Can Historically Massive VIX Pops Predict Future SPY Moves?

The VIX action of late looks a lot like 2011 -- which means we could have three more months of this left to go

Aug 27, 2015 at 8:37 AM
facebook twitter linkedin

This CBOE Volatility Index (VIX) pop isn't exactly in the rearview yet -- but hey, we can still throw out some interesting comps to VIX pops of years past!

VIX and the S&P 500 ETF Trust (SPY) have an inverse relationship, as we all know. Over the course of VIX history, there's a correlation of negative 0.71. Square that, and it implies that the move in the S&P on a given day explains about half of the VIX move. That sounds about right, since VIX ostensibly looks forward. It proxies implied volatility 30 days out, and that implied volatility is more or less an estimate of the volatility the S&P 500 Index (SPX) will actually realize.

So, another way we look at it is that VIX is half-looking backward, and half-pricing forward. And since we know backward, we really want to solve a bit for "forward." In other words, what's the unexplained move in VIX in a given day?

Since inception, VIX moves about negative six times the move in SPY. Ergo, if SPY drops 1%, we'd expect to see VIX rise 6%. Of course, that assumes a linear relationship -- and in reality, VIX often reacts a bit more when the move in SPY is larger. If we look at the 100 largest one-day VIX moves ever, they moved about negative 6.7 times the SPY move. If we narrow that down to the top 20 VIX one-day moves, it's about negative eight times the SPY move.

Anyway, with all that in mind, I wanted to see how the recent pops stack up in the context of VIX-story. Here are all the instances where VIX popped 30% in one day, going back to 1993, as well as the ratio of the VIX move to the SPY move.

VIX pops since 1993

We actually just saw the third- and fourth-biggest one-day VIX moves, in percentage terms, since 1993. Pretty impressive! If you're of the opinion that it looks similar to 2011, well, that's the only other time VIX pops clustered like this. We see three within a 10-day period in August 2011, as well as another one in September, and then a final blast in November before the all-clear whistle blew. So, only three more months of this to go!

And how about if we rank these pops by the degree to which VIX moved vs. SPY?

VIX pops ranked by SPY move

Well, we're kind of in the middle. We overreacted a bit this go-around, but not historically so. And it doesn't say a lot for future market moves; it's pretty much all over the map.

So I guess it's more evidence that these pops were historic, but looking at it through the one-day lens, not terribly predictive.

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

Grab Your FREE 5 Rules for Trading Stock Trends Report



Special Offers from Schaeffer's Trading Partners