Is there a quantitative method in monitoring trades to reduce the possibility of correlated trades?


The most straightforward approach is to develop a covariance matrix to ensure that you are not overweighting to the same factors or bets in your trading. The covariance matrix can be built off of a factor model, for example, or you can construct a covariance matrix based on your prediction signals if you have multiple models. In this way you can understand the ex-ante historical correlation of your trades.

Note that there is a considerable amount of art in designing a covariance matrix (See Fabozzi).

Without understanding more about your approach it's hard to be more helpful than the above approach.

  • $\begingroup$ Can you be more specific about Fabozzi? $\endgroup$ – Zarbouzou May 9 '11 at 14:56
  • 2
    $\begingroup$ Frank J. Fabozzi has a couple of books that cover risk modeling as well as covariance matrix estimation. A good one is "Robust Portfolio Optimization" $\endgroup$ – Ram Ahluwalia May 9 '11 at 22:08
  • $\begingroup$ I currently trade a basket of currencies which have correlation between them to a certain extent. As I know that correlation is ever changing, I can only minimize it through gut feeling. This approach gives me unease as gut feeling is not something that should be counted upon in the long run. To test whether minimizing correlation, I have, in the past few days, come up with a idea of systematically monitor trades. I always take the first signal of trading system. If another signal is generated in X days after, I will calculate the correlation between them for the past Y days. Will need test. $\endgroup$ – user1234440 May 11 '11 at 4:36
  • $\begingroup$ If the correlation between them exceed a certain threshold, i will skip it. This is only my initial answer to my problem of reducing correlated trades. As a trend follower, I need to be exposed to as much market as possible so as to take advantage whenever any one of them explode. Any advise would be great! Thanks! $\endgroup$ – user1234440 May 11 '11 at 4:38
  • 1
    $\begingroup$ Developing an exponentially weighted covariance matrix that weights more recent observations is probably the way to go $\endgroup$ – Ram Ahluwalia May 23 '11 at 15:13

Historical correlation isn't as useful as you might think. Like volatility, correlation is not constant. During times of stress, it is common for the correlation of many different assets to increase/change.

As examples, look at the data for August 2007, the last quarter of 2008, and May 6, 2010. Any trading scheme that minimized correlation before those periods probably had a much different affect during those periods.

Edit 1 (05/10/2011) ===========================

I've bumped into all sorts of problems with this issue, with the estimation itself involving large errors. If you dig around, you'll find several papers with important improvements.








  • $\begingroup$ At a previous trading group, we regenerated the risk model every month. The backtester would use only the risk model available for that month of simulation. $\endgroup$ – chrisaycock May 10 '11 at 0:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.