I guess the guys who sell hurricane insurance either re-insure some of the risk or hold sufficient capital to pay claims. I am not convinced that Karen does either .... I think Sosnoff gave an example of the kind of trade she might do .... useful for illustrative purposes if nothing else ... - SPX at 1820, - 56-60 DTE - Sells Put at 1630 strike for $6.25 - You can repeat this trade every 60 days, or 6x per annum, - Maximum return is 6x $6.25 = $37.50 x100 = $3750 per annum - To keep the maths simple, to generate a return of say 37.5%, you would need to allocate capital of $10,000 to this trade. - Simple scenario; SPX crashes 20% to 1450, you would need at least $18,000 capital to cover the loss from 1630 - 1450 strikes, plus additional capital to cover extrinsic and initial margin Anyone else care to explain how she can make 20-40% per annum selling tail risk and be able to survive the next market dislocation. Nickels waiting for steamroller ......
It was actually a very clever trade. It didn't work out for him (obviously). He was the only seller of vol to an unprecedented market of buyers. Every bank/hedgefund would have sold that vol if they could. But no one could manage the marked-to-market on their personal compensation. I would never compare his trade to anything a systematic short vol trader does here. It was supply/demand related and he exploited a real edge other than "2 SD moves don't happen."
Well, then the reinsurer is taking that risk, right - somebody carries that catastrophe expsure at the end (unlike the mortaility risk, which is much more balanced, for example). Really, who cares what Karen does, she's been trading since 2009 (5 years) and trading 3-month options gives us a sample size of of 20-25 trades, not very statistically significant at all. Pekelo et al, in their usual manner, are getting hung up on various stupid details of the strategy and performance. For all you know, everything she said is true, but she got lucky on the path dependence. The real question is "is it profitable to underwrite tail risk?" and the empirical evidence is that it is. E.g. selling 1 month var from 1990 to today is perfectly profitable, with worst DD whiping out 3 years of your returns. There are good reasons why risk premium is overpriced and it has to to with the micro-structure of the financial industry. As an outside player, retail investor has a little bit of a loophole - though I would not be surprised if that closes soon. Now, the strategy itself can be smart (with some sort of actuarial analysis, holding extra reserves, making bets in a remoted account etc) or stupid, which does not really change the general expectation, but introduces strong path depenency to your final result.
There where plenty of ways to sell vol in a limited downside way at the time - I bought ITM up and out calls, for example. PS. Buffer exploited a number of edges over the years, main one being AAA counterparty (sell long-dated puts, don't post, sit tight - that one worked out for him).
I think I lost my agenda, let me find it first.... I think the way how she adjusts the positions after a sudden big move what makes her strategy different from others doing the same (and helps her to survive). There are some red flags about her, but at this point we don't have enough info to doubt her (and Sosnoff's) credibility... Otherwise I just like to discuss interesting people/strategies, if you don't mind...
My assumption is that she beat the market big time. She might have ended up with a loss, but the market was down -38%. Otherwise big money wouldn't have came in with only 1 year good performance. So it had to be at least 2 years, me (and common sense) thinks... Or big money is actually dumber than I think it is...
I don't disagree with the premise .... it's just that she doesn't carry sufficient capital to withstand a worst case draw down Option trading is a pretty simple game .... short premium .... don't go bust doing so!
20% is a considerable move in something like an index (SPX in this case). If this event was to occur in such short notice to where people didnât have time to respond, I would personally speculate that there would be some type of mean revision in that same time period - similar to the flash crash. On the other hand, if itâs just a slow increase/decrease, there may not be any revision but a least time to respond. In my own experience (although limited), Iâve noticed that large moves in magnitude in short periods of time more often mean revert by some percent. I witnessed/traded this recently in /NG a couple of weeks agoâ¦Hell of a ride!