VIX Below 20: All-Clear Signal, or Intermediate Top?

OK, we've often talked about the positive side of an overbought CBOE Volatility Index (VIX). But what about the reverse -- the negative side of underbought VIX, in honor of breaking under 20?

Is 20 a key level? Well, it's the whole number that most closely approximates the all-time mean in VIX. And, hey, we're all about mean reversion here, so it has to mean something.

So here's the "system." We go long the market when VIX first closes below 20, after it has just spent at least three trading weeks closing above 20. We hold for one- and three-month time frames. How did we do? Well, take a look for yourself:

Recently, the trade has worked out well. The last four instances, going back to February 2010, the trade has performed quite admirably on the one-month time frame. And it did very well the last three instances, looking at the three-month time frame.

Going back further though, not so much. All in all, the one-month trade saw an average return of negative 0.87% in one month and negative 2.8% in three months, compared to returns of 0.65% and 2.01% in randomly timed one-month and three-month trades. Using medians instead of means didn't change the conclusions. I bring this up because it sets up again now, as we just broke below 20 after nearly seven weeks closing above.

So should you sell now? Maybe, but I wouldn't sell because of this. I find it interesting in principle that, by and large, a dip in vol to these levels after a period of "high" vol has looked more like an intermediate top than an all-clear signal.

But, realistically, I'm picking somewhat arbitrary numbers here. I used 20 because it's nice and round and pretty average for VIX, particularly psychologically. Most find VIX "cheap" under 20, even though that's not really the case. The VIX median back to 1993 is 18.26, so if nothing else, I'd use that as the "cheap" point.

And cheap isn't really that meaningful a concept in a vacuum. VIX is only cheap if it underprices future volatility. Fifteen may be fat if the next month sees realized vol of 10, for example. Since we don't know the future, we don't know whether VIX now is underpricing or overpricing future risk. That's all a long-winded way of saying I like using moving averages better, and putting VIX in the context of realized vol.

Bottom line is, it all comes down to whether you believe this was a blip within the longer-term bull, or whether it was the start of more of a down move. Since 2009, all of these VIX pops have been blips. That will end someday, but for now, we just can't say yet.

Disclaimer: Mr. Warner's opinions expressed above do not necessarily represent the views of Schaeffer's Investment Research.