The Best of All Possible Worlds: Using Analysts' Scenario-Based Valuations to Assess Target Price Optimism

Using a unique dataset of scenario-based investment reports, we examine whether the placement of an analyst's valuation forecast, relative to his/her own subjective assessment of the distribution of scenario-based valuations for the covered firm (i.e., embedded "tilt"), conveys information to investors. We document that this embedded tilt is a function of both fundamental risk attributes and behavioral factors. We find that among analysts forecasting price appreciation over the next twelve months, embedded tilt incrementally predicts ex-post valuation errors and realized returns, with valuations tilted towards upside (downside) scenarios generating more negative (positive) valuation errors and returns. In contrast, when analysts forecast price declines, the predictive value of tilt disappears, suggesting that the negative valuation forecast is a sufficient statistic for analyst pessimism. Additional analyses reveal that estimates of implied tilt derived from observable firm characteristics can be used by investors to distinguish among target price forecasts lacking scenario-based information. Although we cannot distinguish between rational and behavioral explanations for our results, we show that analyst valuation forecasts that predict equally optimistic implied returns to investors are not all created equal, and that observable firm characteristics correlated with asymmetric valuation risks and biases can be used to distinguish between investments.