Quote from atticus:
Right I run the marked price against a theoretical flat-vol fly of the same strikes. I use the PF to give me a dollar figure on index skew.
theoretical flat-vol...Numerically.. flat vol...skew=0 smile=0
. i'm going to read this thread and see if i can make sense of how this three put one call pitch fork translates into a good representation of dollar skew.. my first thought is.. that if flat vol is 0 dollars.. using the straddle.. or some combination of short vol at the money.. IE pitchfork would give you a sort of index of skew to continually work from in dollars.. so you could continually graph, compare, etc skew values in dollars over time..
i've been theorizing about how to map some option volatility metric to actual realizing volalitity in one graph.. meaning.. if i take some bollinger bandwidth delta. or some other indicator of standard devs.. Keltner Channels.... literally take the delta of the indicator (difference between the band and the moving average) and chart it against some implied vol.. the implied vol would have to be taken from a static delta point of view... meaning ATM would have to be exactly at the money or would have to be calculated to be.. Dollar skew would be interesting..
i have an indicator that measures trending volatility of one member against another.. IE... RSI of ATR(apple)/ against RSI of ATR(SPX) ... and as well.. ATR(aapl)/ATR(SPX)
OBVIOUSLY I HAVE A TENDENCY TO OVER COMPLICATE

wouldn't these guys with large books have a complete cross book of index options and constituents?