I am currently reading his journals and his strategy seems reasonably simple. Buy 121/132 flies on vol spikes cover the position at around 20 % gain. Cut the losers when they exit the body (no hedging with underlying). It also seems he only traded 5 stocks/indexes. He also posted some nice returns of + 30% after commisions when commissions was $2 a contract. He also has a spread sheet that qoutes flies (i assume for comparison purposes). All of his positions also have a small delta to them when put on. Since butterflies are a kurtosis play. It makes sense to put them on within 2 weeks to expiration. (Which he did). Has anyone talked to Atticus who could provide some more depth to his strategy. I mean it is so simple I am surprised with the consistency. In stocks KISS makes sense since everything is linear but with options KISS does not usually work. Anyways it would be great to here some opinions.
Can you expand on this, specifically the timing piece? I'm familiar with kurtosis as it relates to distributions but I've never considered it was possible to trade. Is the fly then a bet that the kurtosis is understated and increases toward expiration?
1-2-1 flies are a good pre-earnings IV trade. Enter before earnings - exit after earnings. If you don't think the stock will move much after earnings a fly might be the way to trade it. Ideally you want the options to expire at the body strike. The debit to enter the trade is relatively low.
Most of your PnL on a butterfly is from volga (how each of your strikes move with respect to the change in vega). Kurtosis means very fat tails. So when you put on a proper butterfly you are selling expensive ATM vol and hopefully buying cheap wings relative to the realized distribution
Just like a ratio collar (Vega neutral) is a bet on skew, a Vega neutral fly is a bet of kurtosis (or, if unwound before expiration, on slope and convexity respectively). However, in a notional neutral structure it’s just a bet on distribution.