Rolf Agather: I think it's important to recognize hat even as we are able to produce a particular performance or risk-return profile based on some exposure, there are implementation issues that can come along. Equal weight is a great example. It's been around a long time and is a relatively naïve strategy that has performed better than some other strategies. But when you equal-weight companies, you are putting the same amount of money into the largest company in the market as the smallest company, so potentially you can introduce liquidity or capacity issues.
In terms of persistence, we do think that markets are generally efficient and therefore, true anomalies will go away over time. But when you talk about risk premia, like value, momentum and quality factors, there is the notion that you are bearing additional risk for which you have a reasonable expectation to benefit. We would expect a risk premia to persist.
But it's important to recognize relative valuation levels. As money flows into certain strategies, they can become overvalued. That doesn't make the potential benefit go away, but it can create an artificial imbalance in a particular factor or strategy.
Kal Ghayur: There are two schools of thought here. One echoes what Rolf is saying, the risk premium explanation, that suggests that these things can persist over time because they are a compensation for bearing extra risk. But the risk premium explanation is a little bit hard to justify when it comes to the low-volatility anomaly. There's a lot of research that shows that low-volatility stocks tend to outperform high-volatility stocks. This is contrary to everything we know in finance, where the most basic argument is that you can only earn higher returns if you take higher risks.
There is another school of thought that argues that things that have worked in the past and continue to persist over time do so because of behavioral or institutional biases. According to this explanation, investors have some behavioral biases when they invest in stocks. They like glamour stocks and stocks with nice stories, so they tend to invest in growth stocks as opposed to the more depressed value stocks. As a result, the prices of these glamour, high-growth stocks tend to bid up and their future returns fall, whereas the prices of the less exciting value stocks are bid down and their future returns tend to be higher.
Low volatility is still a puzzling anomaly here, but it can potentially be explained through an institutional bias. Most managers are measured against a cap-weighted benchmark — passive or active managers. Low-volatility stocks are low beta stocks, so there is little incentive for managers to hold low-volatility stocks. In fact, because the beta of low-volatility stocks is less than one and the market beta is one, it introduces tracking error. So managers shy away from these types of stocks. As a result, the future returns of high beta stocks are low and the future returns of low beta stocks are high, explaining why the low-volatility anomaly persists over time. This would probably only disappear if the focus on benchmarks disappears.
Dan Draper: I think smart beta strategies adapt over time. Some factors — large cap, small cap, value, growth — which were identified back in the 1970s are now construed as to some degree traditional beta, even though they are not the true overall market.
I expect that smart beta strategies that perform well and stand the test of time to become more mainstream. If that happens, we think we're doing our job!