Lazlo Birinyi and his associates recently completed a study of the VIX. Their conclusions were that the VIX is not a very good indicator of future market movements. Instead, it is simply a mirror image of current market conditions. Their conclusions prompted me to look at the VIX over the past twenty years. Truth be told, GT made me do it. LOL He brought the study up in chat and challenged me to back up any argument for or against with real statistics. Always good advice...
First, let me state that I agree with Mr. Birinyi's conclusion. I overlayed a chart of the S & P 500 on a chart of the VIX from 1990 to present. At least on the surface, it even looks like a mirror image. Granted, I do not have the time nor the inclination to replicate his study methods. Anecdotally at least, it looks like he is on to something.
Next, I became curious if the VIX had any other uses outside of options strategies or if there was a VIX trade with a high probability of success. Looking at the charts as well as monthly VIX averages, I found that it was not unusual for VIX levels to fall into and stay at levels in the low teens for extended periods of time during Bull markets. The VIX was at levels below our current level(16 and change) from September '03 until June '07. A similar pattern emerges in the period from Feb '92 until March '96, except for a single spike around March '94. From this review, I would conclude that a low VIX is in no way indicative of a near term correction or bear market. Shorting the market using a low VIX level as your only indicator seems a very a bad decision to say the least.
In fact after reviewing this data, the only other thing that became apparent to me was that the VIX rarely falls below 11. A 10 handle is rare and a 9 handle is almost unheard of. With a possible multi year time frame in mind and as a hedge against your long positions, I would buy VXX when the VIX falls below 12. and consider adding more if the VIX falls below 11. Unlike actual equity trades, the VIX has a floor that is nearly set in stone. Trying to catch a falling knife doesn't apply in this instance. Keep in mind, the VXX has some tracking issues so it won't be a perfect correlation in performance. I would sell half of the position when the VIX reached the 35 level and set a stop at 12% or so(or set a very tight stop when it broaches the 30 level). Sell out completely at a VIX level of 40. You may miss a little more upside, but at least you've protected your profit. For the record, both bear markets involved in the 20 years of data had VIX levels broach 35 and at least, fleetingly, broke the 40 level.
http://fwd4.me/Te source for daily VIX data
also used Yahoo Finance ^VIX for monthly averages