In The Presence of Persistent Market Anomalies Part 2
Financial markets are incredibly efficient at processing corporate information, such as 10-K reports, product announcements, and management changes, among many others. It also reacts quickly to seemingly random news, like weather events and celebrity tweets. However, as adept as investors are at interpreting these signals, they are far from perfect. The Part 1 article discussed the two primary schools of thought on markets: the Efficient Market Hypothesis vs. Behavioral Finance. In Part 2, we will delve into a series of market anomalies documented by academic researchers, and most importantly, discuss potential strategies to exploit them.
The efficient market hypothesis states that we cannot use past price movements to predict future returns. Professors Jegadeesh and Titman discovered a market anomaly called the “Momentum Effect” that directly contradicts this. In their paper “Returns to Buying Winners and Selling Losers”, they found that significantly positive alpha (excess return over a benchmark, i.e. S&P500) could be achieved by buying stocks of companies that have returned well over the past three months (or beat earnings estimates in the previous quarter), and shorting ones which performed poorly in the past three months. This effect was successful in the short term, about three to twelve months. What this means is that past performance can be used to predict future performance in the short-term, a significant finding when one considers markets to be efficient and stock prices to follow a random walk. However, the strategy returned significantly higher losses during the 2008-09 Financial Crisis (drawback of -77%) than a simple index strategy. This demonstrates that volatile business conditions drastically decrease the effectiveness of a momentum strategy.
Another anomaly is the stock-split effect. This is largely an accounting transaction, where the firm increases or decreases the number of shares on the open market. Management rationale is that the stock price will be lower, thus increasing trading volumes and buy-in attractiveness. For example, a 2-for-1 stock split doubles the number of shares outstanding. An investor who used to hold 10 shares, would hold 20 shares after this split occurred. Since no value has been created, one would expect the stock price to simply halve, i.e. the investor’s 10 shares at $10 are now 20 shares at $5. However, investors often view stock splits as positive news and reverse stock splits as bad signals, even though a stock split does not affect firm value. This means that the stock price does not follow the stock split arithmetic, and diverges from the pre-split firm value. Due to this signal perception, investors could achieve abnormal returns in the short term by buying shares shortly before or directly after a split announcement and shorting reverse split stocks. However, the alpha diminishes to nothing after the very short term (three to six months). The persistence of this anomaly has been traced by some academics to institutional investor bias, which makes it more likely to persist than if it were solely due to small investors.
Our next anomaly is the accruals anomaly. Investors favor companies with low accruals (the non-cash portion of earnings) because cash earnings are thought to be more valuable and persistent. This is why analysts estimate the “persistent portion of earnings” when they disaggregate GAAP earnings-per-share (EPS). Researchers have found that there is a negative relationship between high accruals and stock returns. The accruals anomaly is the practice of this relationship, where a portfolio would hold long positions in firms with very low accruals and short positions in firms with very high accruals. What is fascinating about this anomaly is that since its discovery over 50 years ago, both the magnitude and presence have not diminished, despite institutional arbitragers trading on the information. The reasons for this are twofold. First, trading on the accruals anomaly often involves investing in companies from which institutional investors are barred, due to their legal responsibilities to investors and certain unfavorable firm characteristics. However, the second impediment is the massive amount of analysis time the strategy incurs. A study by Lev and Nissim estimated that individual investors would have to periodically (quarterly) analyze the accruals of at least a few hundred firms to find the 50-60 high accruals firms necessary to profit from the accruals anomaly. So, while this would be costly both in terms of transaction costs and time, any investor could arbitrage this strategy.
The last anomaly is the Beta Anomaly, which has important investing lessons regardless of arbitrage strategy. Scott Murray, a researcher at Georgia State, analyzed stock returns from 1963 to 2012 and showed that high-risk stock portfolios generate about the same return as low-risk stock portfolios. This disagrees directly with the efficient market hypothesis because investors demand higher return for higher risk. If a stock has a high beta (high risk), we expect it to have a higher return, using CAPM theories. However, Murray et al. found that high risk stocks tend to be viewed by investors as “lottery tickets”, increasing the demand for them (and thus the price). If an investor used a naïve approach to these stocks, the timing of the purchase generally lowers their overall return to a low-risk return, despite the higher risk they are bearing. However, the researchers found that the anomaly was eliminated when they constructed their low and high risk portfolios with the same proportion of lottery stocks. So, in addition to the lessons on high risk portfolios and timing, the researchers also found that “lottery stocks” had very low institutional ownership, meaning that small investors were driving the demand (and thus the price). This is a behavioral bias that normal investors must actively work to overcome for the sake of their returns.
Bali, Turan G, et al. “A Lottery Demand-Based Explanation of the Beta Anomaly.” Journal of Financial and Quantitative Analysis, 13 Mar. 2014, poseidon01.ssrn.com/delivery.php?ID=471009102086027124067106005066094123061037004077091082097072004101007117124114108091001009026029106125054080121099112118084077024034037055050079127025088014076101008029054039095030072113021085021014098100092071067076126092030110088083022101092082096088&EXT=pdf.
Griffin, Carroll Howard. “Abnormal Returns and Stock Splits: The Decimalized vs. Fractional System of Stock Price Quotes.” International Journal of Business & Management, Dec. 2010, www.ccsenet.org/journal/index.php/ijbm/article/viewFile/6989/6286.
Jegadeesh, Narasimhan, and Sheridan Titman. “Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency.” Freshwater Biology, Wiley/Blackwell (10.1111), 30 Apr. 2012, onlinelibrary.wiley.com/doi/abs/10.1111/j.1540-6261.1993.tb04702.x.
Lev, Baruch, and Doron Nissim. “The Persistence of the Accruals Anomaly .” Contemporary Accounting Research, Columbia University, www.columbia.edu/~dn75/The Persistence of the Accruals Anomaly.pdf.
Lochstoer, Lars A, and Paul C Tetlock. What Drives Anomaly Returns? Columbia Business School, May 2016, www0.gsb.columbia.edu/faculty/ptetlock/papers/Lochstoer_Tetlock_May16_What_Drives_Anomaly_Returns.pdf.
“Momentum Effect in Stocks.” Quantpedia, quantpedia.com/Screener/Details/14.
Murray, Scott. “High Risk, High Return? Maybe Not: The Stock Market’s Beta Anomaly.” Georgia State University News Hub, 22 May 2018, news.gsu.edu/2018/05/23/stock-market-beta-anomaly/.