obviously just catching up...doing taxes all day mine, kinds, friends of kids.... sorry for the redundent posts...looking at my fav journal for relaxation
As expected -- the large % bull runs cause the most damage, from declining strip vols and gamma losses. Nice to see it bags >5% on average.
absolutely ; for the real nondirectional trades I would assign put bias to capitalize on the spike in 60d vols when market collapses ( while the 30d PnL is still done on intrinsic value bases). This also could be done by using above the current price strike as ATM (both shorts at 1020 when price is 1000), right CAL?
Hmmm, I don't follow why you'd want put bias. Yes, a bear-delta time fly would earn a bit more due to ramping of strip vols, but it would lose a bit on skew. The backtest shows us that the neutral time fly does well in moderate declines. The data suggests trading a bull-delta time fly. It would increase the expectancy on the trial-set you've compiled. Small edge loss from -skew on the straddle/strangle is offset by delta-position. The bull delta time fly will carry larger put vegas as the only long vol $gain in a decline is the long wing put -- so increase the wing-put exposure and increase bull-market gains. Backtest the time fly with the straddle + strangle above the SPX by 15 handles; i.e., SPX at 1100 / 1115 short straddle / 1100x1130 outside strangle. I didn't catch your strike selection. Use the same wing-depth; simply skew the position 15 handles or approx. 5d to the upside. I'd imagine that the average over the sample would increase to 7-8%. Only if you have the time. The point being that the position does well due to the "compactness" of strike selection[+net vegas] into declines, but loses on rallies due to gamma. We can mitigate some of the gamma losses and retain bear performance. To smooth out the equity curve will require an initial bull delta. I realize it's curve-fitting, but we're not talking lagging indicators here.