Is It Time to Hit the Bunkers?

SPY returns have been pretty good when the VIX/VXV ratio tops 1

Jan 7, 2016 at 8:36 AM
facebook twitter linkedin

So North Korea is setting off bombs. Is it time to find some duct tape and jump into the bunkers? After all, nuclear war is likely one of only two things that scare even the toughest of investors.

Fear is a choice. So is bidding up options and prompting overbought alarms in CBOE Volatility Index (VIX) indicators. But for whatever reason, we keep falling just short of setting off the bells. We've looked at VIX vs. its own 10-day simple moving average the last couple days. Bloomberg highlights another indicator. VIX vs. the CBOE 3-Month Volatility Index (VXV).

VXV is VIX methodology, but with a 90-day option. The thinking is that as VIX goes above VXV, it indicates we're overbought. Bloomberg notes this:

"Since the beginning of the bull market, the S&P 500 gained an average 0.46 percent in the three sessions following instances when the ratio reached a value of one. That compares with an advance of 0.2 percent on average in a three-day span since March 2009."

Well, we didn't quite get to 1 on a closing basis this go around, though we did as recently as Dec. 11, 2015. Anyway, the data goes back to July 14, 2006, so we're talking 9.5 years. And it includes a quiet stretch, a violent bear move, then a persistent rally. In other words, we've seen just about everything.

There's very small correlation between the VIX/VXV ratio and forward SPDR S&P 500 ETF Trust (SPY) returns. Here's a scatter plot of the ratio vs. SPY returns two weeks forward:


Kind of an amorphous blob. The correlation is .08741. For one-month returns, it's 0.03579.

But what about the extremes? Specifically, what if the ratio goes over 1?

It's happened 446 times, so its not terribly uncommon (about 25% of the time). But when we cull out one-month "incidents" (and overlaps), we get it down to 61 times.

And the results are decent. If you go long when the ratio gets over 1 and hold for a month, you gain an average of 1.91%, and a median of 2.59%. That's favorable vs. returns of 0.73%/1.46% over all times.

That's the good news. The bad news is, again, we didn't hit the 1 level yet. In fact, we're not quite a month out from the last signal in December, so it wouldn't have "counted" anyway.

Also, it basically has gotten you long for half of the last decade, so it's not the most refined signal. It does suggest, though, that like basically every other overbought vol indicator, it makes sense to fade fear.

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

JUST RELEASED: The Complete Guide to Earnings Season Profits



Special Offers from Schaeffer's Trading Partners