Loss Function Asymmetry and Forecast Optimality: Evidence from Individual Analysts' Forecasts

• Main Authors: Markov, Stanimir, TAN, Min Yen
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2006
• Subjects: Accounting; Finance and Financial Management
• Online Access: https://ink.library.smu.edu.sg/soa_research/165; https://ink.library.smu.edu.sg/context/soa_research/article/1164/viewcontent/markovandtan_may09_2006.pdf
• Institution: Singapore Management University

We examine the optimality of quarterly earning forecasts issued by individual analysts. When we conduct Ordinary Least Squares (OLS) and Least Absolute Deviations (LAD) analyses, which assume loss function symmetry, we reject the null of forecast optimality at 5% significance level more than 5% of the time. Relaxing the symmetry assumption reduces the frequency of rejections below 5%. We demonstrate that the cross-sectional variation in the asymmetry parameter of the loss function is related to analyst employment. Overall, our evidence is consistent with the joint hypothesis of asymmetric loss and forecast optimality rather than the alternative of symmetric loss and lack of optimality.