Stocks quoted in this article:
Remember the iPath S&P 500 VIX Short-Term Futures ETN (VXX)? We haven't checked on our favorite exchange-traded note (ETN) in quite some time now.
VXX, of course, did quite well recently, albeit for a relatively brief time. It hit all-time intraday and closing lows on Dec. 5, then proceeded to rally 32.5% close-to-close over the next week and a half, before peaking on Dec. 16 at 34.63. And then the market explosion happened, and with it the inevitable VXX implosion: it dropped 16% in three sessions.
On the year, VXX is down 29.7%, which actually isn't a terrible year for VXX on its own. Contango in CBOE Volatility Index (VIX) futures always knocks down VXX over time. Thing is, VIX itself is up 20% in 2014, so using VXX as a general volatility or portfolio hedge worked about as well as always. That is to say, it didn't work well at all.
I can put a positive spin on it, though! VXX has hit 316 new all-time lows in its illustrious 1,484 trading-day history; a 21.3% "success" rate. The pace of ineptitude has slowed, though, in 2014, with only 36 new lows in 246 tries (11.39%)! And buying VXX into an overbought VIX has actually worked OK, at least in the very short term. The following table is a VXX-centric version of the overbought VIX table we often run. What happens if you buy VXX on the first close when VIX closed 10% above its 10-day simple moving average and hold for five, 15, or 25 trading days? Here's each play since the invention of VXX in 2009.
Keep in mind that VXX almost always declines over time, so negative returns in and of themselves aren't surprising. Buying VXX on overbought VIX and holding for five trading days actually worked well in 2014: four wins in five tries!
Going back for all of VXX history, though, it still hasn't paid off. It incurred a median loss of 4.53% vs. a median loss of 2.24% on any random five-day hold. The average loss is 1.95%, vs. a 1.43% randomly timed purchase and hold. Buying and holding for 15 trading days is worse off of overbought VIX (-10.46% median vs. -7% random median). If you use averages, you get a different story, but that's wholly because of a huge win if you bought VXX on the Aug. 29, 2011 overbought VIX. Out 25 trading days, there's little difference.
As always, the best advice is to never buy and hold VXX. It's fine to play in the very short term. But if you do that, the numbers suggest you are better off anticipating a VIX pop than trying to ride one in progress. Unfortunately, that's almost by definition a crapshoot. So, just stick with good old-fashioned index puts!
Disclaimer: Mr. Warner's opinions expressed above do not necessarily represent the views of Schaeffer's Investment Research.