Using economic indicators to improve quant performance

In today’s post I’m going to present how using a simple economic indicator can increase returns and lower drawdowns in quant portfolios. I’ve alluded to this in many of my quant posts in the past. Here I present the results of such an approach across a range of quant portfolios. Let’s jump right in.

For background on using economic indicators for investing see the following posts.

And for background on the quant portfolios I present below see quant portfolio section of the Portfolios page. There you’ll find a link to each of the quant portfolios.

Let’s combine these two strategies together. We’re going to use the SPY-UI indicator (based on the trend in the unemployment rate and the SPY moving average) as a risk-on risk-off indicator in each of the quant portfolios. The quant portfolios are normally formed once a year and then held for one year. We follow the same quant procedures except now we will check the status of the SPY-UI indicator once per month. If the indicator is risk-on we remain invested, if the indicator is risk-off we sell the quant portfolio and go to cash. The idea is to avoid being in an active aggressive strategy at the times of highest economic risk. Let’s see how this concept has performed since the beginning of 1999 through 2016. The chart below gives you an idea of how this works. The chart shows the number of positions in one of the quant strategies over time.

Basically, the quant system remained invested most of the last 18 years and only exited during the last 2 recessionary periods thereby avoiding two periods of poor stock market performance.

Now, we’ll look at how powerful this combination can be. The table below shows the annual performance and max daily drawdowns for the eight quant portfolios I most often discuss here on the blog. The first two rows show the results for the ‘normal’ quant portfolios that are formed and held for one year with no other changes in between. The last two rows shows the same strategies but with the SPY-UI indicator used as a risk-on risk-off overlay on the quant portfolios.

The use of a a simple economic indicator as a risk-on risk-off overlay in quant portfolios significantly improves performance, by almost 5% per year, and lowers drawdowns, on average by an absolute 21%, over the buy and hold quant portfolios. For me, the reduction in drawdowns is the most significant improvement. The drawdowns are maybe the hardest part of quant investing and makes sticking with the strategies quite difficult. Having a risk-on risk-off trigger certainly helps. And it’s pretty easy to build and implement the simple system. All you need is the unemployment rate and one moving average.

That’s about it for this post. In short, using a simple risk-on risk on overlays based on economic indicators can markedly improve performance and lower drawdowns in quant strategies.

P.S The effectiveness of the risk-on risk-off trigger can be improved further. This was one of my main motivations for diving deeper into using economic indicators to build investment systems. I’ve posted on an improved model using six indicators here. The culmination of that effort is the the COMP economic indicator that we use in the Economic Pulse Newsletter. Not only is it useful for building TAA models but it very effective as a risk-on risk-off overlay in quant portfolios. If you’re interested check out the newsletter about page. There’s a short screencast presentation there explaining what we do in the letter.

Full Disclaimer - Nothing on this site should ever be considered advice, research or the invitation to buy or sell securities. These are my personal opinions only.

6 thoughts on “Using economic indicators to improve quant performance

  1. Are the CAGR’s listed an average of all weekly rolling 1 year periods since 1999 (800+ time periods) or just 18 time periods?

    And are the drawdowns based on a daily rebalance? Or a single simulation?

    1. 18 time periods. I’ve also done the analysis with rolling 1 yr periods and the performance and drawdown deltas are similar.

      The drawdowns are based on a single simulation.

      Paul

  2. This is a great article. I am currently using the Enhanced Dividend Income for my 80 year old mother. I was forced to look at stocks for her for obvious reasons. What do you think of running 12 month moving averages on each of the stocks and combining that with the COMP indicator?

    1. Not a fan of using moving averages on the stocks in the quant portfolios. In my experience it doesn’t work very well.

      Paul

  3. Hi Paul, long time lurker and first time commenter. When the indicator takes you out and then back in to a quant strategy within a short time period (say within a couple of months), do you re-screen to decide which stocks to purchase or do you repurchase the same stocks that were held before the indicator went negative?

    Thanks

    Mike

    1. Hey Mike, great question. You can do either but it usually pays to re-screen. For most screens (except Pure momo and TV2) the turnover is quite low.

      Paul

Comments are closed.