Quote from rew:
Nothing in your post explains why trading iron condors requires Black Scholes to be correct. There are any number of other pricing models that assume fatter tails than lognormal, and if options are priced according to accurate assumptions an iron condor will have zero expectation. Yes, once in a while you will get wiped out. But in between you make enough to pay for the occasional blow up.
Typical iron conder statistics: win $500 19 times in a row, lose $9500, win $500 19 times in a row, lose $9500. The expected return is 0.
If you believe that far OTM options are still under priced, despite the volatility smirk, then buy them. If you are wrong and options are instead priced correctly your payoff will look like this: lose $500 19 times in a row, win $9500, lose $500 19 times in a row, win $9500, etc. Again, this is a 0 expectation strategy-- it doesn't have the dramatic drop offs on the P&L curve that iron condors do but it's just as lacking in positive expectation. (Of course, allowing for the bid/ask spreads and commissions both strategies will lose money in the long run.) Your one big win with the Lehman strangle no more shows that long strangles are a winning strategy than the fact that Howard has made a good deal of money selling hundreds of credit spreads shows that he has a winning strategy.
Atticus' gripe is that Howard has not demonstrated that he has an edge, i.e., positive expectation. It is in the nature of selling credit spreads that you can have a long string of wins that makes it look as though you have an edge, when in fact you just haven't rolled the snake eyes yet that wipe you out. But this has nothing to do with whether BS is correct. Even if the lognormal distribution of prices was correct and BS was used to price options iron condors would still blow up often enough to give you 0 expectation.
