20D means.. = so what your saying is you try to construct an outlier packet of securities that are a part of an index and arb the difference in volatility.. meaning buying straddles on a selected group of securities that are part of an index. Then shorting volatility in the same index with short straddles.. Maybe straddles aren't the appropriate structure for margin requirement .. but either way.. short vol in the index and long vol in a few of the securities that make up the index.. what structures would be best to exploit this? you gotta find the securities that are stretching the vol out the most in the index stripping them out and buying their vol against the sale of the index vol...
Delta. Two dimensions to the vol. IV higher individually than in the index (correlation risk), as you're not fully-replicating the index, and the premium to the down&out skew (20D vol / 50D vol >1). You can straddle ATM in dispersion, but you're not making an attempt to capture skew. Straddles are a poor choice for dispersion. The goal of long disp is to see outliers in street(component) vol that is diametric; abc +1%, XYZ -1% with little impact to the index. Plus you isolate a whole lot of vol-edge in 20-delta SPX puts.
God theres so much options lingo code to crack in the things you say... you could literally do it on any spread vol verse vol.. butterflys. debit spreads.. straight puts straight calls.. excerdera etcetera.. ?
the variables involved in a "what if " analysis of a trade like that are out of my reach as of right now... you would have to rip apart alot of data and model a few different assumptions to try to figure out which stocks to pick .. how many to buy vol on and how many index vols to sell against them
"Thanks for reading all this. I honestly donât know how you managed, I fell asleep while proof-reading half-way through."