Stocks quoted in this article:
On Wednesday, March 16, the CBOE Market Volatility Index (VIX) closed at 29.40 -- a rather substantial 35% above its 10-day simple moving average (SMA). Three sessions later, one had to wonder whether this VIX spike had really happened. By the close on March 22, the VIX had dropped to 20.21, almost exactly 10% below its 10-day. And then kept going and going and...
We don't see too many VIX extensions this large to begin with, much less such a quick turnaround. Here's a chart of VIX vs. its 10-day SMA, in red, vs. the S&P 500 Index (SPX), in black, over the past half-year.
Pretty amazingly, we hit both a six-month high and low for the VIX vs. its 10-day SMA within the space of six trading days. (Technically, it was a 10-month high, and about a nine-month low.)
I went back further in time and lowered the criteria that defined a "reversal." Going back to the beginning of 2000, I found 11 independent instances where the VIX closed less than 3% below its 10-day SMA within a week of closing more than 20% above this trendline. To avoid clustering, I count any VIX moves 20% above the 10-day as a single instance until it closes somewhere under the 10-day.
Here they are in reverse chronological order, with the date the VIX first closed above the threshold, the date the VIX closed below the threshold, SPDR S&P 500 ETF (SPY) levels in both cases, and a brief comment about what the market did going forward.
- Jan. 20, 2009: VIX 56.65, 23.3% above 10-day; SPY 80.57. Reversed Jan. 26: VIX 45.69, 4.55% below 10-day; SPY 83.68. VIX generally huge and market near what ultimately became a double bottom.
- Nov. 20, 2008: VIX 80.86, 22.1% above 10-day; SPY 75.45. Reversed Nov. 24: VIX 5.32% below 10-day; SPY 85.03. A big bounce off the bottom of a brutal bearish move.
- Oct. 24, 2008: VIX 79.13, 23.6% above 10-day; SPY 87.04. Reversed Oct. 30, VIX 6.52% below 10-day; SPY 96.3. Happened one month prior to the above; very similar setup.
- Oct. 19, 2007: VIX 22.96, 23.4% above 10-day; SPY 149.67. Reversed Oct. 26: VIX 3.56% below 10-day; SPY 153.62. Market shaky after highs; hindsight shows this as the beginning of the bear market.
- June 26, 2007: VIX 18.89, 26.9% above 10-day; SPY 148.29. Reversed July 3: VIX 3.33% below 10-day; SPY 152.34. VIX lift proved a decent buy signal; market OK for three months, as the quick reversal had no impact.
- June 7, 2007: VIX 17.06, 23.2% above 10-day; SPY 149.1. Reversed June 14: VIX 6.72% below 10-day; SPY 152.86. Random event in the middle of a churn; no market impact.
- June 12, 2006: VIX 20.96, 21% above 10-day; SPY 123.99. Reversed June 15: VIX 13.9% below 10-day; SPY 126.12. Market churns for a month, then the three-year rally resumes.
- Jan. 20, 2006: VIX 14.56, 24.4% above 10-day; SPY 125.97. Reversed Jan. 27: VIX 5.33% below 10-day; SPY 128.54. Market stalls and ultimately grinds higher.
- July 23, 2002: VIX 44.92, 22% above 10-day; SPY 79.95. Reversed July 26: VIX 7% below 10-day; SPY 85.6. Short-term rally ensues, but two more big bear moves within the next eight months to follow.
- Oct. 12, 2000: VIX 29.47, 26% above 10-day; SPY 133.13. Reversed Oct. 19: VIX 4.85% below 10-day; SPY 139.31. Market pretty close in time and price to a high it would take six years to revisit.
- April 14, 2000: VIX 33.49, 21.7% above 10-day; SPY 136. Reversed April 18 (only two sessions later): VIX 26.12, 6.4% below 10-day; SPY 144.47. Market higher two weeks out, then generally plowed for almost three years.
So, what can we tell from this data?
Well, last week did not see the biggest three-day VIX drop ever, but it did see the largest drop after first hitting our definition of "overbought." None of these other 11 instances saw the VIX start at greater than 27% overbought, and on March 16, it closed 35% above its 10-day SMA. This spike/reversal was also tied for second-quickest round trip.
As far as market relevance, it's tough to find a real pattern in previous such signals. You can pretty much disregard instances with extremely high VIX readings and low SPYs, on account of that's not the backdrop we're facing now. So off with January 2009, November 2008, October 2008 and July 2002. Likewise, January 2006 and June 2007 look extreme the other way. A 20% move in the VIX from a very low level contains more noise than signal, in my opinion.
So, that doesn't leave us with much, beyond a small sample size. The very first instance -- April 14, 2000 -- looks most similar to now in terms of the VIX level and duration. But even there, the market had already had a choppy three months. And, of course, this was on the heels of the tech bubble.
Most intriguingly, this go-around took relatively little in the way of market motion to set off this whole lightning-quick "fear/complacency" cycle. SPY dropped about 2.9% on March 16, then more or less gained it all back by the next Monday. We're not talking "flash crash" here.
My gut says the combo of a VIX explosion followed by a near-instant reversal does not bode well for the market. Instead, it seems like a sort of crescendo marking the end of the VIX downtrend. The data suggests it's more of a process than a one-off event, though.
Disclaimer: The views represented on this blog are those of the individual authors only, and do not necessarily represent the views of Schaeffer's Investment Research.