Trading Off Correlations Between European and U.S. Markets

Includes: IEV, SPY
by: Digital Trading

Tracking the European market closes is a study that I started several months ago and one that I have revisited a few times since then. I have primarily taken a look at correlations between each of the major European indices and the S&P futures. I did previously write a blog on this analysis back in September, but to recap and update my findings, I will start with the correlations between the European and U.S. markets over a few different time frames (click to enlarge):

Note: The STOXX is an index of 50 blue chip European stocks and hence is a broad respresentation of the European market, as opposed to the FTSE (UK), DAX (Germany), and CAC (France). The above data confirms that there is definitely a large correlation between the U.S. market and the European bourses. One point that I will mention is that we see a noticeable breakdown in the correlation between the STOXX and the S&P since January 2010. This drop is mostly attributable to the performance of the DAX. I broke the 2010 data further by month:

The last time I ran this study at the end of September all of the correlations remained high, but those numbers of dropped off since and now we are hovering around the 0.80 level. Similar to the concept of pairs trading, I believe you can use these instances when the correlations break down as trading opportunities.

For example, back in March of this year the correlations between the two regions started coming down, and this move was confirmed in the following month in April when we essentially became uncorrelated. The last few days in April / beginning of May was the start of the sell off in the market so perhaps we could have interpreted the correlation breakdown as a leading indicator to the impending sell off.

Now shifting the focus to the past couple of months, we are beginning to see that the correlations are in fact backing off again. The October numbers may have been telling of the pull back in the market that we saw in early November. Having said that, looking at the month to date correlation for November, I would say that I’m more inclined to be on the short side.

I have revised my spreadsheet analysis to calculate probabilities of the S&P futures closing up or down a certain amount of basis points based on where Europe closes. This functionality will provide more value on an intraday basis. Below is a snapshot of the model:

Given that we are most correlated with the FTSE, I am using this index as the basis for comparison. The yellow cells are the input fields. There are two separate scenarios for up days and down days. The “FTSE % Chg” denotes where the FTSE closed on the day. If it ended green on the day, the left column (labeled “Up Days”) is used and vice versa for down days. The “S&P % Chg” is a threshold level that is arbitrarily chosen.

The input for this field is interpreted as follows: If the FTSE closes up x%, then what is the probability of the S&P futures closing up y%? The y variable in this case is what you would input for the “S&P % Chg” field. So for instance, you could choose to use 0.00% to calculate the odds of the S&P closing green on the day when the FTSE closes up x%. Perhaps more useful would be to input where the market is currently trading into this “S&P % Chg” field to view the odds of the market closing above or below its current level (the odds are given as an output at the bottom).

So in sum, I would track the European markets in order to view correlations, and particularly to look for breakdowns, between their markets and the S&P. Trading off these correlations is parallel to the concept of pairs trading, shorting one and going long the other when we see a divergence. Also, using this data to calculate probabilities of the S&P closing above or below a certain level based on where the European markets closed may be a useful tool to employ on an intraday basis.

Disclosure: No positions