Anchoring Heuristic and Asset Prices

Abstract

Earnings volatility of a firm depends on a large number of factors. Optimal Bayesian inference requiring accurate assessment of all the relevant factors is generally intractable. Many factors are sector-specific, that is, common to all firms in the same sector, while others are firm specific. A pragmatic approach is to use the sector specific factors to furnish an informative starting point and then attempt to make adjustments for the firm specific factors. This article incorporates this pragmatic approach into the standard consumption-based asset pricing model and shows that such adjustment gives rise to a new mechanism capable of generating the puzzling data patterns. The model also makes a novel prediction: less volatile stocks outperform more volatile stocks. Empirical evidence supports this prediction.