Ronnie Sadka

Liquidity and the Post-Earnings-Announcement Drift

The post-earnings-announcement drift is a longstanding anomaly that conflicts with market efficiency. This study documents that the post-earnings-announcement drift occurs mainly in highly illiquid stocks. A trading strategy that goes long high-earnings-surprise stocks and short low-earnings-surprise stocks provides a monthly value-weighted return of 0.04 percent in the most liquid stocks and 2.43 percent in the most illiquid stocks. The illiquid stocks have high trading costs and high market impact costs. By using a multitude of estimates, the study finds that transaction costs account for 70–100 percent of the paper profits from a long–short strategy designed to exploit the earnings momentum anomaly. One of the most persistent anomalies that seem to violate the semi-strong-form market efficiency as defined by Fama is the post-earnings-announcement drift (PEAD), or earnings momentum. This anomaly refers to the fact that companies reporting unexpectedly high earnings subsequently outperform companies reporting unexpectedly low earnings. More specifically, a company’s standardized unexpected earnings (SUE) is defined as the difference between the last available quarterly earnings and the earnings during that same quarter in the previous year, scaled by the standard deviation of this difference over the previous eight quarters. A trading strategy that each month goes long the stocks in the top decile of SUE and short the stocks in the bottom decile of SUE earns, on average, 90 bps per month (10 percent annually) over the 1972–2005 period. The goal of this article is to demonstrate that stock liquidity is an important consideration for understanding the persistence of the PEAD anomaly over the years. Previous studies have not taken trading costs into account in the calculation of abnormal returns. We studied the impact of illiquidity on the profitability of the PEAD trading strategy and show that this strategy is likely to be unprofitable after adjusting for transaction costs. First, we studied the relationship between the PEAD and illiquidity by using double-sorted portfolios. Our findings suggest that the PEAD is prevalent mainly in illiquid stocks. We examined the profitability of the long–short SUE strategy after sorting stocks into decile portfolios on the basis of their illiquidity. For this analysis, we used the Amihud measure of stock illiquidity, which is the average of the daily price impacts of the order flow (i.e., the daily absolute price change per dollar of daily trading volume). Returns to the long–short SUE strategy increased monotonically from 0.04 percent per month for the most liquid stocks to 2.43 percent for the most illiquid stocks. Because we found that the PEAD is more prevalent in illiquid stocks, following a PEAD trading strategy will generate high transaction costs and a substantial price impact. We used several transaction-cost estimates to calculate the net returns to PEAD trading strategies. Our results show that transaction costs consume 70–100 percent of the potential profits. This lack of profitability can thus explain the persistence of the PEAD anomaly and is consistent with Jensen’s definition of market efficiency and Rubinstein’s definition of minimally rational markets.

Seasonality in the Cross Section of Stock Returns: The International Evidence

This paper studies seasonal predictability in the cross section of international stock returns. Stocks that outperform the domestic market in a particular month continue to outperform the domestic market in that same calendar month for up to 5 years. The pattern appears in Canada, Japan, and 12 European countries. Global trading strategies based on seasonal predictability outperform similar nonseasonal strategies by over 1% per month. Abnormal seasonal returns remain after controlling for size, beta, and value, using global or local risk factors. In addition, the strategies are not highly correlated across countries. This suggests they do not reflect return premiums for systematic global risk.

Hedge-Fund Performance and Liquidity Risk

This paper demonstrates that liquidity risk as measured by the covariation of fund returns with unexpected changes in aggregate liquidity is an important predictor of hedge-fund performance. The results show that funds that significantly load on liquidity risk subsequently outperform low-loading funds by about 6.5% annually, on average, over the period 1994-2009, while negative performance is observed during liquidity crises. The returns are independent of share restriction, pointing to a possible imbalance between the liquidity a fund offers its investors and the liquidity of its underlying positions. Liquidity risk seems to account for a substantial part of hedge-fund performance. The results suggest several practical implications for risk management and manager selection.

Smart Money or Smart about Money? Evidence from Hedge Funds

This paper introduces a measure of fund-flow impact, based on a funds contemporaneous return-flow relation, and demonstrates that the smart-money phenomenon predominantly stems from high-flow-impact funds. A smarter-money strategy, one that concentrates in high-flow-impact funds, fares significantly better than a strategy which concentrates in low-flow-impact funds and earns a premium of 6.5% annually over 1999-2008, after controlling for various risk factors and trading restrictions. Although the effect is strongly apparent for outflows, the analysis shows that a smarter-money-conscious long-only-investment portfolio significantly outperforms the hedge-fund index. The paper suggests that the smart-money effect is not necessarily indicative of superior investor ability to predict manager skill, but rather may reflect the ability of some investors to predict the behavior of other investors whose flow affects fund return.

