Factor investing is a long-term pursuit. Successfully following a factor-based approach is a function of studying what works over time, implementing it in a disciplined way, and then staying the course through what will inevitably be long periods of underperformance to achieve your long-term goals.
Because factors can move in and out of favor for extended periods of time, implementing timing strategies to move in and out of factors based on valuation is either very difficult or impossible, depending on who you ask.
In our studies of factors, we have seen benefits from factor timing in the past when two conditions were present.
- The factor was at extreme levels relative to the past
- You have the patience to wait for the factor to turn around, understanding that the over or undervaluation will persist much longer than you think it will.
A month or so ago, I wrote about one such situation on the positive side with the Price/Book ratio. Our data indicates that the Price/Book is extremely cheap relative to history, and for patient investors, that may present a long-term opportunity. For this article, I thought it would be interesting to look at the other side of the coin and examine the most expensive factor according to our data.
But first, some long-term background.
The initial capital asset pricing model was based on a fairly simple assumption. The premise was that in order to get a better return, you need to take more risk. That principle makes sense logically, since there typically are no free lunches in investing, or in life in general.
Even as the Fama French factor model has expanded over the years, the principle has held. For instance, the outperformance of something like value can be explained by the model based on the fact that value stocks tend to be more risky.
One factor stands out, however, in its deviation from that long-term risk and reward trade-off. That factor is low volatility. Investing theory would tell you that low volatility stocks should underperform the market as a whole since they are less risky. The problem, however, is that they don’t. Many historical studies have shown that low volatility stocks tend to outperform their high volatility counterparts over time. So maybe when it comes to these types of stocks investors can have their cake and eat it too.
But why would this occur? Since traditional finance models equate return with risk, they aren’t going to help in explaining it, and we need to go outside that. Some people believe low volatility stocks work because they perform better in down markets, and due to the nature of compounding, bad years tend to hurt investors more than good years help. Other firms like AQR have come out with much more detailed explanations. The bottom line, though, is that what is going on cannot be completely explained by traditional models, so either there is a risk factor out there associated with this that has not yet been discovered or this is a function of the market mispricing these assets.
According to Research Affiliates, the low volatility factor outperforms the market by a little more than 1% per year long-term. So it isn’t among the top factors from a performance standpoint, but it has added value over time. The recent performance has been even better. For example, the iShares Edge MSCI Min Vol USA ETF, the largest ETF in the space, has outperformed 87% of the funds in its category over the past 3 years. Now three years is not a long period of time, but the recent performance is strong nonetheless.
That recent performance has led to a situation where valuation has become an issue in the space, even when compared to the lofty valuations of the market as a whole. For example, the median PE ratio of the holdings of the minimum volatility ETF is now 28.2 vs. 25.7 for the S&P 500.
If you couple the valuation with earnings growth, the overvaluation of the factor really becomes evident. When we calculate historical earnings growth for companies, we like to use an average of the 3, 4 and 5 year rates. We do this to smooth out issues related to individual growth rates. Using this metric, the median annual EPS growth for firms within the low volatility universe is 7.9%. If you look at future analyst estimates instead of historical growth rates, the situation doesn’t get any better. The median projected 5 year EPS growth rate for these firms is 8.9%.
So the low volatility space is full of companies trading at lofty valuations whose past and projected future EPS growth rates are low. That doesn’t seem like a recipe for outperformance going forward.
Looking at valuations using the great Research Affiliates Smart Beta Interactive Tool also leads to a similar conclusion. The chart below looks at the valuation of low volatility relative to the other factors they track. The vertical bars represent the historical ranges of each factor and the white dot is the current valuation. As the chart shows, low volatility is currently one of two factors at the absolute top of their historical range (along with Dividend Growth).
None of this, of course, means that an imminent decline is coming for low volatility stocks. Factors can stay in and out of favor for long periods of time, and when you bet on a return to historical averages you will almost inevitably be early. But mean reversion is a powerful force in factor investing, and eventually what goes up must come down. So with low volatility stocks trading at premiums to the market based on valuation, and not exhibiting the earnings growth to back up those valuations, it may be a good time for factor investors to look elsewhere.