditionally conservative fair value measurements when market imperfections enable them to exercise discretion over those measurements, consistent with firms attempting to attain the well-documented contracting and other benefits of conditional conservatism (e.g., Watts 2003).
About the Author: Wesley Gray, PhD
—
Important Disclosures
For informational and educational purposes only and should not be construed as specific investment, accounting, legal, or tax advice. Certain information is deemed to be reliable, but its accuracy and completeness cannot be guaranteed. Third party information may become outdated or otherwise superseded without notice. Neither the Securities and Exchange Commission (SEC) nor any other federal or state agency has approved, determined the accuracy, or confirmed the adequacy of this article.
The views and opinions expressed herein are those of the author and do not necessarily reflect the views of Alpha Architect, its affiliates or its employees. Our full disclosures are available here. Definitions of common statistics used in our analysis are available here (towards the bottom).
Join thousands of other readers and subscribe to our blog.