Search

We study the interaction of short sellers and social media and the effect on stock prices. We use 75.1 million investment-related social media posts for 3,683 unique Chinese firms. Prior to high short interest, social media tone is abnormally positive. Once highly shorted, the tone flips and is abnormally negative. No such pattern exists with traditional media. Compared to firms that are just highly shorted, highly shorted firms with pump-and-dump patterns in social media tone have abnormal returns that are 2.7x higher before, and 3x lower after, the initiation of high short interest. Evidence from natural experiments involving China’s introduction and subsequent suspension of shorting also suggest social media manipulation. Manipulation is more likely in firms located in provinces with weaker legal environments. Our findings show that in the realm of social media, short sellers may profit more by creating mispricing than by correcting it.

We study how 9 different market participants trade with respect to 130 different stock return anomalies and how each participant’s trades predict returns. Retail investors trade against anomalies, while firms’ and short sellers’ trades agree with anomalies. Institutional portfolios are weighted against anomalies, although some institutions’ trades agree with anomalies after the anomaly-measurement date. Retail trades predict returns in the wrong direction, firms’ and short sellers’ trades predict returns in the intended direction, institutional trades do not robustly predict returns. The return-predictability of trades by retail investors, firms and short sellers can be either mostly or completely explained by anomalies.

Firm-level variables that predict cross-sectional stock returns, such as price-to-earnings and short interest, are often averaged and used to predict market returns. Using various samples of cross-sectional predictors and accounting for the number of predictors and their interdependence, we find only weak evidence that cross-sectional predictors make good time-series predictors, especially out-of-sample. The results suggest that cross-sectional predictors do not generally contain systematic information.

Some U.S. firms have women directors and executives, while many do not. We seek to explain this heterogeneity. Using U.S. Census data from 1900, we find that firms in regions with populations originating from countries with more gender-egalitarian beliefs have more women executives and directors, and more women chairing important board committees. These same regions also have more women in the labor market and in STEM occupations, and lower gender-pay gaps. Our findings show that regional differences in beliefs about gender roles are persistent and have significant impacts on corporate leadership and the labor market.

The precautionary motive for finance predicts that firms make large issues to build financial liquidity when capital market conditions are favorable, and issue less or not at all when conditions are unfavorable. The liquidity squeeze framework predicts that firms issue to fund liquidity squeezes, regardless of market conditions. We find that both frameworks help explain the timing of issuance decisions, but only the precautionary framework explains issue size. With respect to timing, the majority of firms facing ex-ante liquidity squeezes do not issue. Squeezed firms issue if capital market conditions are favorable, but cut spending otherwise. With respect to size, large issues are not associated with large liquidity squeezes. Instead, firms raise more capital when issuance conditions are favorable, and use the majority of the proceeds to build financial liquidity, both by increasing cash and by refinancing debt to extend maturity.

Using a sample of 97 stock return anomalies documented in published studies, we find that anomaly returns are 7 times higher on earnings announcement days and 2 times higher on corporate news days. The effects are similar on both the long and short sides, and they survive adjustments for risk exposure and data mining. Moreover, anomaly signals predict errors in analysts’ earnings forecasts — analysts’ forecasts are systematically too low for anomaly-longs and too high for anomaly-shorts. Taken together, our results support the view that anomaly returns are the result of biased expectations, which are at least partially corrected upon news arrival.

Previous studies link equity market liberalization to economic growth in emerging markets. In 24 emerging markets, liberalizations always coincide with other economic reforms, making identification tenuous. Theories linking liberalization to growth predict that at the firm-level liberalization leads to significant foreign ownership and influence on management, more external finance, and higher investment. We fail to find evidence that is consistent with any of these theories. Our findings suggest that either liberalization does not cause growth or that it promotes growth through a channel that is absent from the current literature.

Using patents granted to firms in 59 countries by 76 patent offices to measure innovation, we study the relation between innovation and productivity growth at the country-level. Consistent with endogenous growth theory, patents forecast faster total factor productivity growth. This relation is stronger in countries where patents are widely held by small private firms rather than concentrated in public firms. Relatedly, patents granted by local patent offices have stronger growth effects than those granted by the trilateral patent offices (U.S., Europe, and Japan). Greater intellectual property rights weaken the patent-growth relation, suggesting strong patent protections slow innovation diffusion.

