The VIX is extremely overbought right now, but still requires caution on the fade
We're overbought in the CBOE Volatility Index (VIX) again. And not just any kind of overbought -- rather, it's something akin to super, extra special overbought.
On Friday, VIX closed 35.09% above its 10-day simple moving average (SMA). That's not a very common occurrence. In 6,548 trading days, VIX has closed at least 35% above the 10-day only 34 times. That's 0.52%.
It's becoming a bit of a non-rarity, however. You might say we're in a bubble of VIX bubbles. It's happened seven times since August -- and that includes the mother of all (relative) VIX pops. VIX closed a whopping 121.65% above its 10-day SMA on Aug. 24. That broke the record of 81.72% set ... one session earlier. No. 3 on the list happened on Aug. 25.
Before that, you have to go all the way back to August 2011 (OK, that's not so far back). The record then was 74.76%.
I'm not sure why VIX-plosions are getting more common. My best guess is that the way I define this, you need a relatively low VIX to begin with. In 2008, for example, VIX was routinely in the 30s (at least), so the percentage moves on particularly ugly days was not always epic.
But whatever. The big question, of course, is what it means going forward. I culled out incidents so as not to duplicate readings on the same event. They start the first close of VIX 35% above its 20-day SMA. Friday marks the 20th distinct incident.
Anyways, here's how it looks, with one-week, one-month, and three-month returns vs. average and median returns of randomly timed holds. All go back to the start of the SPDR S&P 500 ETF Trust (SPY) in February 1993:
I used a little color coding for when the returns deviated over 100 basis points (BPs) from the random medians. Again, you'll see a bias towards later dates. Three incidents since last August, five back only to October 2014. It only happened four times for the first 14 years of VIX, then 16 times in the last nine years. So maybe it's not a low VIX thing; VIX itself really is getting more volatile.
Fading extreme VIX pops looks OK, but unexceptional. You definitely have an edge in the very short term, though it did not work well the last two times. You did well three months out, also. 2008 was an outlier, all around. One month out is simply mediocre.
And this doesn't factor some ugly "paper" drawdowns. SPY dropped about 8% in three days after the Aug. 20, 2015 signal, then rallied back to make the one-week loss look reasonable.
It's always worth noting that crashes happen off oversold conditions. It's also worth noting that crashes are low-likelihood events, and it's generally not the best idea to overestimate the probability of a huge tail occurrence.
I wish I had a simple answer here, but I don't. I think it's right to fade this action gingerly, knowing that there could be some serious pain first.
Disclaimer: Mr. Warner's opinions expressed above do not necessarily represent the views of Schaeffer's Investment Research.