The equity market question of the hour is “When will U.S. stock market volatility return?”
History tells us this is purely a question of “When and why”, not “if”. A look back to 1990 – the starting point of the modern CBOE VIX Index – shows that domestic stock market volatility goes through distinctly pendulum-like movements that takes years to develop. The daily VIX can run annual averages below 15, as it did from February 1993 – September 1996 (43 months), February 2005 – October 2007 (32 months), and August 2013 to July 2015 (23 months). Once the annual VIX average crosses 15 to the upside (as it did on July 13, 2015), it typically spends years above this level and can average +20 for much of that time.
So for all the chatter about low actual/projected volatility at the moment, that’s where we are now. The pendulum is swinging, albeit slowly, back to higher volatility. That explains the “When”: it is already happening. As for the “Why”… We’ll find out soon enough.
Fill in the blank:
“U.S. equities are currently showing little price volatility because ______________________”
I took a poll of friends and coworkers on the buy and sell side for some answers, and the most comprehensive one came from Pete Coleman, our own head trader at Convergex. His answer:
“… stocks are fully/fairly valued, economic growth is slow but positive, and interest rates are low but not heading much higher.”
Let’s take a quick turn though each of Pete’s points:
#1: Full Valuation. According to FactSet, the current bottom up earnings estimates for the S&P 500 for 2016 and 2017 are $119/share and $134/share, respectively. With the index at 2180, that works out to valuation multiples of 18.3x and 16.3x, respectively. Top down earnings estimates for next year are understandably lower (strategists can cut numbers without ticking off company managements, unlike the single stock analysts that feed the bottom up estimates) at $128/share for a 17.0x multiple.
Now, no one is a point smart on P/E multiples, so 17x and 18x are essentially the same thing. But keep in mind that S&P 500 earnings have not grown in 3 years and that makes paying a rich multiple on this market something like plunking down Gucci money for Wal-Mart store brands. There’s nothing wrong with Wal-Mart, or Gucci for that matter, but you want to see value commensurate with quality. The best thing you can say about U.S. stock valuations is that they are “Fair”. And by that we mean that enough investors see the same value that volatility is low as a result.
See the most recent FactSet report here (it’s free, very complete, and up to date): https://www.factset.com/websitefiles/PDFs/earningsinsight/earningsinsight_9.2.16
#2: Slow Economic Growth. First half GDP growth in the U.S. has averaged 1.0%. Retail sales (July) are running 2.3%. Light vehicle sales stopped growing 18 months ago, albeit at high levels. Personal savings rates (5.7% in July) are double pre-Crisis levels (late 2007). New home sales have picked up, but existing home sales have stalled (like autos, at high levels, but not rising). August ISM survey results were 49.4, below the 50 line that denotes contraction. Labor market data has been fine, but productivity declined 0.4% in Q2. Core CPI (+1.6% in July) and PCE prices (+0.8% in July) are still below the 2% Fed target.
On the bright side, such as it is, the Atlanta Fed’s GDPNow model of future economic growth is calling for 3.5% in Q3, well above the Blue Chip consensus of 2.7%. This model is best used later in the quarter (ideally, 30 days before the first release of the data), but we are 2/3rds of the way through and the model is still remarkably sanguine. So there is some hope that the back half of 2016 will play catch up after a weak first half.
See the Atlanta Fed model here: https://www.frbatlanta.org/cqer/research/gdpnow/?panel=1
And the Richmond Fed’s excellent slide deck on the U.S. economy here: https://www.richmondfed.org/-/media/richmondfedorg/research/national_economy/national_economic_indicators/pdf/all_charts.pdf
#3: Low and stable rates. It isn’t just the equity market with a case of the yawns; 10 Year U.S. Treasuries have traded in a tight band of 1.40 – 1.60% since mid-July and the spread between 2 and 10 Year U.S. sovereign debt has been 0.76 – 0.90 over the same period. Yes, that 2-10 spread has been tightening this year but we’re still far away from the zero point where recession becomes a real possibility.
Just as importantly, markets of all stripes (stocks, bonds, currencies) are locked into a “Fade the Fed” mentality that heavily discounts the possibility of a rate increase before the December meeting. Fed Funds Futures currently place only an 18% chance on a move to 50-75bp at the September 21st meeting and little more than a coin toss (52.4%) at the December 14th FOMC meeting.
See the CME Group’s FedWatch Tool here: http://www.cmegroup.com/trading/interest-rates/countdown-to-fomc.html
In short, Pete’s explanation has all the necessary components – cash flows (earnings), economic outlook (sustainability of those earnings), and interest rates (the discount rate for those cash flows) – to make for a useful framework. Market participants have sufficient confidence in each leg of this stool to comfortably sit on their stock portfolios. The big question now is “How long will that confidence last?”
To get a historical perspective on prior periods of similar low-volatility market behavior, we pulled the daily data for the CBOE VIX Index back to 1990. Since the VIX is closely tied to actual volatility, this is a useful measurement of both current price action and expected near term moves in the S&P 500. To get a sense of “Structural” volatility expectations through time, we ran a historical one year average VIX level using the daily close for the VIX. The chart with our findings (1991 to present) is in the attachment to this note, and here is what the data tells us about prior periods of low actual/expected volatility:
Even though the long run average of the VIX is 20 (19.74013, to be exact), there are long periods of time when it can trade consistently below 15 on an average annual basis. That’s not quite one standard deviation (that is 8, or 7.9867 if you want to be precise), but far enough away from the long run mean to be noticeable.
The three times this has occurred since 1990 are:
- February 1993 to September 1996, when it averaged 13.4 over this 43 month period.
- February 2005 to October 2007, with an average of 13.2 over 32 months.
- August 2013 to July 2015, with an average annualized reading of 14.4 over 23 months.
Once the annual average VIX crosses back over 15 on its way higher, it is years before this measure of volatility begins to decline again.
The picture here is essentially one of a pendulum of market volatility, swinging back and forth in very long movements. Volatility can remain suppressed for years, and we outlined 3 periods where average annual VIX readings are consistently below 15. The last one of these ended a little over a year ago. Yes, it may feel like we are still in it (today’s VIX close was 12.3) but recent bouts of market churn (China last year, Brexit earlier in 2016) are holding those averages above 15. Today’s one year average, for example, is 17.2.
As to where we go from here, the historical patterns are clear. We had our run of low volatility and it ended last year. We are now in a new phase where volatility will rise.
The logical question is, of course, “Why?” Prior periods of low average VIX readings did not always end badly, with equities rallying from 1996 to 2000 and from July 2015 to the present day. Still, coming full circle to Pete’s comments – fair/full valuation, a slow economy and low interest rates all supporting stocks – it is harder to make the bull case. What else could go right?
Which leaves the opposite question as the elephant in the room: “What could go wrong?”