مقاله انگلیسی رایگان در مورد نگاهی دیگر در مهار کردن و قابلیت پیش بینی بازده سهام – الزویر 2018

Since the seminal study by Jegadeesh and Titman (1993), stock return momentum has been extensively studied in the literature. While conventional momentum strategies rank stocks based on their recent past returns, an alternative momentum strategy based on stocks’ 52-week high prices is proposed in George and Hwang (2004). George and Hwang find that the stocks trading near their 52- week high prices significantly outperform the stocks trading far from their 52-week high prices over next 6 to 12 months. This return predictability is attributed to investors’ “anchoring and adjustment” bias (Tversky and Kahneman, 1974) that causes them to underreact to positive (negative) information about stocks with prices near (far from) their 52-week high prices.1 More recent research also examines return patterns associated with stocks’ historical or all-time high prices. Interestingly, this research documents an opposite pattern: the stocks trading near their historical high prices underperform, consistent with investor overreaction.2 Li and Yu (2012) document such reversals at both the aggregate market level for Dow Jones Industrial Average, and find consistent evidence in the cross-section of individual stocks. Lee and Piqueira (2017) study short-selling behavior around 52- week high and historical high prices, and find that short-selling is positively associated with proximity to the historical high, and negatively associated with proximity to the 52-week high. Thus, they conclude that their evidence is consistent with short-sellers exploiting the mispricing resulting from investors’ anchoring bias. The goal of this paper is to examine the impact of low-priced stocks and January seasonality on return predictability associated with these psychological price anchors. It is common in the momentum literature to exclude low-priced stocks to avoid microstructure biases. Bhootra (2011) shows that the strong reversal experienced by loser penny stocks over short horizons reduces the momentum strategy returns significantly. Therefore, inclusion of penny stocks in the sample can alter the inferences, especially in case of equally-weighted portfolio returns. The effect of January seasonality is also well-documented. The momentum profits are significantly higher when January returns are excluded due to the large positive returns of loser stocks in January. We find that with either of these adjustments, excluding the stocks priced below $5 or excluding January returns, the stocks trading near historical high prices significantly outperform the ones trading far below their historical high prices. In our sample, we also find that in the presence of these sub-$5 stocks, the momentum strategy based on 52-week high price does not earn a significant positive all-months raw return, on average.3 However, when both low-priced stocks and January returns are excluded, a long-short strategy based on decile portfolio sorts earns an average monthly return of 0.93% (t-statistic = 5.27) in case of the historical high price, and 1.64% (t-statistic = 7.72) in case of the 52-week high price.