We study the out-of-sample and post-publication return-predictability of 56 characteristics identified in the academic literature as predictors of cross-sectional stock returns. The average out-of-sample decay in return-predictability due to statistical bias to be 15%, but not statistically different from zero. The average post-publication decay, caused by both statistical bias and price pressure from aware investors, is about 50%, and statistically different from both 0% and 100%. These results are not explained by time trends in anomaly returns or measures of attention from the academic literature. Consistent with academic research drawing the attention of informed traders, after an anomaly is published it experiences higher correlations with other anomalies that have been published.

U.S. financial development varies a good deal over the last half century, primarily increasing since the 1980s. We ask whether this variation had consequences for the real economy. Difference-in-difference tests reveal that increases in financial development have disproportionate effects on industries that depend more on external finance. Higher financial development forecasts externally dependent industries using more external finance, having higher turnover of leading businesses, greater variation in firm-growth rates, more new firms entering, more mature firms exiting, lower concentration, and at the aggregate level more innovation and faster growth. The mosaic of our evidence is consistent with a Schumpeterian framework linking the supply of finance to competition, innovation, and growth. Our findings suggest that the growth in finance had some real effects that are socially beneficial.

Using data for emerging market firms that are either investable or uninvestable to foreigners, we study the effects of investability on fundamental volatility, excess volatility, and external finance. We find that fundamental volatility is lower for investable firms, while excess stock return volatility is greater. Share issuance is lower for investable firms, while debt issuance is essentially unchanged. These firm-level findings pose challenges to current theories of equity market liberalization, which contend that investability increases risk sharing, does not increase stock return volatility, and reduces financial constraints.

At year-end, some allege that institutional investors try to mislead investors by placing trades that inflate performance (portfolio pumping) or distort reported holdings (window dressing). We contribute direct tests using daily institutional trades and find that year-end price inflation derives from a lack of institutional selling rather than institutional buying. In fact, institutional buying declines at year end. Consistent with pumping, institutions tend to buy stocks in which they already have large positions. We find no evidence of window dressing, as institutions are not more likely to buy high-past return stocks, or sell low-past return stocks, at year- or quarter-end.

Legal investor protection is associated with how firms choose to issue shares. The likelihood of private placements relative to rights offerings increases with investor protection, as does the likelihood of public offerings relative to both private placements and rights offerings. These findings are consistent with investor protection benefitting minority investors and reducing the benefits of control. Commonly used measures of equity market development are not associated with how shares are issued, nor are measures of market inefficiency. Our study helps shed light on several current issues in the literature, including choice of share issuance mechanism, how investor protection promotes finance, whether investor protection reduces ownership concentration, and the rights offering paradox.

We study the relations between precautionary motives and cash and external finance and cash internationally. It is well established that in the U.S., firms with high precautionary motives hold more cash, have weak internal cash flow, and often build cash with large equity issues. We find similar effects in countries with high access to external finance, but not in those with low access to finance. This difference suggests that in low access countries the benefits of building cash via external finance are outweighed by high issuance costs. Our results show the importance of capital markets with respect to cash management, and are consistent with recent dynamic models linking cash management to capital market conditions.

The recent financial crisis highlights the importance of understanding how financial market conditions impact the real economy. We ask whether access to external finance typically varies over time, and if so what the effects are on investment and employment. Consistent with time-varying external finance costs, we find that both investment and employment are less sensitive to Tobin’s q and more sensitive to cash flow during recessions and periods of low investor sentiment. Share issuance and debt issuance both play roles in causing these effects, although our findings suggest that share issuance plays a bigger role. Alternative tests, which do not rely on q and cash flow sensitivities, support the idea that both low sentiment and recessions increase external finance costs, which in turn limit investment and employment.

We study how investor sentiment affects real investment. We find that both investment and external finance increase with sentiment. The relation between q and investment increases with sentiment, while the relation between cash flow and investment declines with sentiment. These findings suggest that sentiment both creates and alleviates financial constraints. We find that investment in low sentiment states leads to increases in operational efficiency, whereas investment in high sentiment states leads to declines in operational efficiency. Investment and external finance in high sentiment states predict worse stock returns than investment and external finance in low sentiment states. The findings suggest that investor sentiment creates inefficiencies in the allocations of both financial and real resources.

This paper studies the link between momentum and long-term reversals in stock returns. I find that stocks that are making a long-term reversal generate more than 70% of the momentum portfolio's profit. Stocks that have had extreme long-horizons returns but not reversed diminish the momentum portfolio's profit and eventually cause the momentum portfolio to reverse over long-horizons. A combined momentum-reversal strategy has double the profit of momentum, persists over long-horizons and is profitable in both good and bad economic states. These results are consistent with a broad set of behavioral theories, which predict a momentum-reversal process in stock returns.