فروش استقراضی توسط سرمایه گذاران انفرادی: بی ثبات سازی و یا کشف قیمت ؟

This paper examines how individual investors' participation in short sale affects the efficiency of stock pricing using a unique regulatory change in Korea. The change enables individual investors to sell short some – but not all – domestic stocks, without affecting the short-selling ability of institutions. We find no evidence that individuals' participation in short sale destabilizes stock market. Specifically, our difference-in-difference estimates indicate that stocks show little change in their return volatility or skewness after they become shortable by individuals. Moreover, we find that stocks are traded within a narrower bid–ask spread and deviate less from the random-walk process after becoming shortable by individuals. Overall, our results suggest that at least some individual investors are privy to private information and they contribute to more efficient pricing via their short sales.

Since the path-breaking works by Black (1986) and De Long et al. (1990), there has been continued interest in whether “noise” can affect stock price. It is exactly in this context that individual investors started to attract academic attention. Allegedly more subject to psychological biases and unexpected liquidity shocks, individual investors have often been characterized as a natural empirical proxy for noise traders. Subsequent studies have shown that individual investors indeed trade for non-fundamental reasons and their trades are capable of affecting or destabilizing stock prices (e.g., Lee et al., 1991, Odean, 1998a, Odean, 1998b, Odean, 1999, Kumar and Lee, 2006 and Barber et al., 2009). Despite these findings, viewing individual investors as a whole as noise traders may require further scrutiny. As some authors convincingly argue, not all individuals are made alike and thus at least some individual investors might contribute to pricing efficiency (e.g., Coval et al., 2005, Dhar and Zhu, 2006, Griffin and Zhu, 2006 and Nicolosi et al., 2009). Short sales are particularly instructive in this regard, since they are largely motivated by private information (e.g., Boehmer et al., 2008). By focusing on short sales by individual investors, one can thus identify a subset of individuals who believe that they have private information, and then examine how their trades affect stock prices. In this paper, we examine a recent regulatory change in the Korean stock market, which provides a rare opportunity to isolate short sales by individual investors. The change, which occurred in January 2008, is to resume the share-lending business to individual investors 22 years after this business had been banned. As institutional investors have had access to an over-the-counter market in which they can borrow shares directly from one another, this regulatory change does not affect the short-selling ability of institutions. In other words, our “event” creates an ideal setting to compare a stock market with short sale by individual investors to the one without.1 Moreover, the lifting of the short sale constraint applies only to a subset of stocks in the market; consequently, we are able to compare the stocks shortable by individuals with other stocks, both before and after the regulatory change. Such a difference-in-difference framework significantly mitigates the endogeneity issue in making inferences. To examine the effects of individual investors' short sales on stock prices, we examine four variables in a difference-in-difference framework. First, we examine stock return volatility. If individual investors only create noise through short sales, volatility will increase. Since the regulatory change is exogenous, we do not expect the economic fundamentals of the shortable stocks to change. Thus, any increases in volatility must be largely attributed to increases in noise. Second, as short sale transactions speculate on a fall in stock price, we examine stock return skewness to test whether shortable stocks exhibit more negatively skewed returns after they become shortable. Again, unless there is a sudden downward revision in the economic fundamentals of those stocks around the regulatory change, there is no reason to expect a more negative skewness. Rather, changes in skewness – if any – are likely to reflect the speculative influence of individuals' short sale. Third, to the extent that short sales reflect private information, individual investors' short selling might improve price discovery and may lead to a narrower bid–ask bound. Finally, we examine stock return variance ratios for both shortable and non-shortable stocks around the regulatory change. If stocks are more efficiently priced after becoming shortable by individuals, their price changes will be less serially correlated and the variance ratios will be closer to unity. Our results, based on monthly panel data from January 2007 to August 2008, can be summarized as follows. First, we find little change in stock return volatility attributable to the lifting of the short sale constraint on individual investors. While shortable stocks tend to have greater volatility than others in the first place, these stocks do not become more volatile after they actually become shortable. Similarly, we find little change in stock return skewness related to the lifting of the short-sale constraint on individuals, a result that makes it further implausible that individual investors play a detrimental role in the stock market through their short sale transactions. Finally, we find that the bid–ask spread narrows and the variance ratio becomes closer to unity after the stock becomes shortable by individual investors. We ensure the robustness of the above results in many different specifications. First, we employ two-way clustered standard errors in all of our panel regressions. In other words, we take into account both the correlation among the observations for a given firm over time and the correlation among the observations across firms at a given point in time. We also address this issue by modifying the panel dataset in such a way that the cross-time or the cross-firm correlations are eliminated. Specifically, we “collapse” the panel data into a single cross-section using the ratio of the post-event volatility (bid–ask spread, skewness, variance ratio, or the absolute value of one minus variance ratio) to the pre-event mean volatility (bid–ask spread, skewness, variance ratio, or the absolute value of one minus variance ratio). Alternatively, we minimize the cross-firm correlation in the panel dataset by constructing it with the residuals from the market model. The results with those alternative datasets are qualitatively the same as the main panel regression results. As might be expected, policymakers choose stocks with sufficient trading volume as shortable stocks, so that any negative impact of individuals' short sale is minimized. To address this selectivity issue, we identify a “matching non-shortable stock” for each of our sample shortable stocks based on turnover. The panel regressions with those turnover-matched stocks continue to confirm the reduction in bid–ask spread and the smaller deviation from random walks. Finally, the decrease in bid–ask spread is also confirmed by the actual short sale transaction data. Overall, our results show that at least some individual investors are capable of bringing in private information to the market via their short selling, thereby helping the stocks to be priced more efficiently. This suggests that the characterization of individual investors as noise traders may need further verification. Our findings are thus in contrast to those reported in Foucault et al. (2011) who examine both short sales and margin trading in the French stock market and conclude that portraying individual investors as noise traders is warranted. The differences in results can be reconciled if the individuals' short selling is motivated differently than their margin trading. In this regard, it is worth noting, again, that the investors who engage in short sale are particularly more likely to be informed than ordinary investors, and the argument for which our results provide support is that at least a subset of individual investors has private information and this is incorporated into prices through their short sales. This paper proceeds as follows. In the next section, we provide a brief literature review and detail our experimental setting. Section 3 explains our sample and data, and Section 4 reports the empirical results. Section 5 concludes the paper.

