For evidenced-based investors, it’s easy enough to measure the best-returning investment styles.
In their January 2014 study “Risk and Return Within the Stock Market: What Works Best?” authors Roger G. Ibbotson and Daniel Y.-J. Kim examined the return-predictive characteristics of some well-known investment styles: beta, volatility, size, value, liquidity and momentum.
Their study covered the 42-year period from 1971 through 2012, and included the largest 3,000 stocks in the U.S. CRSP database. Following is a summary of their findings, which basically serve to confirm prior research:
Defying economic theory that risk and return should be positively related when ranking stocks either by beta or volatility, the authors found that the highest quartile of stocks had the lowest returns.
However, the relationship isn’t linear. The annual average return from the lowest quartile to the highest quartile beta stocks were 16.0 percent, 16.5 percent, 15.4 percent and 13.0 percent, respectively. Similarly, when ranking stocks by daily volatility, the returns from the lowest quartile to the highest were 15.3 percent, 16.4 percent, 16.5 percent and 12.8 percent, respectively.
The common explanation is that leverage-aversion on the part of some market participants resulted in the popularity of high-volatility, high-beta securities. Their popularity pushes up prices and reduces returns. The bottom line is that it’s not that low volatility is good, it’s that high volatility is bad.
Small-cap stocks outperform large-cap stocks. When ranking stocks by market capitalization from the smallest to the largest quartile, the annual average returns were 17.4 percent, 15.2 percent, 15.0 percent and 13.3 percent, respectively. Here, risk and return were positively related as the smallest stocks had the highest standard deviation and the relationship was monotonic.
Value stocks outperform. It doesn’t matter whether the valuation metric was price-to-book (P/B) or price-to earnings (P/E), returns decreased monotonically as the sample moved from value to growth quartiles. For example, ranking by P/E, the annual average returns by quartile were 19.1 percent, 15.9 percent, 13.2 percent and 12.7 percent, respectively.
However, again, defying economic theory, the stocks with the lowest P/E ratios not only had higher returns, but lower volatility (a standard deviation of 23.2) than the stocks with the highest P/E ratios (a standard deviation of 28.5).
Illiquid stocks outperform liquid ones, and the relationship was monotonic across quartiles. Ranking stocks by turnover, from lowest to highest, the annual average returns by quartile were 17.3 percent, 16.3 percent, 15.2 percent and 12.1 percent, respectively.
However, yet again defying economic theory, the least liquid stocks produced the highest returns and had lower volatility (a standard deviation of 20.2) than the most liquid stocks (a standard deviation of 28.4). And the relationship was monotonic.
High-momentum stocks outperform low-momentum stocks, although the relationship isn’t monotonic. By quartile, from high momentum to low momentum, the average annual returns were 16.1 percent, 16.5 percent, 15.8 percent and 12.5 percent, respectively. Also of interest is that the high-momentum quartile not only produced higher returns, but experienced lower volatility (a standard deviation of 23.5 versus 28.7).
However, the relationship here again was not monotonic. The standard deviations of the middle two quartiles were 19.7 and 20.7, respectively. Importantly, the low-momentum quartile had both the worst returns and the highest volatility.
None of these results is surprising to those familiar with the literature. What is surprising is that risk (at least as defined by volatility and standard deviation) and reward have not always been positively correlated as financial theory predicts.
Contrary to the popular wisdom that greater reward comes with greater risk, low-beta and low-volatility portfolios outperform high-beta and high-volatility portfolios. In addition, as measured by turnover, less-liquid stocks both outperform and are less risky. And high-momentum portfolios outperform while also being less risky.
The authors concluded: “Overall the best-returning characteristics are high earnings/price, high book-to-market, and low turnover. On risk adjusted basis, the best performances were low beta, low volatility, and low turnover.”
They also found that: “Whether it be through factors that encode popularity among investors (turnover, growth), academic popularity (citations), or popularity caused by leverage aversion (beta, volatility), popularity underperforms.”
I’d add that we also see this phenomenon in investor cash flows. The most popular asset classes, sectors or individual stocks earn the lowest future returns as cash inflows from performance chasers.
Performance chasers tend to make expensive assets even more expensive in bull markets and cheap assets even cheaper in bear markets, which can lead to buying at the top and selling at the bottom.
Larry Swedroe is the director of research for the BAM Alliance, a community of more than 140 independent registered investment advisors throughout the country.