Pim van Vliet and Jan de Koning, both members of Robeco’s quantitative equities team (with van Vliet responsible primarily for the firm’s conservative equity strategies), have written a book challenging the claim that risk and return are positively correlated. High Returns from Low Risk: A Remarkable Stock Market Paradox (Wiley, 2017) is intended for a broad audience of investors. As a result, even though the authors obviously have quant skills, there’s no razzle-dazzle math on display here.
The book’s results are based on a dataset of monthly closing prices from January 1926 to December 2014 of the U.S. traded stocks of the largest 1,000 companies by market capitalization at any given moment in time. For each of these 1,000 stocks he (I assume van Vliet) measured the rolling three-year historical monthly return volatility and ranked them by risk (throughout risk is equated with volatility). Then he constructed two portfolios, one containing the 100 stocks with the lowest volatility, the other containing the 100 riskiest stocks (the high-volatility portfolio). He rebalanced the portfolios every quarter. Assuming that a person put $100 into each portfolio on New Year’s Day 1929 and reinvested any capital gains for 86 years until New Year’s Day 2015, the low-volatility portfolio was worth $395,000 at the beginning of 2015, the high-volatility, $21,000. Put another way, the low-volatility portfolio returned 10.2% annually on average whereas the high-volatility portfolio returned only 6.4%.
The disparity might be inflated somewhat by virtue of the fact that the calculations start in 1929 and “the low-volatility portfolio wins by losing less during times of stress.” The high-volatility portfolio would have been worth a little over $5 when the market bottomed out in the spring of 1932; the low-volatility portfolio would have been worth $30. Nonetheless, the authors contend, “if we were to start both portfolios in the spring of 1932, the low-volatility portfolio would still ‘win’ by a very significant margin.”
A low-volatility portfolio, it should be noted, doesn’t produce maximum returns. Given ten portfolios, each containing 100 stocks and ranked according to volatility (low to high), and using the same 86-year time frame, the portfolio in the fourth decile performed best (about 12% a year). Even the portfolio in the ninth decile performed better than the low-volatility portfolio. But the portfolio of the 100 stocks with the highest volatility performed far worse than any of the others.
The authors analyze why low-volatility stocks are overlooked in the market, thereby providing an opportunity for solid returns. Basically, “virtually everybody seems to be drawn to the dark and risky side of the stock market.” So, even though the paradox was first discovered over 40 years ago and even though it may become more well known, “there is every reason to believe the paradox will continue to exist and may even become stronger.”