The general sentiment on the Convergex trading desks continues to be bearish, so today Nichaolas Colas reviews seasonal patterns for the CBOE VIX Index going back to its starting point in 1990 to see what that math says about current market risk.
Over the last 27 years, the VIX has tended to bottom in three specific months: January (15% of the time), July (22%), and December (30%). When the VIX does bottom in January, its average low reading is 12.2; today’s close was 11.9. That doesn’t guarantee that we’re at the lows on the VIX for the year, but given January’s propensity to represent an annual low it does merit your attention. And since changes in the CBOE VIX index are strongly (and negatively) correlated to equity market returns, this is a warning sign about the near term direction of US stocks. Potential hiding places are few and far between, but we’d look at yield sensitive groups like Utilities, Consumer Staples and REITs, plus precious metals.
Nothing much typically happens in January, right? It is cold out, daylight is in short supply, and everyone is economizing after spending too much around the Holidays. More recently, the “Dry January” movement seems to have emptied out the bars and nightspots in New York as well.
Yet a quick look at the history books shows that January (and not just the one in 2017) does get its fair share of excitement. A few examples:
- January 1, 1863. President Lincoln signs the Emancipation Proclamation. (Which, by the way, was a Presidential Executive Order.)
- January 1, 1959. Fidel Castro rolls into Havana.
- January 8, 1982. Breakup of the American Bell System into regional phone companies and AT&T long distance service.
- January 10, 1946. The first meeting of the modern United Nations, in London.
- January 16, 1979. The Shah leaves Iran after mass protests led by religious clerics.
When it comes to US stocks, January has an even-money chance of being the best or worst month of the year when it comes to one key measure of expected near term volatility. Since the start of the modern CBOE VIX Index in 1990, January has marked the high point for risk pricing in a given year 4 times. Likewise, it has been the low point for expected volatility 4 times as well.
A few details:
- In case you aren’t up on your options math, the CBOE VIX Index is essentially the price of insurance against a sudden drop in the S&P 500. Mathematically it is the implied volatility imbedded in the options prices for near dated S&P 500 futures contracts. But all you really need to know is that when the VIX goes up, stocks tend to go down. The range on the VIX since 1990 is 9 to +60. Lower numbers indicate little current volatility and fear, while higher numbers correlate to periods of market turmoil.
- The VIX has peaked during any given year in January four times since 1990: On the 14th in 1991, the 14th in 1999, the 20th in 2009 and the 27th in 2003.
- The VIX has troughed 4 times for any given year in January since 1990: on the 28th in 1994, the 26th in 1996, the 22nd in 1997, and the 24th in 2007.
- This shows that January has an abnormal number of VIX highs and lows; if the distribution were even through a year then January should have only 2 apiece (27 years, 12 months in a year).
- The VIX has a very strong seasonal pattern, especially as it relates to annual low points. Just three months – January, July and December – account for 66% of the annual lows for this measure back to 1990 (when it shifted from measuring expected volatility on the S&P 100 to its current look at the S&P 500 index).
- As a point of interest, January, August and October are high points for the VIX some 48% of the time since 1990. Only January is on the “Top 3” list for both VIX highs and lows.
Now, the obvious question is “Could the current CBOE VIX reading represent a low point for the year?” There are 3 reasons why it could:
A reading this low is more than one standard deviation away from the VIX’s long term average of 20. Additionally, the VIX rarely goes below 10. Statistically, it is an unusually low reading for the VIX.
The seasonal factors we outlined above show that the VIX often visits its extreme point, high or low, for the year in January. And since we know the current level is not likely to the high point for the year, it could well be a low.
Fundamentally, there are rafts of reasons why stocks may become more volatile as the year progresses. US large cap equity valuations are still high at 17x this year’s earnings for the S&P 500 of $130. President Trump’s agenda may be largely pro-business with tax cuts and less regulation on offer, but it comes with uncertainties over trade policy and tariffs. And the timing and cadence of those changes is still a wild card even if the promised reforms do eventually come to pass.
When the CBOE VIX Index moves higher, stocks tend to go lower, so the next thing to consider is what sectors might be safe parking lots during a shift in risk appetites. Assuming that whatever shock that drives equity volatility higher also puts a bid into Treasuries, the answer are dividend yield plays such as Utilities, REITs, and Consumer Staples. Gold and silver may also be safe havens, especially if the dollar weakens.
As with any historical analysis, we should always remember the old saying about history rhyming more than repeating. We do know January tends to show extremes of the market’s perceptions of near term risks, for good or for bad. And we know that at current levels the VIX highlights a complacent market. Does that assure us that things will get choppier from here? Of course not. But to be boldly bullish here is to ignore the historical patterns.
And that seems riskier than staying aware of both history and current market dynamics.
Furthermore, as Gavekal Capital's Eric Bush details, given the level of economic policy uncertainty, VIX should probably be higher...
Global economic policy uncertainty is near 20-year highs while the VIX is nearing 20-year lows.
This is an odd configuration for these two series. Usually as economic policy uncertainty moves higher, the VIX moves higher with it. That is not the case today. In the scatter plot below which plots monthly data points for the two series going back 20 years, you can clearly see how the latest data is an outlier. The other three data points surrounding the latest data point are from June 2016, July 2016, and November 2016. Based on the level of the economic policy uncertainty in the world, a regression model would have predicted that the VIX would be pushing 30 instead of hovering around 10.
All in all, it would seems more likely to us that the VIX will climb higher to close this gap rather than a swift drop in economic uncertainty.