Stocks

Analyst Forecast Skewness And Cross Section Stock Returns

Analyst Forecast Skewness and Cross Section Stock Returns via SSRN

Cai Zhu

Hong Kong University of Science & Technology – School of Business and Management

May 5, 2015

Abstract:

In the paper, we show a significant economic linkage between analyst EPS forecast skewness and cross section stock returns. The effect on stock return of our skewness measure is quite different from that based on skewness calculated from options or high frequency data. Literature shows that, using such skewness as a signal, trading profit is generated mostly from over-valued stocks with high positive skewness, which is consistent with Barberis and Huang (2008)’s lottery arguments. However, we find that for our analyst forecast skewness, trading profit mainly comes from those stocks with negative skewness. Long-short strategy purchasing stocks with low forecast skewness and shorting those with high forecast skewness earns annualized abnormal returns 11% with sharpe ratio 0.64. Our study suggests that negative skewness stocks tend to be undervalued (risk-adjusted returns for negative skewness stocks are significantly positive), while stocks with high positive skewness have fair prices (risk-adjusted returns for positive skewness stocks are not significant). Our empirical results are closely related with investors learning behavior and consistent with Veronesi (1999) theory. In the model, Veronesi shows that when investors cannot observe cash flow growth rate, they tend to overreact to bad news, push current stock price down, such behavior will lead to higher future stock returns. Our results also hold when using robust skewness defined as the gap between analyst EPS forecast mean and median.

Analyst Forecast Skewness And Cross Section Stock Returns – Introduction

There is a long history for studies on higher moments in returns. Researchers such as Rubinstein (1973) and Kraus and Litzenberger (1976, 1983) develop models of expected returns that incorporate skewness. In these models, the higher moments that are relevant for individual securities are co-moments with the aggregate market portfolio. Harvey and Siddique (2000) explore both skewness and co-skewness and test whether co-skewness is priced, and Dittmar (2002) tests whether a security co-skewness and co-kurtosis with the market portfolio might influence investors expected returns.

More recent, empirical work provides evidence that, besides co-movement with market, higher moments of the return distribution themselves are important in pricing securities. Works of Barberis and Huang (2008) and Brunnermeier, Gollier, and Parker (2007), together with the empirical evidence presented in Mitton and Vorkink (2007) and Boyer, Mitton, and Vorkink (2010) imply that the skewness of individual securities may also influence investors portfolio decisions. Investors prefer positive skewness assets based on speculative desire, and such preference increases demand, pushes current price high and generate lower future returns. Not only for stock returns, Boyer and Vorkink (2014) also find support for such theory from option returns.

In this paper, we consider a different measure for skewness: analyst EPS forecast skewness. Such skewness proxy is different from most widely used measures in literature fundamentally, such as realized skewness from tick-by-tick data (Amaya, Christoffersen, Jcobs, and Vasquez, 2015) and risk neutral skewness from option data (Conrad, Dittmar and Ghysels, 2013). Our measure is skewness in analysts’ belief, which is closely related to firm fundamentals, since analysts use fundamental data heavily to predict future earnings.

Different sources of skewness may have different effects on asset returns. Our findings are significantly distinct from common wisdom in many papers, representing by Barberis and Huang (2008). Such branch of literature shows that, using skewness from return distribution as a signal, trading profit is generated mostly from over-valued stocks with high positive skewness, due to lottery property. However, we find that for our analyst forecast skewness, trading profit mainly comes from those stocks with negative skewness. Long-short strategy purchasing stocks with low forecast skewness and shorting those with high forecast skewness earns annualized abnormal returns 11% with sharpe ratio 0.64. Our study suggests that negative skewness stocks tend to be undervalued (risk-adjusted returns for negative skewness stocks are significantly positive), while stocks with high positive skewness have fair prices (risk-adjusted returns for positive skewness stocks are not significant). Our empirical results are closely related with investors learning behavior and consistent with Veronesi (1999) theory. In the model, Veronesi shows that when investors cannot observe cash  ow growth rate, they tend to overreact to bad news, push current stock price down, such behavior will lead to higher future stock returns.

Stock Returns

See full PDF below.