It really is a matter of "perspective". If you, in the context of your portfolio, value your option positions by looking at the notional value of the underlying shares (and manage risk/leverage accordingly)... then intestinal fortitude becomes a little easier to build. I think in that sense, it would be like "being long the underlying". If you value your positions by looking at premium alone... you're never going to be -200% by being long underlying, that's for sure.
Short index straddles have a historically-high Sharpe ratio. Does that mean we should all sell atm straddles on the S&P?
It really doesn't. Sharpe ratio is notoriously useless for any option strategy. Investing in most financial instruments is (for the sake of argument) a 50/50 thing. You buy GOOG today; 50% chance it'll be up tomorrow, 50% it'll be down tomorrow. It's a coin-flip. If someone comes to me claiming that they've flipped 10 heads in a row with a fair coin (and/or have a high Sharpe ratio), I'm impressed, and I will invest in their coin flipping skill. Options are completely different. The probability of gain/loss for any option position is determined strictly by the manager, by selecting the strikes they're dealing with. A manager might have flipped 50 heads in a row... but if they're using a coin that's biased to turn up 99% heads, then I'm not at all impressed. That's precisely what MANY option strategies are: flipping a biased coin. You simply can't determine skill by looking at their returns alone.
I am not talking about a discretionary strategy. What I meant is a predetermined static option construct that gets triggered by some indicator such as Kurtosis or a similar market trigger.
Well, actually, if a manager came to me and told me they achieved a high Sharpe over time with ATM straddles using strikes half a standard deviation away from the underlying... I personally would invest in them. If a manager came to me and told me they were selling 1 std dev out, I'd tell them they should be profitable at least 68% of their trading months. They better have at least 5 years of track record with fewer than 10 losing months and low DD, so I can see with some confidence that they have skill. If they're selling strikes 2 std dev out, then they should be profitable 96% out of the year... in 20 years (240 months), they would be expected to have 10 losing months. So, unless they have 20 years of track record with fewer than 5 losing months, I'm not investing.
I have no idea what you mean by this. Any option "construct", regardless of the trigger, has a probability distribution of the manager's selection. You simply can not look at historical returns without understanding the probability distribution the manager has targeted.
Have you looked at the risk-reversal premiums at 15 on VIX? Any strategy using such a trigger is doomed.