one thing people i think fail to realize is how many expressions they are making by doing such things.. adding complexity without much edge doesn't make sense alot of times.. your speculating about roll cost.. and how well the options on the vxx have that priced in.. its not like you can just come up with an equation to tell you the number of deltas you need to short based on how many short vxx shares you have on.. MESSY i'm sure there are quants out there that have a decent idea how to handle the additional parameter of "roll cost" in pricing those puts.. but they are making a market probably and making spreads to get an edge.. unless you have a solid view on term structure steepness, vola direction, and spy direction.. spy can go down while vol stays low.. contango curve can flatten reducing roll cost and causing the put to lose value.. hell the market can go up and vol can go up.. your talking about trying to hedge a third order derivative...
Shorting SPY calls is more risky than shorting VIX puts. SPY has no upper limit and may continue to march higher and higher for months. VIX, on the other hand, is range-bound with 13-14 being a low end of that range. Similarly, if VIX shoots up to 80, selling VIX calls (and VXX calls for that matter) is a lower risk strategy as VIX will not stay at 80 for an long period of time.
I agree with you. However, as I said there are many variations. One might go long a bear put spread on vxx and long a bear call spread on SPY....
similiarly selling vix calls is a lower risk strategy? go back to the drawing board haha.. Rocky man.. all i'm saying is your argument makes it sound like there is a smooth linear relationship between all the variables in which you "clearly" can make money.. vix can go low and stay low your valuating more recent data higher.. the curve can flatten... etc.. i think your shooting off the hip.. my thought is. if you have a view on some particular thing.. find a way to isolate it and then take a limited risk position.. your adding a few parameters that you don't know much about and thinking its giving you an edge..
Yes, it is highly complex. I agree... However, what I want to say is this: Lets suppose ATM IV of VXX is 80%. VXX has a strong trendiness. Lets suppose you find a stock that also has an ATM IV of 80%. It is very likely that this stock does not have the strong trendiness that VXX has. So buying a straddle on VXX and shorting a straddle on the other stock in a dollar neutral way might generate profits. It is the trendiness that matters. Let me give you another example. Stock A has an IV of 30% and a price of 100 USD. Stock B has an IV of 30% and a price of 100 USD. Both stocks will have the same ATM call and put prices. However, suppose that stock A has a strong trendiness, while stock B is more or less mean-reverting. In this case, you should be buying straddles on stock A and selling straddles on stock B. This is why VXX is so interesting. The trendiness of VXX provides opportunities. So it comes down to the fact that BS (= Black and Scholes, not Bullsh*t) does not consider trendiness...
now your throwing in new variables.. implied correlation between delta and implied correlation between vola.. why don't we just turn it into a complex mix of constituents and our view on vol,delta, correlation against our view on the index's and throw term structure risk on top of that..
You're taking my words out of context. Selling VIX call is a low risk strategy IF (BIG IF) VIX is at 80. However, doing so when VIX is at 15 is high risk. This is similar to iron condor: you don't put one on when IV of the UL is low. You wait for the high IV and then open an iron condor. Same with long/short straddle. Low IV - go long. High IV - go short.