There are several ways to attempt to measure fragility. This BAML graph illustrates the recent divergence. I don't personally do a generic measurement of risk. For me, it's how often do I lose x dollars in a day on non-positional trades or how often am I subject to sudden non-reverting price dislocations. Since I primarily am a market maker, large non-reverting dislocations with no increase of volatility to follow through is the worst type of environment. These are certainly becoming more frequent. Not to say that there isn't plenty of compensatory good stuff going on in some markets but most MM firms are fairly specialized to 1-2 product spaces. Also. Nobody wants to sell equity index vol at these levels. The only people who are doing it are forced to because that's their main business. The best thing a retail trader can do with options is to buy individual equity names. Those get mispriced because the MMs can't practically keep on top of all the nuances of every individual stock so they take shortcuts.
I think HF returns don't really indicate much considering how much that space has changed. My view is that every pundit out there will be blabbering about his own indicator of FdM and yet nothing of true predictive value comes to life.
I prefer selling vol when the VIX is low(<20). I personally find low realized vol much easier to manage and there is "usually" enough premium in the implieds/forwards to compensate. Just thought I'd offer the OP an alternate viewpoint.
Out of curiosity, how does that work out with vol being mean-reverting? E.g. when you are selling vol at extreme (say selling 3-month variance at 25), you know that your downside is limited, at least somewhat. On the other hand, when selling vol at historical lows (say selling the same tenor variance at 15) you are taking a serious downside risk. IMHO, of course.
That's a bit beyond the scope of this thread but I'll humor you. There is serious downside risk regardless of whether i sold 15 or 25 vix. If I am selling vol it's usually done against some replication that I have going on elsewhere and I find it much easier to plug the hole when spot vol is low. At 25 vix I am pretty much just hoping mean reversion bails me out or that the juice I sold was high enough. Again, just my preference.
Indeed it's beyond the scope. However, let me translate this into my own language and you tell me if I am understanding you correctly. Are you selling vol as a part of some sort of relative risk premium strategy? And so you feel that in a low vol environment it's easier for you to find some offsetting cheap risk premium? It's an interesting topic in itself
Yes, exactly. And this is also off topic but then there is the frequency issue. How often do you get to sell a 25 vol line nowadays? If vix prints 25 tomorrow we will be lucky if the 60/90 day implied/forward vol lifts to 19. So, I'd rather sell the 15-16 line and manage the risk hole into that lift. Those 3-4 extra handles I sold at 25 vix won't give me much sleep on a 50 vix print