Why the CBOE Volatility Index (VIX) shouldn't shape Fed policy
I flipped on financial TV yesterday morning, and lo and behold, former Treasury Secretary Lawrence Summers was on talking about the CBOE Volatility Index (VIX). That wasn't all that impressive in its own right -- I mean, I'm pretty sure Fed governor types all know there's something called the VIX and it's a "Fear Index" and all.
No, what impressed me was that he used VIX to suggest Fed policy. Specifically, he said the Fed shouldn't hike because the VIX was high.
So, by "impressed," I meant, "Wait, what?"
Fed policy guided by VIX? I know VIX has gained more and more prominence over time as an indicator, but seriously. If VIX has one iota of impact on Fed thinking, we're all in trouble. But hey, if they want to look at it, I have an economics degree and know VIX stuff -- make me a Fed governor!
OK, seriously. My opinion of the Fed rate hike debate is that it's wildly over-covered and kind of misses that point. The Fed is by definition part reactive and part anticipatory. But let's face it: it's mostly reactive. How else can we derive forward-looking numbers than via looking back first?
And come to think of it, that describes the VIX too? So maybe Summers is on to something!
VIX predicts future volatility, but the best way to "predict" future volatility is to look at recent volatility. Over the course of VIX history, the one-day move in VIX has a -0.71 correlation to the one-day move in S&P 500 Index (SPX)/SPDR S&P 500 ETF Trust (SPY).
Another way of saying it is that you can explain 50% (0.71 squared) of the VIX move on a given day if we simply knew what SPY did. So already we can see it's going to be the largest single factor -- and we can also see that literally half of the VIX move tells you nothing you wouldn't already know if you just looked at SPY.
So, what explains the other half? A little bit is just the calendar: day of the week, day in the cycle, holidays, et al. And a little bit is simply anticipation: Is news pending? Is news out? Etc.
But that's all over the course of time. In 2015 VIX has become even more tied to simply aping the SPY move. The correlation is -0.86, meaning the Index Move of the Day explains nearly three-quarters of what we see in VIX. Throw in weekends, holidays, and Imminent Fed Obsession Days, and there's really not a lot unexplained these days. And that correlation is pretty persistent. Here's how rolling 20-day backwards-looking correlation looks this year:
The measure peaked at -0.97 on Aug. 24… which, technically speaking, means that VIX was its absolute least insightful on the worst and most volatile market day in eons. We didn't need a vol index to tell us it was volatile.
Other than a blip around May, VIX has had abnormally high negative correlation to SPY pretty much all year. Remember" -0.71 is "normal."
What's it all mean? Well, VIX is getting more and more predictive of what we've already seen. Using it as some sort of forecasting tool becomes less and less sensible. It's always fine for us to try to derive short-term trading cues from the VIX, but let's hope the Fed doesn't actually use it to instruct policy in any way, shape, or form.
Disclaimer: Mr. Warner's opinions expressed above do not necessarily represent the views of Schaeffer's Investment Research.