I'm working on a simple tool for a brokerage(not trader) that needs an indicator for an equity's "volatility". I've been given the CDS(5 yr) and equity history for Ford for about 18 months, and from that I've been unable to correlate the variables because clearly a two-variable regression for something so complex would have correlated error terms(I haven't tried to fit this data to a marginal distribution, but I'm not convinced it's even elliptical, and so how the hell can... whatever: I'm not getting into copulas yet). I looked at the % changes of both the cds and stock price(using various sample sizes, shifts, etc), and I get correlatations that fluctuate like mad, like I've no idea how to even bs this one. Right now, since people want results immediately, I'm telling them to use a parity change in the 4 day sample mean of cds percent change as an indicator of volatility. They really want me to find how the cds changes lead the stock changes, but this seems an insanely difficult problem to solve! Ideas?

I'm thinking I should just use the time-series data to estimate coefficients for a LTI model and be done with it.

correlate (implied) volatility of Ford with absolute levels in CDS.btw, evidence so far is that equity markets lead CDS markets not the other way around

i appreciate the help! I was totally clueless, and I figured to assume CDS leads equity after reading a paper(some guys at LBS) on insider trading and CDS's. http://www.moodyskmv.com/conf05/pdf/pre ... charya.pdf

QuoteOriginally posted by: Lepperbecorrelate (implied) volatility of Ford with absolute levels in CDS.btw, evidence so far is that equity markets lead CDS markets not the other way aroundother evidence point to cds markets leading

HiAlso take a look at the change in MTM of the CDS, and the change in F Equity. Making a market in CDS is done also by looking at Equity prices. Finally, pls take into account LBO risks, which have the opposite directionality in credit vs equity. Using equity and other exogenous data, the guys at http://www.creditgrades.com/resources/resources did some work on this. But just because of LBO risks, along with things like dividend payments, etc, which are equity positive yet credit negative, means that it's difficult to implement any reliable regression.

How would one define implied volatility for a cds? Are we treating the CDS as a call?