BUNK 21

WHEN THE VIX IS HIGH, IT’S TIME TO BUY

And when it’s low it’s time to go. Or is it?

For readers not familiar with this investing Wall Street “wisdom,” good for you—your life is infinitely better. The VIX is the ticker symbol for the Chicago Board Options Exchange (CBOE) Volatility Index. It’s meant to show the market’s expectations of 30-day volatility for the S&P 500. There’s nothing wrong with the index; it’s forward-looking and well constructed from a range of S&P 500 index options—both puts and calls. And it pretty much does what it’s supposed to do—tells you the market’s expectation for future volatility.

But that’s it! Volatility is volatility. It doesn’t tell you where stocks are going to go next—similar to the problem with beta (Bunk 19). But folks who read charts like carnival fortune tellers read tea leaves (and with about as much success) say you can replace “volatility” with “fear.” The VIX is sometimes called “the fear index.” So a spiking VIX means spiking fear. And because stocks love to climb a wall of worry, increased fear should mean capitulation selling and good times for stocks ahead. And an absence of volatility means less fear—even excess complacency—and possibly a downdraft ahead. Hence, the saying goes, “When the VIX is high, it’s time to buy. When the VIX is low, it’s time to go.”

Coincident Plus Relative Equal Useless

In theory, it’s not bad. It is true stocks like to climb a wall of worry. So does it work? Normally, I advise looking at long periods—the more data the better. But VIX fans tend to be short-term trading oriented, trying to capture short-term swings. Again, for most people, a short-term view is wrong. But to be fair to those cute little VIXens, let’s look at a short period where the VIX worked beautifully.

Figure 21.1 shows S&P 500 returns (dark line) and the VIX (lighter line) for 1999—a pretty volatile year. And look! Overall, the VIX was pretty high all year, and peaks and troughs on the VIX tended to mirror troughs and peaks on the S&P. Success! Had you bought at relative high points and sold at lows, you’d have timed near-term stock swings pretty darned well.

Except if you’d just bought and held the S&P 500 through 1999, you’d be fine too—up 21.0 percent.¹ And you wouldn’t have had all those extra transaction costs or cap gains taxes to pay. Plus, you wouldn’t have had to watch the VIX like a hawk—which your spouse likely prefers.

But how would you have done had you bought the VIX peaks and sold the lows? I have no idea. Why? Well, how do you define it? Do you buy at every VIX peak? Or just the major ones? And how do even you know if it’s a peak? Look at the graph again—the highs and lows are all relative—you wouldn’t know a peak had formed until after the fact. The VIX peaked higher at the start of the year, then stocks moved choppily sideways. Mid-August there was a peak—but lower than many points earlier in the year—though stocks were pulling back sharply. Weird! And the VIX bottomed just before that pullback—signaling a near-term peak. Cool! Except there were some VIX bottoms late in the year when stocks were generally moving higher. How would you know which peak or trough to act on?

Figure 21.1 When the VIX Is High, Part I—1999

Source: Thomson Reuters, CBOE SPX Volatility Index, S&P 500 price return from 12/31/1998 to 12/31/1999.

Worse, by the time you can look back and see the VIX peak or trough, you’ve missed the corresponding S&P 500 bottom or top too (if there was one). There’s no lag in stock performance, giving you time to build conviction you are indeed seeing a VIX peak. Coincident, relative indicators might look cool, but they are close to worthless when making forward-looking forecasts.

Useless . . . But Also Inconsistent!

But hang on—that’s assuming the VIX always works. It doesn’t. Like most all “technical” indicators, for every chart you show me that works, I can probably show you lots more where it falls apart. (See Figure 21.2.)

Figure 21.2 When the VIX Is High, Part II—1995

Source: Thomson Reuters, CBOE SPX Volatility Index, S&P 500 price return from 12/31/1994 to 12/31/1995.

1995 was a great year—US stocks up 37.6 percent²—almost a perfect, straight shot up! But the VIX was just static—noise. What do you do with it in a year like this? It gives you no guidance. It doesn’t even hint to you at the start of the year to get in now! VIXens will say the VIX works during heightened volatility. Ok, but 1995 was one heck of a volatile year—it was just all upside volatility! (Folks forget—up or down—it’s all volatility.) Overall, the VIX doesn’t look very helpful this year. There are peaks and troughs, but they seem disjointed from market action.

In some years, the VIX works great. In others, it’s just static. How do you know when the “right” short time period for using the VIX has started? You don’t. You couldn’t, not until it was over. And then, it only helps if you can build a time machine and go back in time to take action precisely at the right peak or trough. And if you could do that, you would license the time machine and be a multi-billionaire 50 times over and not fuss with stocks.

VIXens claim the tool works when there is “heightened fear”—that’s how you know. But how can you measure that? Only hindsight bias lets you look back and say, “Well, I wasn’t nearly as fearful then as I am now.” Folks never remember fear and pain right—memory fades the pain—it’s a survival instinct. (If you don’t believe that, ask any woman who’s given birth.)

Sure, VIX can sometimes work really well. But you can’t forecast from it. And when you look at it longer term, there’s no way to know when it will work or won’t. So when the VIX is high, the VIX is high. That’s about it.