A large body of evidence demonstrates the persistence of numerous anomalies in stock prices, which suggests they can depart from fundamentals for periods of time. The anomalies include:
- Failure Probability: Stocks with a high probability of failure have lower future returns.
- O-score: Stocks with higher O-scores (a higher probability of bankruptcy) have lower future returns compared with stocks with lower scores.
- Net Stock Issuances: The stocks of firms that issue equity underperform the stocks of nonissuers.
- Composite Equity Issuance: Firms with higher equity issuance underperform firms with lower measures.
- Accruals: Stocks with high accruals underperform stocks with low accruals.
- Net Operating Assets: Stocks with higher net operating assets underperform stocks with lower net operating assets.
- Momentum: Stocks with higher past performance outperform stocks with lower past performance.
- Gross Profitability: Stocks with higher gross profitability have higher future returns.
- Asset Growth: Stocks with higher asset growth have lower future returns.
- Return on Assets: Stocks with higher return on assets have higher future returns.
- nvestment-to-Assets: Stocks with higher past investment (scaled by total assets) have lower future returns.
Evidence that these anomalies persist indicates so-called smart money is not fully able to erase mispricings, which are created by “dumb money.” The persistence of these anomalies—despite the large number of hedge fund strategies that trade on various anomalies documented in the academic literature—is a significant puzzle.
Limits To Arbitrage
As anomalies became common knowledge, we would expect them to vanish. However, the “limits-to-arbitrage” literature provides several explanations for why the anomalies may not completely vanish:
- Many institutional investors (such as pension plans, endowments and mutual funds) are prohibited by their charters from taking short positions.
- Shorting can be expensive. You have to borrow a stock to go short, and many stocks are costly to borrow, because supplies of available stock from institutional investors are low. The largest anomalies tend to occur in small stocks (which are costly to trade in large quantity, both long and especially short) because the share volume available to borrow is limited and borrowing costs are often high. In other words, market frictions prevent prices from being fully corrected.
- Investors are unwilling to accept the risks of shorting because of the potential for unlimited losses. Even traders who believe a stock’s price is too high know they can be correct in the long term (the price may eventually fall), but still face the risk that the price will go up in the short term before it eventually goes down. Such a price move, requiring additional capital, can force the traders to liquidate at a loss.
Ferhat Akbas, Will Armstrong, Sorin Sorescu and Avanidhar Subrahmanyam, authors of “Smart Money, Dumb Money, and Capital Market Anomalies,” which appears in the November 2015 issue of the Journal of Financial Economics, contribute both to the literature and to our understanding of why anomalies occur and persist even after their discovery.
They used flows into retail mutual funds as a proxy for “dumb money,” and flows into hedge funds as a proxy for “smart money.” Their use of these proxies is based on the findings in the literature, which include:
- Retail investors tend to “chase performance” by directing money to mutual funds with strong recent performance while failing to redeem capital from funds with poor recent performance.
- Retail investors tend to direct “dumb money” to mutual funds that hold overvalued stocks.
- When mutual fund managers receive new flows from retail investors, they usually increase positions in existing stock holdings. As a result, net money inflows are associated with higher contemporaneous returns and subsequent return reversal.
- High-performing mutual funds tend to attract relatively higher flows, which are then reinvested by fund managers into their existing stock holdings. Similarly, mutual funds with poor performance tend to liquidate existing holdings to meet redemptions. Price pressure from the purchases of recent winners (or liquidation of recent losers) causes return continuation (helping explain the momentum anomaly).
Flows, Returns & Anomalies
Motivated by these findings, the authors examined the intertemporal relationship between two time-series: the aggregate mutual fund flows, and an aggregate measure of monthly cross-sectional equity mispricing that includes the 11 well-known equity anomalies listed above.
- They also examined the relationship between flows and returns to each individual anomaly. The period covered by the study is from 1994 through 2012. Following is a summary of their findings:
- Aggregate flows to mutual funds (“dumb money”) appear to exacerbate cross-sectional mispricing, particularly for growth, accrual and momentum anomalies. The mispricing increases along with mutual fund flows. Mutual fund flows accentuate the mispricing of overvalued stocks (they are “dumb” money) because these stocks go up in value during months when money flows into mutual funds at the aggregate level. The mispricing tends to occur mostly in the “short” leg of a strategy that goes long undervalued stocks and goes short overvalued stocks. Among mispriced stocks, those that are overvalued are more mispriced than those that are undervalued.
- Mutual fund inflows exacerbate overvaluation because they invest new flows into stocks that already are overvalued. The overvaluation eventually is corrected (reverses). However, when facing redemptions, mutual fund managers tend to sell stocks “on the right side,” meaning they sell stocks that are the most overvalued. This does not necessarily imply that mutual fund managers are “smart” when it comes to sales, as the effect could be caused by managers being restricted to selling the stocks held in their portfolio, which the inflow results suggest are overvalued. Unfortunately, the mispricing induced by new investments is stronger than the correction induced by redemptions.
