Low vol strategies. Jon Danielsson. Modelsandrisk.org

Low vol strategies

May 3, 2020
Does it make sense to invest in low vol funds?

The promise of a low volatility — low vol — fund is that volatility is mispriced. Buy a portfolio of low volatility stocks and leverage it up, and because of the mispricing, the return on the low vol portfolio will be higher than other portfolios with the same volatility.

There are, of course, many ways to ascertain whether low vol makes sense, and one can download an ocean of studies on this very question.

Here, I am opting for one of the simplest ways. I downloaded all of the monthly stock returns from the CRSP, spanning 1926 to 2019. Furthermore, only use the 500 largest stocks (or lower before 1953). In particular, the same number of stocks as CRSP uses in its calculation of the SP-500. I identify stocks with PERMNO and use RET for returns, that is, total returns that include dividends.

I’m basing the comparison on the same stock in two subsequent years, where I restrict the analysis to stocks that have 12 observations in the first year.

We start with one dollar in 1926 and create six portfolios.

  1. All of the stocks;

  2. Only invested in US treasuries;

  3. Then four based on volatility buckets. Stocks that have annual volatility below 25%, then 25%-50%, 50%-75% and above 75%.

The portfolio is rebalanced at the end of each year, and performance is ascertained by the total return for the next year.

No volatility adjustment

The most interesting thing is that the highest vol portfolio has the worst performance.

I would have expected based on standard mean-variance analysis that highest vol portfolio should perform best in the long run.

Alas, data doesn’t lie.

But what about the low vol strategy. Take the data used to create the previous plot, and create portfolios with the same volatility.

In other words, take the volatility of the below 25% volatility portfolio, and rebalance the other four to have the same volatility, by allocating some of the money into treasuries.

The below 25% will be all equities, and the 75% to 100% will be partially high volatility stocks and partially treasuries.

And now we see the confirmation of the wisdom of the low vol strategy. The below 25% volatility portfolio has the best performance.

Constant volatility

Again, the most interesting thing the performance of the low vol portfolios mostly come from eliminating the highest vol stocks!

What I haven’t done here is ascertain whether the below 25% has the best performance if I want more risk, because, after all, I would have to borrow to leverage the portfolio up, and that costs much more than the three-month treasury I use for comparison. Cumulatively, that can be quite a large number, most disadvantaging the lowest vol portfolios.

Footnote

I did the analysis both in R and Julia just to get a comparison, and see which language I prefer for this.


The five principles of correct riskometer use
Of Julia and R

Models and risk
Bloggs and appendices on artificial intelligence, financial crises, systemic risk, financial risk, models, regulations, financial policy, cryptocurrencies and related topics
© All rights reserved, Jon Danielsson,