Research

Using a new and comprehensive sample of more than 87 million Twitter posts referencing Russell 3000 firms between 2012 and 2022, we introduce a novel, unsupervised method of scoring the sentiment of emojis. Our method generates point-in-time dictionaries that map individual emojis to the contextual sentiment of recent tweets that contain them. In out-of-sample tests, we find that even controlling for the sentiment extracted from words, news, and corporate events, emoji sentiment correctly predicts future firm-level stock returns. Importantly, we show a newly emergent generation of Twitter users drive emoji-based return predictability, while more experienced users better predict returns using words. Understanding the sentiment of emojis has become increasingly important as individuals and market professionals continue to adopt these new forms of communication.

We exploit an exogenous shock to dark trading volume in order to identify the effect of dark trading on market quality. Following a 34% reduction in dark trading volume, we find that the cost of trade (e.g., effective spreads, realized spreads, price impact, and quoted spreads) does not change. While our findings stand in contrast to those of prior studies, a number of supplemental tests confirm that conflicting inferences cannot be attributed to different stock samples or time periods. Our research highlights the benefit of structured experimentation from the Securities and Exchange Commission (SEC) for understanding causal effects in capital markets. 

We examine the economics that underlie retail trading costs around discount brokers’ widespread adoption of zero commission trading in October 2019. Our analysis of participating brokers’ Rule 606 filings and financial statements reveals little change in payment for order flow, which suggests brokers absorbed the cost of eliminating commissions in a competitive environment. We then perform a difference-in-differences analysis of effective spreads and report economically trivial changes in retail execution costs around the commission change. Finally, we assess the total trading costs of an aggregate retail portfolio compared to a host of counterfactuals. We find that following the zero-commission change, total retail transaction costs dropped substantially even under the extreme counterfactual that these traders pay exchange quoted spreads and receive zero price improvement. Our findings support the brokerage industry’s claim that dropping commissions helped retail investors and should ease regulators’ concerns to the contrary. 

We present evidence that short sellers alternate between stock picking during expansions and market timing during recessions. First, firm-level short interest is a much stronger negative predictor of the cross-section of stock returns during expansions than it is during recessions. High short interest also only predicts negative future earnings announcement returns during expansions. We attribute these findings to short sellers’ emphasis on collecting firm-specific signals. Second, short sellers appear to make factor bets more so during recessions than in expansions. These bets tend to pay off as we observe a strong negative relation between the betas of highly shorted stocks and future stock market returns, a result which disappears during expansions. Together, these findings are consistent with theories of information acquisition under attention constraints, endogenous information production, as well as theories of time variation in aggregate overconfidence amongst traders.

In a standard stock loan, the borrower reimburses the lender any dividends paid while the loan is outstanding. Since these substitute dividends may be taxed differently than dividend payments themselves, some investors have incentives to either remove their shares from lendable supply – if they pay high taxes on substitute dividends – or lend out their shares to arbitrageurs – if they pay high taxes on dividends. Consistent with these incentives, we find a significant tightening of the equity lending market on dividend record days driven by both a contraction of supply and an expansion of demand – although the demand effect appears to dominate. We then exploit the plausibly exogenous nature of these shifts to causally link tightness in the lending market to wider effective spreads in the stock market.

Using proprietary data on millions of trades by retail investors, we provide the first large-scale evidence that retail short selling predicts negative stock returns. A portfolio that mimics weekly retail shorting earns an annualized risk-adjusted return of 9%. The predictive ability of retail short selling lasts for one year and is not subsumed by institutional short selling. In contrast to institutional shorting, retail shorting best predicts returns in small stocks and those that are heavily bought by other retail investors. Our findings are consistent with retail short sellers having unique insights into the retail investor community and small firms’ fundamentals.

We analyze the role of retail investors in stock pricing using a database uniquely suited for this purpose. The data allow us to address selection bias concerns and to separately examine aggressive (market) and passive (limit) orders. Both aggressive and passive net buying positively predict firms’ monthly stock returns with no evidence of return reversal. Only aggressive orders correctly predict firm news, including earnings surprises, suggesting they convey novel cash flow information. Only passive net buying follows negative returns, consistent with traders providing liquidity and benefiting from the reversal of transitory price movements. These actions contribute to market efficiency.

Empirical methodologies used to assess intraday price efficiency are described. The metrics presented here summarize the extent to which actual prices deviate from an underlying efficient price and, therefore, attest to overall market quality. Two different approaches are discussed, and their relative merits are evaluated. The first approach uses a vector autoregression model over transaction‐based returns and trade information. The second one uses simple autocorrelations of intraday midquote returns. Because it explicitly decomposes price changes into permanent (efficient) versus temporary (pricing error) components and it is set in transaction time as opposed to clock time, the first approach is preferred. 

Using a broad panel of NYSE-listed stocks between 1983 and 2004, we study the relation between institutional shareholdings and the relative informational efficiency of prices, measured as deviations from a random walk. Stocks with greater institutional ownership are priced more efficiently, and we show that variation in liquidity does not drive this result. One mechanism through which prices become more efficient is institutional trading activity, even when institutions trade passively. But efficiency is also directly related to institutional holdings, even after controlling for institutional trading, analyst coverage, short selling, variation in liquidity, and firm characteristics.

Reversal is the current stylized fact of weekly returns. However, we find that an opposing and long-lasting continuation in returns follows the well-documented brief reversal. These subsequent momentum profits are strong enough to offset the initial reversal and to produce a significant momentum effect over the full year following portfolio formation. Thus, ex post, extreme weekly returns are not too extreme. Our findings extend to weekly price movements with and without public news. In addition, there is no relation between news uncertainty and the momentum in 1-week returns.

In spite of the growing research concerning investor protection, the relation between investor protection and real investment by foreign multinationals is largely unexplored. Recognizing this relation, however, is especially important in light of the surge in cross-border activity in recent decades and the potential impact cross-border investment can have on a country's economic development. We find that U.S. multinational foreign investment is significantly greater both when shareholder protection is poor and when creditor protection is poor. Consistent with existing literature, our results suggest that U.S. firms have greater comparative advantages when local firms in poor investor protection countries either i) invest suboptimally due to agency problems or ii) have constrained access to debt capital. The increased investment by U.S. multinational in poor investor protection countries is of particular interest, because it suggests an important way in which adverse outcomes related to poor investor protection may be mitigated.