Long time lurker, first time poster. Itâs clear a lot of the folks were scared away from credit spreads and iron condors due to the market volatility in recent years. Most notably, the meltdown during the last quarter of 2008 and the flash crash of May 2010. I see many folks are very critical of some of the credit spread trading journals because, as one quickly learns, one extreme market move can blow out credit spread traderâs profits and capital. Nevertheless this is such an enticing strategy because of its high probability of success. Iâve done some back testing using a sort of credit spread hedge with the VIX that Iâm certain someone has thought about before. I wanted to share the idea and see if you someone can explain why it would not be a good plan. The idea basically involves hedging your SPX credit spreads with a long VIX 40 call of the same expiration month using a 3:1 ratio (3 SPX credit spreads for each VIX call). Why 40? If you look at the historical meltdowns mentioned above that killed credit spread traders, in each case the VIX reached 40 or greater. Using TOS thinkback and your classic 9% - 10% profitibility bull put spread 45 - 60 days in advance one can normally get a VIX 40 call from anywhere between $17 to $40 on average. Using this criteria, I tested by selling 3 1140/1130 OCT 08 bull put credit spreads in late August 08 and 3 1020/1010 NOV 08 bull put credit spreads in late September 08 (both dooms day trades ). You see that each of these reach maximum loss. Now add the very cheap VIX 40 call and see what happens? The 2008 crash is now survivable to a credit spread trader. In fact, if your timing and stomach acid level was right, November reached a few profitable points. Do the same test by putting on a MAY 10 vertical spread in March 10 and the flash crash is profitable as well. If market fundamentals or stomach acid level dictate a period of volatility may be looming you could increase the ratio to 3:2 for even better protection. I understand the VIX is based off futures, but back testing seems to show that this could be a life saver. The flip side to this is how much profit is lost when all is well? Clearly that will vary based off market conditions and the price of VIX calls, but under stable conditions it seems to me an affordable insurance policy netting only a 1.20% â 1.50% profit sacrifice for the insurance. For example a VIX 40 call for JUN 11 as of this writing is $35. If I put on 3 SPX 1230/1220 receiving a credit of $255.00 (risking $2745.00, 9.29% profit), the VIX 40 call for $35 buys me a comfortable hedge. Thatâs a 1.28% cut into the credit received. Any thoughts?
Off the top of my head, i can right away suggest that VIX over 40 does not mean that your front futures will settle over 40. In May 2010, front vix futures traded around 35. What you are suggesting is protecting your OTM gamma with OTM vega/volvol. It is pretty easy to backtest if this has worked well in the past, especially if you make some reasonable assumptions about the term structure of vols etc.
VIX calls trade off of VIX futures, NOT the spot VIX. The market could drop sharply with VIX hitting 25 and your VIX 40 Call will not budge at all if the futures show no indication of pushing that high. So basically you will still lose money on your credit spread and your hedge will also lose money. VIX options far OTM are horrible hedges for credit spreads since they trade against VIX futures and not spot VIX.
Thanks for the comments. Assuming by "settle" you mean expiration? I suppose I wouldn't suggest waiting until expiration in these scenarios. Normally at this point you're just trying to stop the bleeding. The exit plan would roughly be to close out the entire position (doomed credit spread and profitable VIX call) when total losses are equal or better than 20%. I understand that the VIX call trades off futures and not the current volatility, but back testing does confirm that in each of the devestating events to credit spreaders (OCT, NOC, DEC 2008 and MAY 2010) that each months VIX expired over 40 with ample oportunity to close it out when it is above 40 beforehand. You are suggesting it is horrible because it is not directly correlated...I understand and agree with that point. However, it doesn't appear to be horrible when it comes to being profitable in those scenario's. If VIX 40 isn't the magic number, maybe bump it down a notch to VIX 35 or 30, it still would seem to be an affordable hedge. A more conservative trader could increase the SPX credit spread to VIX call ratio from 3:1 to 3:2 and still be profitable. I'm not saying that this plan would be profitable in black swan events, rather survivable...meaning a 20% or even 30% loss would be acceptable and prevent a complete blowout of the capital at risk.
+ I recalled in 2008 there is a time where the intrinsic value of the VIX is actually negative, e.g. VIX >40, but the intrinsic is only ~35 ..
I've thought about this but never tested it. As you say you would need to exit both at or near the same time. The market could tank and the VIX would go up, then the market stabilizes lower and the VIX might go down. Can you test it for the little dip we had a couple of months ago that wiped out some spread traders? Also, it seem that the long sides of the spreads might be an unneeded expense, so you could just go short and use some of the extra premium to buy VIX calls. Or maybe do VIX spreads. PowerOptions started offering something along these lines, I think they call it Safety Net or something.
It's a non-fungible forward - there is no such thing as "intrinsic value" in VIX, just like there is no such thing in Eurodollar futures, for example.
I gonna start up a journal and see how it goes since I'm green. For now here are the open positions: -30 SPX May 11 1340.00 Put 30 SPX May 11 1330.00 Put 30 VXX May 11 24.00 Call Credit spread = $2400
Just to note the differences with the OP, he said to use 3:1 ratio of SPX:VIX, and to buy VIX 40 calls. Don't know how VIX correlate with VXX, but that might pu the VXX calls out at 45 or higher (VXX=23.78, VIX=15.62). Even the 40 VXX calls are only .09 now.
Despite the very real problem of futures/spot fidelity, etc, I think your biggest problem is one of misleading data analysis. That is, I see your problem not as one of strategy, but as one of implementation. When backtesting, you can pick strikes that work perfectly, but you have no idea going forward if you've backtested a "sweet spot" for the strategy as a result of the particular circumstances of the chosen time period. It's seductive to think that the most relevant stress test would be the Oct 08 meltdown, but as they say, "maybe, maybe not." My suggestion is to look at as many variations of your basic approach as possible to see how robust it is. That is, 30d verticals, 5d verticals, high vol, low vol, vix calls struck at 20, 30, 40, high vol of vol, and all manner of ratios. Then test with market advances, declines, minimal changes, etc. Be sure to leave enough untested time series so that you can test out-of-sample when you decide on an approach. You've got a fair amount of work ahead, but hey, there's no reason not to spend some time on a potentially profitable approach to trading.