TIME-VARYING FACTORS AND CROSS-AUTOCORRELATIONS IN SHORT-HORIZON STOCK RETURNS

In this paper I show that the lead-lag pattern between large and small market value portfolio returns is consistent with differential variations in their expected return components. I find that the larger predictability of returns on the portfolio of small stocks may be due to a higher exposure of these firms to persistent (time-varying) latent factors. Additional evidence suggests that the asymmetric predictability cannot be fully explained by lagged price adjustments to common factor shocks: (i) lagged returns on large stocks do not have a strong causal effect on returns on small stocks; (ii) trading volume is positively related to own- and cross-autocorrelations in weekly portfolio returns; and (iii) significant cross-autocorrelation exists between current returns on large stocks and lagged returns on small stocks when trading volume is high.     

Disposition effect and analyst forecast dispersion

Behavioral finance theories posit that behavioral biases are more pronounced when there is higher information uncertainty about fundamentals. This paper examines the relation between the disposition effect, the tendency to ride losses and realize gains, and dispersion in financial analysts’ earnings forecasts for a sample of large U.S. discount brokerage accounts from January 1991 to December 1996. I find that the disposition effect is exacerbated in stocks with higher analyst forecast dispersion. In particular, the disposition effect is 10% in stocks in the highest forecast dispersion quintile and not significant in the lowest forecast dispersion quintile. The driving factor behind these findings is investors’ higher propensity to realize gains when facing higher information uncertainty. The results are robust to controlling for firm size, analyst coverage, idiosyncratic volatility, turnover, and past market-adjusted returns. The results provide supportive evidence for a behavioral bias explanation of the disposition effect consistent with mean-reversion beliefs for winners and loss actualization avoidance for losers.     

Biases in computed returns: An application to the size effect

Previous estimates of a ‘size effect’ based on daily returns data are biased. The use of quoted closing prices in computing returns on individual stocks imparts an upward bias. Returns computed for buy-and-hold portfolios largely avoid the bias induced by closing prices. Based on such buy-and-hold returns, the full-year size effect is half as large as previously reported, and all of the full-year effect is, on average, due to the month of January.     

Short Selling and the Weekend Effect for NYSE Securities

Using short-sale transactions data, we examine the relation between short selling and the weekend effect. We do not find that short selling is more abundant on Monday than on Friday, even for stocks that have higher Friday returns. We find that short sellers execute more short-sale volume during the middle of the week, and that the positive correlation between short selling and returns on Monday is greater, on average, than the correlation on the other days of the week. Our results are robust to subsamples of stocks with larger weekend effects and stocks that do not have listed options.     

The survivorship bias, share price effect, and small firm effect in Canadian markets

After controlling for survivorship bias, we examine the relation between average returns, firm size, and price levels for Canadian stocks during the 1975-1994 period. Our findings indicate that there is a significant inverse share price level effect in Canadian markets. When we compare the results of the overall sample with the groups of surviving firms and delisted stocks, the latter group shows strong performance for large-size, high-priced stocks. Evidence that supports an independent size effect is less clear for Canadian stocks. A small size effect exists only among the higher share price denominations, which suggests a confounded size-price effect. Although the delisted group returns are statistically different from those of the survivor and the overall groups, which implies some evidence of survivorship bias, the difference between the survivor group and the overall group is weak at best.     

Is momentum trading profitable from Shari'ah compliant stocks?

This research offers fresh evidence supporting the pervasiveness of the momentum effect. Two decades after the momentum profitability firstly documented by Jegadeesh and Titman (1993), yet little research has been specifically devised for the momentum profitability on Shari'ah compliant stocks. We assess the momentum profitability over the Shari'ah compliant stocks in a Malaysian setting. We find evidence of strong return persistence as far as toward four-year holding period. Interestingly, no significant momentum returns are found among the conventional stocks. Upon further exploration we find neither an industry-driven momentum effect nor the small size firms can account for the momentum returns. Using return persistency formation criteria, we further find that underreaction seems to well fit in explaining this unique long lasting momentum profitability.     

The turn-of-the-year effect and tax-loss-selling by institutional investors

Prior studies attribute the turn-of-the-year effect whereby small capitalization stocks earn unusually high returns in early January to tax-loss-selling by individual investors and window-dressing by institutional investors. My results suggest that a significant portion of the effect on turn-of-the-year returns that prior studies attribute to window-dressing is actually attributable to tax-loss-selling by institutional investors. Among small capitalization stocks, I find that institutional investors with strong tax incentives and weak window-dressing incentives realize significantly more losses in the fourth quarter than in the first three quarters of the calendar year, and that their fourth quarter realized losses have a significant impact on turn-of-the-year returns. A one percentage point change in these institutional investors' fourth quarter realized losses scaled by a firm's market capitalization results in an increase of 47 basis points in the firm's average daily return over the first three trading days of January, which represents a 46 percent change for the mean firm.     

