Stocks quoted in this article:
Another week, another -- well, nothing much. There were new highs in the S&P 500 Index (SPX), and some modest drifting in the CBOE Volatility Index (VIX). We did expect that last week would stay quiet, and it certainly lived up to those expectations.
But, hey, we did see one modestly interesting development in the volatility complex: iPath S&P 500 VIX Short-Term Futures ETN (VXX) didn't make new all-time lows! In fact, it hasn't made an all-time low since July 3, when it bottomed at 26.95. That's 40 trading days, and if this goes on another week or so, we can start talking about where it ranks in terms of VXX non-ugliness.
VXX outperformance last week makes some sense. The pending holiday weighed on VIX, given that realized volatility tends to lag on holiday weeks. VXX proxies a 30-day VIX rolling future, and a holiday now won't have much impact on where traders 30 days from now expect to see implied volatility 30 days beyond that.
That VXX all-time low on July 3 coincided with a VIX of about 10.30. Since then, VXX is up 3% while VIX is up about 16%. Believe it or not, that's a reasonable performance for VXX.
Selling VXX into VIX pops still looks like a winning strategy. Here's an update of the table of VXX performance after VIX has closed greater than 20% above its 10-day simple moving average.
I looked at five-, 10-, and 15 trading-day returns in VXX if you bought it on the close of sessions where VIX closed overbought. At the bottom I show median and average VXX returns in each category vs. median and average returns if you bought VXX on any random date.
And just for laughs, I included VIX and VXX values on the close of the days VIX went overbought.
Yes, VXX adjusted for reverse splits really was that high back in 2009.
You actually do better on an average basis if you buy VXX after VIX gets overbought. But that's really misleading -- it's thanks almost entirely to the monster returns in August 2011. The median comps provide a better example of the typical VXX ownership experience. And you do quite poorly over the five- and 15-day windows. Over 25 days, you actually do slightly better than random buys.
And it's very important to remember that no matter how you look at it, owning VXX for even a short length of time is a terrible idea. Unless you catch an August 2011-type pop -- and probably catch it before the VIX pop even happens -- you're doomed. Buying VXX randomly is merely less bad than buying after VIX lifts.
Disclaimer: Mr. Warner's opinions expressed above do not necessarily represent the views of Schaeffer's Investment Research.