wilmott.com

Recently I was looking at the P&L performance of a large pairs book. At first glance, the book's P&L was "mostly L" during periods of increased volatility. Further inspection, however, seemed to show that the culprit was really increased index correlation (SPX correlation -- most of the book's stocks seemed to be SPX names) rather than increased volatility. And, to me, this seems to make sense.If one wanted to hedge against P&L losses corresponding to volatility spikes, one could obviously employ a vol swap or maintain an appropriate SPX option position. But if one wanted to 'hedge' against a spike in correlation, must one buy the index options and sell the equity options of the component stocks? Any other suggestions are appreciated. Thanks to all in advance.

An easier, although less than perfect, hedge I would try would be to find an asset that is significantly correlated with the correlation of the index. For example, you might find that gold or the Swiss Franc tends to increase with the correlation of the index. There is the risk that that correlation would also break down at that time, but I would test prospective candidates with a simulation.Another approach you might try, that I am less familiar with, would be to find some principle components in the returns of your portfolio, and use a factor model. If you believe in APT type models, the correlation maybe possible to isolate and cancel with another asset, likely similar to the ones you'd try above.Of course, the hard part about this problem is that your correlations are not a number, but in the case of SPX, a 500x500 matrix...

i would think that putting on a dispersion trade to hedge your correlation risk would cost too much in slippage. you are probably much better off with the simpler solution of grabbing some downside puts in spx. basket vol moves same way as correlation, especially to the downside.

Those were my thoughts, too... I suggested that since the correlation trade would not be "perfect" to the book, owning some SPX premo (I suggested staying in the front two months) was the way to go -- Such a hedge is just as "imperfect" as the dispersion trade, considerably more efficient and keeps one out of "disaster d'jour" land...Thanks for your prescription, apine.Anybody else?