Friday was a day in which the VIX spiked 26.5% and the SPX dropped 3.1%. On the surface, that does not sound unusual: the market plunges and the VIX simultaneously spikes in the opposite direction as sellers get a little panicky.
Over the weekend, I was reviewing the performance of the SPX after a 3% drop and saw the classic mean reversion pattern of a market that usually does very well after a big one day selloff. I looked at the VIX and noticed an even more predictable pattern of mean reversion following a VIX spike. Again, this is what is what you might expect -- particularly if you spend a lot of time with this data, as I do.
Then I saw something that I was sure had to be some sort of mistake: when the VIX spikes 20% or more in one day (see Adam at Daily Options Report for his thoughts on this phenomenon), the VIX reverts, but the SPX does not tend to perform particularly well going forward. But how can this be? I had just looked at data I had crunched to show that following a SPX drop of 3%, the index had a history of exceeding ‘normal’ performance by three to ten times over the course of the next few weeks and months.
As it turns out, prior to Friday, there are 27 instances since January 1990 in which the VIX has spiked 20% or more in a single day. As luck would have it, there are also exactly 27 instances in which the SPX has dropped 3% in a single day during that period. It turns out, however, that only 8 of those 27 instances are the same session. In analyzing the data, I have come to the conclusion that as a rule, the 3% drops in the SPX are better indicators of a future bounce in the SPX than the 20% drops in the VIX. I will be sure to elaborate more on this observation going forward, but thought I should offer up this conclusion this morning for those who are still wondering about whether Friday’s session has bullish or bearish implications.