Idiosyncratic risk and the cross-section of expected stock returns

Theories such as Merton [1987. A simple model of capital market equilibrium with incomplete information. Journal of Finance 42, 483–510] predict a positive relation between idiosyncratic risk and expected return when investors do not diversify their portfolio. Ang, Hodrick, Xing, and Zhang [2006. The cross-section of volatility and expected returns. Journal of Finance 61, 259–299], however, find that monthly stock returns are negatively related to the one-month lagged idiosyncratic volatilities. I show that idiosyncratic volatilities are time-varying and thus, their findings should not be used to imply the relation between idiosyncratic risk and expected return. Using the exponential GARCH models to estimate expected idiosyncratic volatilities, I find a significantly positive relation between the estimated conditional idiosyncratic volatilities and expected returns. Further evidence suggests that Ang et al.'s findings are largely explained by the return reversal of a subset of small stocks with high idiosyncratic volatilities.     

Media tone and expected stock returns

We find that media tone reflects firm-level expected returns—firms with low-negative tone stories over a few months earn higher returns in the medium to long term than do firms with high-negative tone stories. The tone premium is driven by consistent outperformance of low-negative stocks, while high-negative stocks do not produce significant abnormal returns. The media tone effect is associated with some common risk factors, among which the size factor plays a consistent and most significant role. The tone effect also reflects differences in firms' idiosyncratic risk, and there is evidence that it is partially related to mispricing and limits to arbitrage.     

Institutional investors,intangible information,and the book-to-market effect

This paper establishes a robust link between the trading behavior of institutions and the book-to-market effect. Building on work by Daniel and Titman (2006), who argue that the book-to-market effect is driven by the reversal of intangible returns, I find that institutions tend to buy (sell) shares in response to positive (negative) intangible information and that the reversal of the intangible return is most pronounced among stocks for which a large proportion of active institutions trade in the direction of intangible information. Furthermore, the book-to-market effect is large and significant in stocks with intense past institutional trading but nonexistent in stocks with moderate institutional trading. This influence of institutional trading on the book-to-market effect is distinct from that of firm size. These results are consistent with the view that the tendency of institutions to trade in the direction of intangible information exacerbates price overreaction, thereby contributing to the value premium.     

The smallest firm effect: An international study

Using a carefully screened and filtered international database with a wide coverage across countries and size classes, this paper identifies and documents a post-1980s size effect which is persistent, not picked up by a Fama–French-style SMB, and largely due to the smallest-decile stocks. We test for potential explanations (such as market risk, infrequent trading, financial distress risk, missing book values, momentum, liquidity risk, changing business conditions, January effect, exchange risk, time-varying risk loadings and dividend yield effects), but none can quite explain the international size effect, whether separately or jointly. Fully identifying the missing risk factor is beyond the scope of this paper but we do find that dividend yield shows up as a significant characteristic in the cross-section of risk-adjusted returns, even after controlling for time-varying risk loadings linearly related to dividend yield. When we construct two ad-hoc risk factors that jointly capture the documented size effect, and then correlate these factors with characteristics-based portfolios, we likewise find that especially dividend yield seems to play an important role in the missing risk factor. More generally, this paper revives the debate on the small-firm effect and, we hope, will stimulate further research on a class of stocks that are too interesting to ignore.     

The puzzling behavior of short sellers around earnings announcements

We examine the performance of ‘predictive’ and ‘reactive’ short sellers who take relatively large short positions immediately before and after quarterly earnings announcements, respectively. While both types short into advancing markets, it is surprising for reactive shorts since their trades are in stocks that just announced unexpected good news and thus, according to the post-earnings announcement drift anomaly, will subsequently have abnormally high cumulative returns. Nevertheless, we find that for both types of short sellers: (1) subsequent cumulative returns are significantly negatively related to the amount of abnormal short selling, suggesting they are informed, and (2) relative to non-earnings dates, the subsequent returns around earnings announcements are significantly more negative, indicating they appear to be adept at exploiting earnings announcements. Surprisingly, we find that the subsequent returns of reactive short sellers are significantly greater than those of predictive short sellers except for S&P 500 stocks, perhaps due to their greater analyst following. Importantly, we are left with two puzzles. First, reactive shorts would have significantly improved their performance had they based their trades on the size of standardized unexpected earnings (‘SUE’). Second, predictive shorts of Micro stocks would have significantly improved their performance had they simply waited until earnings were announced and then based their trades on SUE.     

Attention Spillover in Asset Pricing

Exploiting a screen display feature whereby the order of stock display is determined by the stock's listing code, we lever a novel identification strategy and study how the interaction between overconfidence and limited attention affect asset pricing. We find that stocks displayed next to those with higher returns in the past two weeks are associated with higher returns in the future week, which are reverted in the long run. This is consistent with our conjectures that investors tend to trade more after positive investment experience and are more likely to pay attention to neighboring stocks, both confirmed using trading data.     

