A Few Takeaways from the VIX Spike

So, I just spent a few days in Montreal doing research on the CBOE Volatility Index (VIX) (or hockey research ... something like that). Anyway, I came up with a few takeaways:

1. I can't understand a word in French.


2. Uber's a really great invention.


3. Hockey is much more popular than inverse exchange-traded funds (ETFs).

Anyway, I'm back, and as luck would have it, nothing moved all that much while I was gone. It's really as if October never happened. VIX now sits at mid-September levels. But, alas, it did happen, and VIX did spike in a big way, and it did trigger a market buy signal in the highly sophisticated "20% Above the 10-Day Simple Moving Average" system that we track.

That happened one month ago yesterday, so we now can update the table of results. And so far, so good:

To refresh, the system looks at returns if you buy on the close of the first day that VIX closes at least 20% above the 10-day moving average and measures returns on a one-month and three-month hold. I also include data on simply buying and holding for one month and three months on any random date. I don't include overlapped triggers. For example, VIX had a separate violation on July 31 that's not on the table because the data would have double-counted the July 17 data. Finally, this only goes back to 2009. I'm of the theory that this "rule" works better when the VIX pop bucks the longer-term VIX trend.

Anyway, the one-month trade worked quite well this go around. The 7.05% return is the strongest since a buy on Dec. 28, 2012. And, it marks the eighth straight time the trade "won," although the last win was all of 0.01%. More importantly, the average returns (2.42%) and median returns (4.26%) blow away the average and median returns of randomly timed one-month holds (1.31% and 1.83%, respectively).

A trade like this -- and really any trade that tries to time a fade into an oversold stock market -- will tend to win more often than it loses. The problem is that the relatively few losses can sometimes dwarf the wins in magnitude. That is, markets implode from already oversold conditions, so you could walk right into a disaster when you fade an overbought VIX.

If you played along this time, you had a bit of a scare before we turned in a big way. On day two (Oct. 13), VIX closed 40% above the 10-day. It then peaked about 55% intraday on Oct.15. And, the SPDR S&P 500 ETF Trust (SPY) dropped 4.5% at its bottom in just over two trading days. It was another win, but we humans tend to hate losses more than we enjoy wins. And, for a few days, it looked like a pretty awful signal. So, definitely some psychic pain before realizing any gains.

Disclaimer: Mr. Warner's opinions expressed above do not necessarily represent the views of Schaeffer's Investment Research.