Definitely are large funds dedicated to short vol. They are unsophisticated, IMHO. I once went around the country analyzing a handful of the better short vol funds for a HNW client. I wasn't impressed, but I'm a dickhead so take it as you will.
Is Credit Spread really profitable?
- Thread starter cmukesh19
- Start date
Definitely are large funds dedicated to short vol. They are unsophisticated, IMHO. I once went around the country analyzing a handful of the better short vol funds for a HNW client. I wasn't impressed, but I'm a dickhead so take it as you will.
https://www.reuters.com/article/us-...l-mageddon-losses-in-u-s-stocks-idUSKCN1GC29H
LJM was one of the more famous short vol funds. I told my HNW client these guys were "100% guaranteed to blowup" back in 2016.
How did I know? Because they would say things like "vol is historically overpriced". Wrong answer!!!!
Vol is historically overpriced IN LOW VOL MARKET ENVIRONMENTS. It's underpriced in high vol environments.
The key is knowing when to adjust your approach. Any hedge fund that does the same thing every day will have trouble. A short vol fund that just sells vol w no eye on market environment will blow up eventually. Guaranteed!
LJM was one of the more famous short vol funds. I told my HNW client these guys were "100% guaranteed to blowup" back in 2016.
How did I know? Because they would say things like "vol is historically overpriced". Wrong answer!!!!
Vol is historically overpriced IN LOW VOL MARKET ENVIRONMENTS. It's underpriced in high vol environments.
The key is knowing when to adjust your approach. Any hedge fund that does the same thing every day will have trouble. A short vol fund that just sells vol w no eye on market environment will blow up eventually. Guaranteed!
This really makes sense.
There is no way to do this consistently. If one can, there are thousand ways to profit.
What would happen if you did this Feb 5th 2018?
Just want to discuss, thx!
"No way to this consistently" is exactly what I am challenging with my data.
Feb 5th was a Monday, in which the week before saw the vix almost double. Many support levels crushed along w follow through outside moves should tell a trader to flaten out and wait. Holdinga short position through that weekend would be a poor decision. Even if you did, as long as you were short spreads rather than outrights you would have simply lost max value once.
Ex: selling the 1stdv / 2stdv condor should equal short premium of 22% of max spread width.
Therefore you'd lose almost 4 winners in that 1 max loss. It happens. As market continued lower in Feb 5th there were long vol indicators firing. Long vol is tough but even if you just stay flat after the Feb 5th until vix drops back to 15, you'd get out of Feb 2018 flat to down a little.
What LJM did was bad risk management. It was a result of an unsophisticated, caveman approach to short vol. "Vol is overvalued"... Until you blow up. Idiots.
@Jerkstore, I get your premise - selling high IV makes more sense in a low-vol overall market - but why would anybody sell a 1SD IC in high vol for 22% of spread??? Of course you'd lose money doing that.
Right now, with VIX at 20.11, ~1SD IC in SPY with 25DTE is paying more than 33% of spread width (0.33nat - 0.36mid.) And SPY IV is not what you'd call stellar at the moment; if it was, that premium would be much higher.
Not so much disagreeing as curious about your take on this - because this is pretty much standard behavior that I've seen for premium during high VIX.
Last edited:
I think coming up with those buckets (< 18, > 28, > 35) retrospectively makes this strategy overfit.
Jeff1228 asks a really good question about Feb 2018. VIX was under 18 on Friday, Feb 2. If you're going to throw additional guidelines on like VIX almost [sic] doubled the previous week, support levels crushed, and others, then you don't have a simple VIX-level based strategy either. You're overfitting something that's already overfit, and each of those additional guidelines also have questionable utility.
One question I would ask about making claims about high VIX involves sample size. Do we have enough of a sample size to make any claims about high VIX environments at all? If not, and if those levels are just enough to filter out the worst historical periods for the strategy, then it's yet another instance of overfitting.
Not sure what you're asking. Take any atm straddle (aka 1 st st deviation of an expiration). From the atm sell the 1 straddle difference out strangle and buy the 2 straddle difference out strangle .
If the straddle is $10, you will probably see a midpoint value of $1.25 in the put spread and $1 in the call spread = 22.5% of the total spread value of $10.
IV level is irrelevant.
If the straddle is $10, you will probably see a midpoint value of $1.25 in the put spread and $1 in the call spread = 22.5% of the total spread value of $10.
IV level is irrelevant.
Yes my methods are more complicated than 1 liner buckets. I'm simply pointing the crowd in the right direction.
We have enough data to know we don't want to be short vol into high vol environments , yes.
I'll share a story. I was in Mexico City years ago at some conference and met the CEO of Quantlabs. I sorta corner the guy and remind him he publicly proclaimed zero losing days in last 3 years (this was back in 2011 I think). So I ask him, how? How does any trading group never lose money?
His answer paraphrased " I have about 30 identified opportunities, and algo strategies ready to take advantage of those opportunities. On any given day I typically only run 6 algos. Why? Because my job is to determine market environment, then to match the proper strategies to that market environment. We do this every day, and paper trade the algos that we aren't live trading. *My most important job as a risk manager is to know what risk will work today. No algo works in every environment.*"
^^ that's what I try to do.
Ah, OK; appreciate the clarification. Seems I didn't understand your shorthand of "1stdv / 2stdv condor".
Quick review of SPY for the expiration I mentioned gives me an ATM straddle of just over $11, with a put side spread worth 1.63 and the call side at 0.72, which is pretty much in the general area. Not sure why I'd ever want to sell such a critter, though.
Expected value of random risk is zero minus costs.
Good thing I don't sell it randomly, but rather into favorable market conditions.