- The dollar volume of new investments is higher (in absolute value) than that of redemptions, helping to explain why mutual fund flows, in the aggregate, have an exacerbating effect on mispricing. The new investment effect dominates the redemption effect, both in terms of volume and intensity.
- Hedge fund flows appear to be invested in a manner that corrects mispricing along the same dimensions in which mispricing is exacerbated by mutual fund flows. In other words, the effect of hedge fund flows on mispricing is significantly positive, because hedge fund flows exert the “right” type of price pressure (they are “smart money”) on mispriced stocks. This is the type of price pressure that brings price convergence toward fundamental value and corrects cross-sectional mispricing.
- Hedge funds’ “corrective” effect is driven primarily by overvalued stocks. Aggregate hedge fund flows appear most effective when they take short positions in overvalued stocks rather than long positions in undervalued stocks. This result is consistent not only with the mutual fund results, but also with the literature that shows short transactions are generally more informed.
- Anomalies related to growth, momentum and accruals are the main channels through which mutual fund flows exacerbate mispricing, while hedge fund flows tend to correct mispricing.
- When mutual fund flows are high, stocks in the undervalued “long” portfolio become more undervalued, while the stocks in the overvalued “short” portfolio become more overvalued. The opposite result is obtained with hedge funds. When hedge fund flows are high, stocks in the undervalued portfolio become less undervalued, while the stocks in the overvalued portfolio become less overvalued.
- Mispricing is more prone to correction during bear markets—as opposed to bull markets—during periods in which small stocks underperform large stocks and during periods in which value stocks outperform growth stocks.
- In the case of five of the anomalies listed above (the O-score, failure probability, gross profitability, return on assets and net stock issuance), mispricings appear to involve primarily “growth” stocks. A characteristic common to “growth” stocks is an unusually high rate of earning growth preceding the portfolio formation period. Thus, investors’ tendency to extrapolate growth rates in past earnings causes them to chase stocks with such “growth” characteristics (precisely the stocks that tend to be the most overvalued). The stocks that are overvalued with respect to the accrual anomaly are also likely to be “glamour” stocks, with earnings partially inflated by the use of accruals.
- New mutual fund flows are directed primarily to the purchase of stocks with “growth” characteristics.
- Using the Fama-French three-factor model (beta, size and value), the average monthly excess return above the risk-free rate on the long portfolio (which buys undervalued stocks) is +138 basis points. The average monthly excess return on the market portfolio is +52 basis points. The average monthly excess return to the short portfolio is -60 basis points. The monthly return to the long/short strategy is +198 basis points, and is highly statistically significant.
- When disaggregating mutual fund flows into index and nonindex funds, the effect on mispricing was significant for nonindex and insignificant for index funds, as well as of a smaller magnitude (about one-tenth the size). Thus, nonindex funds are the main driver of the mispricing. Similarly, the effect on mispricing from retail mutual fund flows is much greater than from institutional mutual fund flows (although it is still negative).
- The effects of mutual fund flows on aggregate mispricing are higher during periods when the VIX (a measure of volatility, referred to as the fear factor) is high, as well as during periods of high investor sentiment. Investor sentiment is the propensity of individuals to trade on noise and emotions rather than facts.
Hedge Funds Vs. Mutual Funds
The authors concluded: “Our results suggest that aggregate flows to mutual funds may have real adverse allocation effects in the stock market [and therefore in the allocation of capital], while aggregate flows to hedge funds contribute to the correction of cross-sectional mispricing.”
They added this hypothesis: “Our analysis thus suggests that despite cross-sectional return predictability now being common knowledge among sophisticated investors, judicious strategies that seek to exploit this predictability should continue to earn positive alphas so long as ‘dumb’ money continues to enter the stock market via the mutual fund industry. Indeed, an aggregate consequence of this ‘dumb’ money is to create a market for ‘smart money,’ hedge funds, where investors can earn alpha by merely trading against the price pressure induced by these ‘dumb’ flows.”
I would add this observation: While the evidence presented above provides support for the assertion that hedge funds provide a useful service by reducing the effects of mispricing created by “dumb money” (thereby improving the allocation of capital), the evidence is also very strong that hedge fund investors have not benefited from the skills exhibited by the hedge fund sponsors. While the sponsors have created wealth for themselves, they have delivered poor risk-adjusted returns to their investors.
Larry Swedroe is the director of research for The BAM Alliance, a community of more than 140 independent registered investment advisors throughout the country.