In my project I tested a Pairs trading strategy in the US equity market for data of 10 years. These 10 years include the financial crisis (2007-2008). The strategy follows a Cointegration approach.

Here is what I observed: In the beginning of the financial crisis lots of the trading pairs that were open closed, many of these trades still closed at a profit (but percentage wise (profitable trades/all trades that closed) less than in other periods. In general during the financial crisis I closed (mostly in the beginning of the crisis) and opened many more trades, compared to periods of lower volatility in stock prices. I am now trying to explain how a market shock affects my trading strategy.

My reasoning so far: When there is higher volatility in stock prices, the spread values of the pairs also deviate more from their mean and thus are more likely to trigger the opening or closing thresholds of the strategy. This ultimately leads me to the question of how or why a move in the overall market disrupts the statistical long-term relationship of two stock prices that the cointegration test found?

Thanks a lot in advance for any help!

  • 1
    $\begingroup$ Hi: I think you answered your own question in the last line. cointegration is a statistical relationship that has existed during the lookback period. ( assuming the pair passed the test ). When the market goes haywire as in 2007-2008, this is not a statistical movement. It's a market movement that statistics-cointegration cant account for because the regime has changed. This is one of the reasons why co-integration is quite the beast to deal with. $\endgroup$ – mark leeds May 1 at 9:26

Your Answer

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

Browse other questions tagged or ask your own question.