In this post I describe a simple low volatility quant strategy that outperforms the market and another strategy that combines low volatility with value and momentum to provide further outperformance.

Stocks that exhibit low volatility outperform more volatile stocks, contradicting the efficient market theory and the capital asset pricing model (CAPM). For a great explanation of the low volatility anomaly and some possible explanations for it, see this post by the Alpha Architect team.

For my purposes here I want to keep things very simple. In my version of a low volatility quant strategy, I start with a universe of large cap, high liquidity stocks, basically the O’Shaughnessy Market Leaders universe and then invest in the top stocks with the lowest volatility, as measured by 1 year standard deviation. And lastly as an added risk protection measure, I use the SPY-UI indicator as a crash protection measure. That’s it. Simple. Performance metrics for a 10 stock version and a 25 stock version of this strategy are listed below. I also included SPY and the USMV (minimum volatility) ETF as the benchmarks. All numbers are from Portfolio123.

As the table shows, low volatility on it’s own, the USMV ETF, outperforms the benchmark on both return and risk-adjusted return (sharpe ratio), as we would expect for the historical research. When you concentrate the approach, equal weight the stocks, and add an extra risk management measure results improve dramatically. Results are similar going back to 1999 as well. The power in these strategies is the risk-adjusted returns, market beating returns with maximum daily drawdowns of just over 10% are impressive to say the least. A more detailed picture of the results of the top 25 strategy is below.

Not too bad. These strategies have the added benefit of being low turnover. Compared to other higher risk quant strategies this one does not seek to knock it out the park in terms of returns but to provide market beating returns with lower risk and low turnover. For more conservative quant investors or those just getting started with quant investing this strategy is not a bad one to start with.

Another approach is to combine low volatility with other factors like value and momentum. That seems to work pretty well also. Check out this post on the ‘Conservative Formula’ that combines low vol, value, and momentum into one strategy. I’ll probably be taking a look at that in the near future.

For now, I plan on adding the low vol strategy to my Quant Pulse Service (see description) in the next few days. If you’re interested subscribing see here.


2 Comments

Brad · March 20, 2019 at 6:21 am

Hi Paul. Love this work. How much of the outperformance here comes from adding in the downside protection of SPY-UI vs. outperformance strictly from the low. vol selection? It would be interesting to see a true “apples to apples” comparison between your Top 10 and Top 25 screens with SPY and USMV that also use the same downside protection scheme.

On an unrelated topic, I was wondering if you have ever seen a screen or strategy that compared SPY with modified SPY that removes any stocks that have negative 12 month momentum. So in any given year instead of holding all 500 stocks long you may only be holding 100 or 200 of the stocks because we have filtered out anything with negative momentum. I suspect this would outperform the SPY index in most years and I would think it would be an easy thing to create with an ETF but I have not seen it.

Thanks!

    paul.novell@gmail.com · March 20, 2019 at 1:35 pm

    Good point Brad. The numbers for the strategy without SPY-UI are 9.79%, -31.44%, 0.89, respectively.

    And no, I’ve never seen that type of strategy. Easy to create as an ETF but not really investable as an individual.

    Paul

Comments are closed.