Weekend Effect, 'Reverse' Weekend Effect, and Investor Trading Activities

Abstract:   In this study, we document evidence of a 'reverse' weekend effect – whereby Monday returns are significantly positive and they are higher than the returns on other days of the week – over an extended period of eleven years (from 1988 to 1998). We also find that the 'traditional' weekend effect and the 'reverse' effect are related to firm size in that the 'traditional' weekend effect tends to be associated with small firms while the 'reverse' weekend effect tends to be associated with large firms. In addition, we find that during the period in which the 'reverse' weekend effect is observed, Monday returns for large firms tend to follow previous Friday returns when previous Friday returns are positive , but they do not follow the previous Friday returns when Friday returns are negative . Furthermore, we find that during the period in which the 'reverse' weekend effect is observed, Monday returns are positively related to the volume of medium‐size and block transactions, but negatively related to the volume of odd‐lot transactions.     

Online Search Frequency,Retail Investor Overreaction,and the Cross-Section of Stock Returns: Evidence from the Chinese Stock Market

I find a strong negative relation between online search frequency and future returns on the Chinese stock market. I suggest that this effect captures retail investor overreaction to unexpected signals, because online search frequency reflects the efforts made by investors to obtain firm-specific knowledge. The effect is particularly strong in stocks with high information uncertainty (high analyst dispersion, big past earnings surprises, low analyst coverage, and large trading volume), whose intrinsic values are difficult or costly for investors to estimate. Online search frequency as a direct indicator of retail investors’ reaction to signals also sheds light on the idiosyncratic volatility (IVOL) puzzle. I find that this puzzle is more pronounced in high-search-frequency subsamples and disappears in low-search-frequency subsamples. Further evidence shows that high search frequency strengthens the negative IVOL effect in stocks with positive signals but weakens this effect in stocks with negative signals. I suggest that the IVOL puzzle in the Chinese market can be partially explained as a reversal following overreaction to positive signals by retail investors.     

The tax-loss selling hypothesis,market liquidity,and price pressure around the turn-of-the-year

In this paper, we use intra-day data for all stocks listed on the ISSM and provide new and direct evidence consistent with the tax-loss selling hypothesis. We find that (a) there is abnormal selling pressure prior to the year-end for stocks that have experienced large capital losses in the current and prior years (b) investors delay realizing capital gain by postponing the sale of capital gain stocks until after the new year (c) there is a significant decrease in the average trade size for stocks with large capital losses before the year-end and for stocks with capital gains in the new year, which suggests that individuals, rather than institutional investors, are the major sellers around the year-end (d) the tax-loss selling hypothesis, and not firm size or share price, is the fundamental explanation for abnormal January returns. Further, small or low share priced firms with capital gains do not experience abnormal returns in January. However, conditional on capital losses, small or low share priced firms magnify the turn-of-the-year effect (e) On average, the increase in selling activity adversely affects market liquidity by increasing bid-ask spreads and reducing depths. (f) The tax-loss selling pressure not only causes the price to be at the bid at the year-end, it also temporarily depresses the equilibrium price indicating the short run demand curve is not perfectly elastic (g) the year-end buying activity suggests that large investors buy capital loss stocks prior to the year-end to take advantage of the temporarily depressed price and capital gain stocks after the new year to reinvest the proceeds of the tax-loss selling.     

Short-term predictability of equity returns along two style dimensions

This study uses daily return data on 20 portfolios split along two dimensions, growth/value and market size, over the period of four decades and employs over 12,000 trading rules to investigate the short-term predictability of portfolio returns. It shows that, historically, portfolios of small stocks and value stocks have been more suitable for active trading strategies since returns on value portfolios exhibit more predictability than returns on growth portfolios and returns on portfolios of large stocks appear to be less predictive than returns on portfolios of small stocks. The predictive ability of trading rules is all but gone during the 2000s. Popularization of exchange-traded funds and the introduction of quote decimalization on the exchanges are the most likely reasons behind the lack of predictability.     

The 2008 short sale ban: Liquidity, dispersion of opinion, and the cross-section of returns of US financial stocks

This study examines the cross-sectional impact of the 2008 short sale ban on the returns of US financial stocks. Motivated by the large cross-sectional variation in the extent to which banned stocks suffer an illiquidity shock, we hypothesize that stocks with larger liquidity declines are associated with poorer contemporaneous stock returns. The evidence supports this hypothesis and suggests that this effect is stronger for more liquid stocks, as predicted by Amihud and Mendelson (1986). Moreover, consistent with Miller’s (1977) model, we report a valuation reversal whereby stocks with higher abnormal returns at the onset of the ban have lower abnormal returns at its removal. Our findings are robust when we control for firms most affected by TARP, include non-banned matched firms, and compare banned firms’ stock returns with their bond returns. From a policy standpoint, the ban reduced valuations, ceteris paribus, of the stocks that were hardest hit by illiquidity.     

New evidence on the turn-of-the-year effect from closed-end fund IPOs

One of the most persistent securities anomalies is the january effect, whereby significant positive abnormal returns occur during the first few days of the calendar year, especially among small capitalization stocks. I detect a statistically significant January seasonal among a sample of closed-end stock funds that went public during the immediately preceding calendar year. However, contrary to prior research, the results indicate that the abnormal January returns are associated with year-end tax-loss-selling, but do not exhibit a small firm effect.