A number of years ago, I researched low-beta stocks and I was convinced that low-beta stocks increased portfolio risk-adjusted returns. A number of different researchers and portfolio managers also concluded that low-beta stocks boosted portfolio performance. See here and here for examples. There were several theories about why low-beta stocks outperformed higher-beta stocks. One theory was that low-beta stocks tend to increase tracking error relative to equity benchmarks and portfolio managers are willing to sacrifice alpha to keep racking error low. Managers’ desire to minimize tracking error is a well-known issue. There is no inherent performance benefit in low tracking error, but many retail investors have been trained to seek out funds with this property. Another theory about why low-beta stocks outperform is investors’ and managers’ reluctance to employ leverage. Low-beta stocks tend also to be low risk. A portfolio manager or investor could use leverage to bring a low-beta portfolio to a risk level comparable to a benchmark, but there is evidence that individuals and managers tend to avoid levered portfolios. Part of this is probably complexity (margin costs, potential for margin calls).
In the last year or so, I have started to wonder whether the growth in index investing has reduced the benefits of low-beta stocks. Imagine that you have two stocks that are part of an equity index. One has beta of 1.3 and one has beta of 0.7 relative to the index (let’s use the S&P 500 as the hypothetical index). Both stocks have a 1% weight in the index. The index goes up 2%. The 1.3-beta stock goes up 2.6% and the 0.7-beta stock goes up 1.4% (just assuming the stocks respond in line with their betas). The market cap of the company with the 1.3 beta has risen considerably more than that of the company with the 0.7 beta. The S&P 500 is a market cap weighted index. The relative weight of the 1.3-beta stock is now higher than that of the 0.7-beta stock. As new money flows into the index, more of it goes to the higher-beta stock than the lower-beta stock. Over time, in a rising market, the higher-beta stocks get more allocation and more of the new money flows and vice versa. This effect suppresses the returns of lower-beta stocks.
Over time, more indexed assets imply lower returns for low-beta stocks. This mechanism seems straightforward. I have also been wondering whether this effect is part of why value stocks have been lagging growth stocks for so long. Historically, value stocks (those with low P/B and P/E) have tended to deliver higher returns than growth stocks, but this has not been the case for a long time now. Low-beta stocks also tend to be value stocks, so a reduced low-beta effect would tend to reduce the returns of value stocks.
One of the challenges for investors and portfolio managers is understanding when and how markets change. There is, ultimately, no certainty that relationships from the past will persist. The growth in indexing is certainly one of the most significant changes in how people invest over the past fifty years. The potential for indexing to hamper the total returns from low-beta stocks, and potentially from value stocks, is significant.
Geoff Considine on Finance and Economics 