By “technical analysis” I mean the practice of drawing inferences about likely future prices based on the visual patterns displayed on a chart. We can exclude rules-based, objective strategies here: those are beyond reproach because they can easily be verified. For example, “buy asset x when the n-period moving average crosses above the m-period moving average” is a strategy we can test historically and rigorously – a computer can do much of the hard work of confirming whether the strategy is valuable. I’m talking here about the subjective, lines-on-charts stuff.
I am generally skeptical about subjective technical analysis because, by definition, it’s so hard to verify. We can spit daily pivot points and support levels at each other just as easily as lottery numbers, and without post-trade performance checks of the validity of the former, there’s no reason to think they are any more meaningful than your great-aunt’s process for deriving Powerball bets.
But for the sake of argument, let’s assume that things like support and resistance levels are really, really meaningful indicators of future market action. They’re still totally irrelevant when it comes to the CBOE Volatility Index (VIX). Here’s why.
The most important reason should be intuitive: VIX doesn’t trade. VIX is just a statistic. It’s an estimate of the annualized implied volatility of SPX derived from options at a weighted 30-day horizon. It’s not a share entitling you to the cash flows generated by a business in the physical economy. Support and resistance and trend lines and momentum effects all depend on the existence of buyers and sellers in the asset being analyzed. But you can’t trade VIX directly, so the VIX can never find “support” because no one previously bought VIX “shares” at that price level. And for those of you who are thinking ahead, the same goes for ETPs: no one has ever bought iPath S&P 500 VIX Short Term Futures (VXX) shares at 16 as a “deep value” play.
Calendar and skew effects also introduce a lot of noise. A lot of VIX ticks tell you absolutely nothing about investor sentiment because the spot index level can change solely on the basis of the day of the week. VIX often declines into a weekend or a holiday as options prices change to reflect the time the markets will be closed, and VIX often moves up on Mondays for the same reason.
Similarly, as SPX prices move higher and lower, the weights applied to individual SPX options in the VIX calculation will change. VIX might be higher today solely because the SPX volatility skew curve is steep and higher-IV options are receiving more weight after a modest SPX decline. Mark Sebastian recently wrote up a great example of this type of phenomenon at Options Profits. Options traders look at the IV of particular strikes directly; technicians who don’t pay attention to the options are likely to make false inferences based on those VIX fluctuations.
Now, this is not to say that VIX and volatility data isn’t meaningful – quite the opposite. Options traders know that volatility is generally mean-reverting. And given the high average inverse correlation to SPX prices, we might expect to see some (inverse) stock-like VIX behavior at short timeframes simply because of that relationship. Analyzing SPX and VIX joint reactions to events can be worthwhile. The term structure of VIX futures communicates a lot about investor expectations. And so on – there are plenty of ways to mine volatility data usefully. It’s just that those useful methods don’t generally require clever lines on charts.
At the time of publication, Jared Woodard held positions in SPX, VIX, and VXX.