Liquidity Risk and Mutual Fund Performance

This paper demonstrates that the ability of fund managers to create value depends on market liquidity conditions, which in turn introduces a liquidity risk exposure (beta) for skilled managers. We document an annual liquidity beta performance spread of 4% in the cross section of mutual funds over the period 1983–2014. Liquidity risk premia explain an insubstantial fraction of this spread; instead, the spread can be attributed to the differential ability of high liquidity beta funds to outperform across high and low market liquidity states, due to a differential rate of either mispricing correction or intensity of informed trading. Tests based on mispricing, proxied by a comprehensive set of 68 anomalies, and tick-by-tick trades, from a large proprietary institutional trading data set, corroborate the contribution of these channels. The results highlight the interaction between informed investors, mispricing, and liquidity beta. The Internet appendix is available at https://doi.org/10.1287/mnsc.2017.2851. ...

Has the U.S. Stock Market Become More Vulnerable Over Time

This paper demonstrates that the cross-sectional variation of systematic risk and systematic liquidity have increased over the period 1963-2008. Both have increased signi ficantly for large-cap firms, but declined signifi cantly for small-cap fi rms. Several implications for investment managers are discussed, such as the declining ability to diversify return volatilities and liquidity shocks by holding liquid, large-cap stocks. The evidence suggests that the vulnerability of the US equity market to unanticipated events has increased over the past few decades.

Are You Trading Predictably

The authors find predictable patterns in stock returns. Stocks whose relative returns are high in a given half hour today exhibit similar outperformance in the same half hour on subsequent days. The effect is stronger at both the beginning and the end of the trading day. These results suggest that strategically shifting the timing of trades can significantly reduce execution costs for institutional traders. Anecdotal evidence suggests that some institutional traders concentrate trades at particular times of the day. For example, index funds may execute market-on-close orders to minimize tracking error relative to their benchmarks. Active managers may choose to hand trades to the trading desk at particular times of the day. In addition, trading algorithms that imply particular time-of-day trading patterns under certain assumptions have been proposed. A separate literature suggests that flows of funds to institutional managers are autocorrelated and that institutional trading is highly persistent—that is, institutional investment managers tend to buy or sell the same stocks on successive days. These two observations (time-of-day trading patterns and autocorrelated fund flows) suggest the existence of time-of-day patterns in both volume and order imbalances. If these patterns were fully anticipated by traders, however, one would not expect time-of-day patterns in stock prices. But we found persistent patterns in stock returns: Stocks whose relative returns are high in a given half hour today tend to exhibit similar outperformance in the same half hour on subsequent days. Although the effect is stronger at the beginning and the end of the trading day, it exists throughout the day. This periodicity is also stronger for small-cap stocks, but it exists for both large and small companies. The magnitude of the return pattern is sizable relative to several components of transaction costs, including commissions and effective spreads. The return patterns that we documented are consistent with investors’ predictable trading patterns. At a minimum, randomizing trades, rather than trading predictably, should help investors avoid buying high and selling low. More strategically, shifting trades to certain periods can substantially reduce execution costs for institutional traders.

Media and Investment Management

This paper studies the relation between the media coverage of funds and their future performance. We classify news items about equity hedge funds over 1999--2008 into three source groups: General newspapers, Specialized magazines, and Corporate Communication. Examining post-exclusive-coverage performance, we document that Corporate-covered funds outperform and General-covered funds underperform, with a performance difference of about 11% annually. Applying a textual analysis to news items, we find that sentiment-related biases do not explain the inter-source return spread. Nevertheless, investor fund flow does not differentially respond to source-based information. The results suggest that the source-based return spread may reflect the extensive costs of processing information across thousands of media sources to generate alpha.

Implied Cost of Capital in the Cross-Section of Stocks

Recent research shows that the implied cost of capital (ICC), measured from analyst forecasts and current stock prices, predicts market returns. This paper studies the cross-section of stocks and finds that ICC negatively predicts returns. An investment strategy that goes long low-ICC stocks and short high-ICC stocks provides an alpha of 6% per year. Evidence suggests that the negative relation is due to the fact that stocks with a high level of ICC are systematically related with overly optimistic earnings forecasts. High-ICC stocks are also associated with a low probability of survival. Investors fail to incorporate this bias, leading to more negative earnings surprises. The findings highlight the need to exercise caution when using ICC as a measure of cost of capital for individual firms.

Illiquidity and Earnings Predictability

This paper studies whether illiquidity affects the predictability of fundamental valuation variables. Firm-level, cross-sectional analyses show that returns of illiquid stocks contain less information about their firms future earnings growth compared to those of more liquid stocks. A natural experiment utilizing an exogenous variation in liquidity amid the reduction of tick size on the NYSE indicates that an improvement in liquidity causes an increase in earnings predictability. At the aggregate level, stock returns contain less information about future growth in aggregate earnings, GNP, and industrial production during illiquid periods. The results highlight the importance of liquidity for forecasting fundamentals and stock-price efficiency.