Do Deviations from Investor Tastes Signal Informed Trading?

This paper provides a novel method for identifying informed institutional trading by conditioning on investors’ tastes. Given an information signal about a stock, an investor’s tastes for certain characteristics of that stock can influence his decision to trade. Thus, trades which deviate from an investor’s tastes are more likely to reflect information. I test this hypothesis with respect to institutional investors’ taste for positive (expected) skewness. Measuring skewness preferences by their past portfolio allocations, I find that the most skewness-averse institutions earn abnormal returns of 152-207 basis points per quarter on their holdings of stocks with high expected idiosyncratic skewness (EISKEW). Their holdings of high EISKEW stocks outperform those of the most skewness-loving institutions by 126-128 basis points per quarter, and this outperformance is stronger for fresh holdings and among the holdings of more aggressive institution types. I confirm the main result using an alternative proxy for skewness preferences based on local religious beliefs. An analysis of institutions’ sin stock holdings provides complementary evidence using another dimension of investor tastes, supporting the hypothesis that trades which deviate from an investor’s tastes are more likely to reflect information. ∗Please address all correspondence to Jeremy Page, Department of Finance, University of Texas at Austin, 1 University Station B6600, Austin, TX, 78712; Phone: 512-431-6157; email: [email protected]. I would like to thank Alok Kumar (my chair), Andres Almazan, Vidhi Chhaochharia, John Griffin, Umit Gurun, Andy Koch, George Korniotis, Bob Parrino, Clemens Sialm, Oliver Spalt, Sheridan Titman, Malcolm Wardlaw, and seminar participants at UT-Austin and the University of Miami for helpful discussions and valuable comments. I am responsible for all remaining errors and omissions. Do Deviations from Investor Tastes Signal Informed Trading? Abstract – This paper provides a novel method for identifying informed institutional trading by conditioning on investors’ tastes. Given an information signal about a stock, an investor’s tastes for certain characteristics of that stock can influence his decision to trade. Thus, trades which deviate from an investor’s tastes are more likely to reflect information. I test this hypothesis with respect to institutional investors’ taste for positive (expected) skewness. Measuring skewness preferences by their past portfolio allocations, I find that the most skewness-averse institutions earn abnormal returns of 152-207 basis points per quarter on their holdings of stocks with high expected idiosyncratic skewness (EISKEW). Their holdings of high EISKEW stocks outperform those of the most skewness-loving institutions by 126-128 basis points per quarter, and this outperformance is stronger for fresh holdings and among the holdings of more aggressive institution types. I confirm the main result using an alternative proxy for skewness preferences based on local religious beliefs. An analysis of institutions’ sin stock holdings provides complementary evidence using another dimension of investor tastes, supporting the hypothesis that trades which deviate from an investor’s tastes are more likely to reflect information. This paper provides a novel method for identifying informed institutional trading by conditioning on investors’ tastes. Given an information signal about a stock, an investor’s tastes for certain characteristics of that stock can influence his decision to trade. Thus, trades which deviate from an investor’s tastes are more likely to reflect information. I test this hypothesis with respect to institutional investors’ taste for positive (expected) skewness. Measuring skewness preferences by their past portfolio allocations, I find that the most skewness-averse institutions earn abnormal returns of 152-207 basis points per quarter on their holdings of stocks with high expected idiosyncratic skewness (EISKEW). Their holdings of high EISKEW stocks outperform those of the most skewness-loving institutions by 126-128 basis points per quarter, and this outperformance is stronger for fresh holdings and among the holdings of more aggressive institution types. I confirm the main result using an alternative proxy for skewness preferences based on local religious beliefs. An analysis of institutions’ sin stock holdings provides complementary evidence using another dimension of investor tastes, supporting the hypothesis that trades which deviate from an investor’s tastes are more likely to reflect information.