Monetising default correlation
Reuters reports that default correlation on the investment-grade Markit iTraxx Europe index reached new highs of 45 percent on Tuesday. Suppose you believe that this was too high: what could you do?
The effect of increasing default correlation is to move value up in a CDO. So if you believe the market spreads of the assets are right - or if you have no view on them - you can still take a view on default correlation by trading different parts of a CDO. For instance in the iTraxx if you believe that the implied default correlation is too high, then that implies that the CDS spread available for writing protection on the supersenior is too high and the CDS spread on the junior is too low. So you buy protection on the junior, sell protection on the supersenior, and make money if the market comes to believe you are right. Alternatively you could buy protection on junior and hedge by selling protection on the individual names (although that is quite a lot of trading for the iTraxx).
With that preliminary you can see why implied default correlation - the level the market demands for trading the index - is so high. Market participants are long risk on the supersenior and are trying to get out. Hence the cost of protection on the supersenior has ballooned out and thus implied default correlation is high. This does not mean that the market necessarily believes that companies are more correlated (or riskier): it just means that there are more buyers than sellers of protection on the senior tranches right now.
Update. FT alphaville has more background and a pretty graph of fair value CDS spread by tranche as a function of correlation here.
The effect of increasing default correlation is to move value up in a CDO. So if you believe the market spreads of the assets are right - or if you have no view on them - you can still take a view on default correlation by trading different parts of a CDO. For instance in the iTraxx if you believe that the implied default correlation is too high, then that implies that the CDS spread available for writing protection on the supersenior is too high and the CDS spread on the junior is too low. So you buy protection on the junior, sell protection on the supersenior, and make money if the market comes to believe you are right. Alternatively you could buy protection on junior and hedge by selling protection on the individual names (although that is quite a lot of trading for the iTraxx).
With that preliminary you can see why implied default correlation - the level the market demands for trading the index - is so high. Market participants are long risk on the supersenior and are trying to get out. Hence the cost of protection on the supersenior has ballooned out and thus implied default correlation is high. This does not mean that the market necessarily believes that companies are more correlated (or riskier): it just means that there are more buyers than sellers of protection on the senior tranches right now.
Update. FT alphaville has more background and a pretty graph of fair value CDS spread by tranche as a function of correlation here.
Labels: CDS, Markets, Securitisation
posted by David Murphy at 07:49
0 Comments:
Post a Comment
<< Home