Quant Investing

Momentum in individual stock quant strategies

One of the challenges of using a quantitative approach to investing in individual stocks is choosing the actual strategies to use among of sea of seemingly ever expanding strategies. In my quant work I’ve narrowed down the universe of strategies to 11 but that is still an overwhelming number of strategies to select from. In practice, you usually have to use some fundamental approach based on historical data to make the strategy selection. But maybe there’s an easier way. Momentum. Does momentum work among different quant strategies? That’s what we’re going to find out in this post. If it does that sure would have some advantages. It would eliminate the need to statically choose among different quant strategies, potentially improve returns and drawdowns, and make individual quant stock investing a lot easier to implement. Let’s dive right in.

First, let’s take a look at the dispersion in annual performance among quant strategies. Using annual return data for the 11 quant strategies that run in my Quant Pulse service, you can see a list of them and their performance here, I ranked the strategies every year by return, assigned an ordinal rank to them, and color coded them by tertile (thirds). The ranking data goes back to 1999. See table below.

As you can see from the table the ranking of the strategies changes quite a bit from year to year. This means with whatever strategy you choose patience is really key to long term results. This is one of the key things that makes quant investing so hard to stick with. But maybe there’s a better approach by using momentum to rank and implement the strategies.

As it turns out momentum among quant strategies tends to be quite persistent so a longer, slower momentum approach tends to work best. I ranked the quant strategies, the top 10 stock versions only, by 3 year compound annual return, and then ran a strategy that invested in the top 1, top 2, or top 3 strategies based on the 3 year CAGR updated on a monthly basis. Performance is backtested to 2002.  For comparison purposes, I used the average performance of all the quant strategies, and the two benchmarks (SP500 and 60/40) over the 18 year period of the backtest. Results are presented below.

As you can clearly see from the results, momentum among quant strategies works really well. For either Top 1 10 stocks), Top 2 (20 stocks), or top 3 (30 stocks) strategies returns and worst year returns improve over the average of all the strategies, and also handily beat the benchmarks. The Top 2 or Top 3 strategies seem to be the sweet spot for returns and drawdowns. The other nice thing about this approach is that it is really easy to implement, just choose the top 1, 2, or 3 strategies ranked by 3 year CAGR on a monthly basis. Rinse and repeat each month. Turnover among the strategies and the number of portfolio holdings also is quite manageable. That is really about it. All the hard work is in designing the underlying strategies but after that just let momentum do the heavy lifting.

In summary, momentum works really well among individual quant strategies. Returns and risk are much improved over a buy and hold approach, and the need for strategy selection is eliminated. In order to implement this in my Quant Pulse service, I will be start providing the monthly rankings of the strategies starting at the next re-balance period. If you’re interested head over and take a look.

 

Tagged ,