Many people had found that shorting vix futures is a strategy with a positive edge. You can simpy look at the performance of CBOE's VPD index, which tracks the result of systematically shorting vix 1-month futures. This index had risen from the initial 100 to 215 in 5 years. This year, the index rose about 50%,a winner among numerous systematic investment strategies. However, if you short vix futures, you will be hurt if the market goes panic. You can see the VPD index dropped sharply in 2008 and 2011 in a short period. To avoid unlimited risk, you can buy OTM vix calls to protect. And the CBOE has a capped vix shorting index, the VPN, which tracks the performance of shorting vix futures and buy OTM vix call to hedge. Strange enough, the VPN index seems to have the same risk as VPD. You can look at its historical performance. The OTM vix calls seem to have no protective ability at all. How can that happen? And if I want to realized the edge of shorting vix futures, can I do any form of dynamic hedging to get it without being hurt by sudden jump of vix?
Dynamic hedging allows you to make the difference in implied / realized, however I dont think it realizes the famous VIX edge, which is contango. The best way to realize it is probably pure short deltas. Personally I use VXX, XIV, UVXY etc to realize it. One of my fav edges and definitely the one I've made the most profits from in total...I got kind of lucky though, went into short VIX at the top in 2011, take a look at XVV / XIV prices at those times and go figure the profits...
There's a paper out there (I'll see if I can find it somewhere) that evaluates a VIX futures roll-down strategy by shorting the VIX futures AND ALSO shorting an E-Mini S&P contact (ES). Since Vix generally spikes when ES drops, you get a dirty hedge.
The paper you refer to might be this one. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2094510 I've just written an article for Active Trader magazine which looks at the trade. I use an even dirtier hedge of 500 SPY per VIX future.
But doing that means you have unlimited risk on both sides. If the spx rise rapidly you will be hurt seriously.
I've seen the paper - looking forward to your article too. Do you think it still works for VIX? I am currently trying to test the idea on VSTOXX. Well, if I get enough historical data first
bs2167, i'm glad you liked the books. A new edition of "Volatility Trading" is coming out in April. It will have about 100 pages of new stuff. I think the article will be out in the December issue but I'm really not sure. I know they have a very fast turn around time. I'm fairly sure that the VIX trade will continue to work. I think generally there are two fundamental types of trades. Sometimes these are loosely called "alpha" and "beta" (very loosely). The first is an active trade that has to be worked at. It is the kind of thing that a trader makes rather than finds. An example would be a pit trader with unusual execution skill or a computer trader who has figured out how to game VWAP. The second type of trade is basically collecting a risk premium. These trades are easy to find. EG the low vol stock anomaly, the carry trade, the variance premium. These tend to exist for deep psychological reasons (this is a conjecture of mine. i don't think there is a concensus about why these exist). The trick here is to really understand what drives the premium, what the dangers are and when to avoid the trade. They are knowledge based, risk management trades. I think this VIX trade is one of these.