I think the real edge is in vol or var swap dispersion... if you got enough capital to play, of course.
This. is. not. an. arb. As a rule of thumb, the formula: index_vol = sqrt(correlation) * weighted_avergage_component_vol isn't a bad approximiation. So index vol at 83% of component vol implies a correlation of 0.69 (.83 squared). Your "arb" is just a bet that realized correlation will be lower than 0.69. Acutal formula is: sqrt( summation(i)[v(i)^2*w(i)^2] +2*sumsum(i,j>i)[v(i)*v(j)*w(i)*w(j)*p(i,j)]) where v is vol, w is weight, and p is rho or correlation. Or in matrix form: sqrt(W'QW) where Q is the covar matrix and W is a vector of weights. Did you read the paper posted earlier in the thread?
Hi Guys, Thank you for all your help the last days. I have read the paper several times and after working a bit with it finally understand dispersion theory. For some reason I was really slow to grasp this. Again, I do really appreciate your help! Kind regards, Steffan
Hi, I am trying to understand the concept of dispersion. Quote from Atticus: "You're best to tinker with OTM calls rather than eat the index skew on short dispersion (in puts)." Can Atticus or somebody clarify/explain the above quote. Does that mean that it is better to be short index call instead of long index put on short dispersion? Thanks in advance