Note: This is a modified version of an analysis I did for subscribers of the Economic Pulse Newsletter where I used the SPY-COMP indicator from the newsletter in world wide markets. Results for the SPY-COMP indicator are significantly better than the SPY-UI indicator presented in this post.

One of the basic premises of several modern TAA strategies, is that the US stock market functions as a good risk indicator for worldwide markets. Most investors realize that in a global economy world stock markets are highly correlated and move together a lot of the time, especially during times of crisis. There is an old and much repeated saying that ‘when the US markets sneezes the rest of the world catches a cold’. The chart below shows rolling 1-year correlations among global stock markets. While recent correlations have gone down recently, they shoot up during time of market and financial stress. The bigger stress the more correlated they become.

But are these correlations strong enough to build an investment system based on them? Yes. For example, look at the success and performance of Antonacci’s GEM strategy which uses the performance of the US market as a risk-on risk-off indicator. In GEM the 12-month performance of the US market is used as absolute momentum filter for a world-wide investment strategy. Turns out that you can also use the US economy in the same regard. On the surface, it might seem counter-intuitive to use the performance of the US economy and US market to make worldwide investment decisions. Let’s dive into some historical data to see how well this approach holds up. I’m going to use the SPY-UI indicator I’ve presented before as a risk-on risk-off indicator in foreign markets.

I performed historical back tests on a few countries and regions using SPY-UI vs using the region/country’s own market and UI as the risk-on risk-off indicator. I was able to get data for Germany, Japan, and the Eurozone. I also did the test in local currency (Euros in this case) where I could. The investment during the risk-off period in the cases I looked at was cash at 0% return. I did not want to cloud the results with various bond choices for the risk-off investment. The results are shown in the table below.

Let’s start with Germany, represented here by the DAX index. I was able to go back to Dec 1987 for the DAX priced in Euros. The table above shows the buy-and-hold performance of the DAX index compared to DAX-UI, using the UI economic indicator in combination with the 6-month SMA of the DAX, and SPY-UI DAX, using SPY-UI to enter and exit the DAX index. The performance of using SPY-UI on the DAX index, 11.26%, is the best option by a considerable margin. Using UI with the DAX 6 month-SMA returned 10.34% and buy-and-hold returned 8.62%. The same holds true from the perspective of a dollar-based investor in DAX, using the EWG ETF, 10.14% return vs 6.86% for buy-and-hold.

Next, consider the European wide index, the STOXX 50 index. I performed similar comparisons, using UI with the STOXX 50 6 month-SMA or using SPY-UI to invest in the STOXX 50 or not, compared to buy-and-hold. Similar to the German case, using SPY-UI was the best option by a considerable margin, 8.17% CAGR vs. 7.3% for UI and the STOXX 6 month-SMA, and 4.78% for buy-and-hold.

Next, we have Japan and Developed Markets in general. I was only able to use the dollar-based indexes for the analysis. Using SPY-UI on EWJ or VEA provides much higher returns than buy-and-hold on EWJ or VEA. I also looked at using the respective 6 month-SMAs of either index and the results were lower than using SPY-UI, tracking what I found with the other markets.

In summary, the SPY-UI indicator is very effective as a risk-on risk-off indicator for US and for global stock markets, even more so than using a region or country’s own market index as the basis for a risk-on risk-off indicator.


2 Comments

Vincent · July 9, 2018 at 3:32 pm

– “For example, look at the success and performance of Antonacci’s GEM strategy which uses the performance of the US market as a risk-on risk-off indicator. In GEM the 12-month performance of the US market is used as absolute momentum filter for a world-wide investment strategy.”

I think it’s wrong because if VEU is positive over the past 12-month and SPY is not GEM is invested in VEU.

– Did you backtest a strategy like GEM but including a SPY-UI for both the US and Developed ex-US.
I am interested in the results

And thanks for your blog !

    paul.novell@gmail.com · July 9, 2018 at 11:31 pm

    I am referring to the absolute momentum component of the strategy which is purely determined by SPY. See Antonacci FAQ here. He says “I mention there that I determine absolute momentum using only the S&P 500 index, since the U.S. leads world equity markets. I cite a supporting reference.”.

    And yes, I have tested a GEM like strategy using SPY-UI. It’s called DM-COMP. I use it in my Econ Pulse Newsletter.

    Paul

Comments are closed.