Understanding the role of individual investors is important not only from a pure academic perspective but also from a policy perspective. This paper directly examines this issue by addressing the question of whether individual investors, when they engage in short sales, enhance the efficiency of stock pricing or simply add noise to the price. To this end, the paper compares a stock market with short sale transactions by individuals to the one without. This clear-cut experimental setting is created by a unique regulatory change in Korea that enabled individual investors to sell short some – but not all – domestic stocks. Around this regulatory change, we examine: (1) whether the return of a stock is more or less volatile; (2) whether the return of a stock is more or less negatively skewed; (3) whether a stock trades within a narrower or wider bid–ask spread; and (4) whether the return of a stock becomes more or less serially correlated. Using data from January 2007 to August 2008, we find no change in stock return volatility or skewness after the stock becomes shortable by individual investors. However, we find that stocks exhibit a narrower bid–ask spread after they are eligible for short sale by individual investors. Moreover, the stock return becomes less serially correlated and shows a pattern closer to random walks. Our study adds to both the literature on individual investors and the one on short sales. By linking the two strands of literature, we thus provide further insight with respect to the ongoing debate on whether short sales, especially those by individual investors, are detrimental to market stability. The findings in this study suggest that perhaps such concerns may be overstated. Moreover, our findings indicate that some individual investors – that is, those who engage in short sales – are capable of bringing private information to the market. Indeed, this interpretation is consistent with the recent argument that heterogeneity exists among individual investors and some individual investors can play a beneficial role in stock market (e.g., Coval et al., 2005, Dhar and Zhu, 2006, Griffin and Zhu, 2006 and Nicolosi et al., 2009).