Virus: Could it be the Catalyst to Change Sentiment?
Heed Keynes’ message about markets staying irrational...
Being a contrarian just to be a contrarian is rarely profitable; but investors need to be mindful of the message that sentiment indicators are nearly-universally sending.
There are lots of extremes; but perhaps comparisons to circa 2000 have been overdone.
Before we get to the matter at hand, we are all mourning the loss of the great Kobe Bryant and his beautiful daughter Gianna. What a heartbreaking period for his family and his fans. His legacy across the world of sports will live on for generations to come.
I’ve often remarked that it’s rarely wise to be a contrarian just for the sake of being a contrarian—especially where the stock market is concerned. The gains over the past year have been remarkable; not just for their magnitude, but also for their consistency. But the power of the momentum move brought with it very extended investor sentiment conditions. Sentiment can move into the extreme optimism zone and stay there; while being right for an extended period. It’s difficult to pinpoint the inflection point in advance. As John Maynard Keynes famously said, “The market can stay irrational longer than you can stay solvent.” What extreme optimism does though is establish a greater amount of vulnerability to the extent there’s the arrival of a negative catalyst.
Before putting more meat on the bones of sentiment in this longer-than usual missive; a shout-out is warranted to the source for much of what follows. I have been an avid user of the work of SentimenTrader for many years courtesy of their robust database and historical studies; but also their unique look into the behavior side of investor sentiment (in contrast to the many attitudinal measures I also track). For sentiment tea leaf readers among our readers, they offer a free daily newsletter, which I highly recommend.
Friday: blip or sign?
Friday’s stock market weakness was neither sufficient in scale or scope to suggest a sentiment-exacerbated pullback is upon us; but if the weakness persists, we may point to the coronavirus as the negative catalyst. That is clearly the case in Monday’s pre-open session (at which point this report was submitted).
To put some color on Friday’s market behavior, the S&P 500 gapped up early in the day; opening above its highest close over the prior five days, but closing below its lowest close over the prior five days. According to SentimenTrader (ST), this type of reversal pattern historically led to short-term weakness over the next week or so.
The S&P 500 was down more than 1% near the close, but recovered a little in the approach to the close. While not a major loss obviously, the index had gone 73 days without a loss greater than -0.9%—ranking among the longest streaks since the S&P 500’s inception in 1928. Historically, there were mixed results in the ensuing month or two; but when ST broadened it out to -1.0% streaks, the returns were weaker, though not yet applicable obviously (although perhaps applicable after today’s close).
Crowd goes wild
There are very few exceptions, as of last week, to extreme levels of optimism among sentiment indicators—save perhaps for domestic equity fund flows. The chart below shows the Crowd Sentiment Poll (CSP) from Ned Davis Research (NDR). I like to show it as a sentiment proxy given it’s an amalgamation of seven individual sentiment indicators—both attitudinal and behavioral.
Crowd Optimism Elevated
Source: Charles Schwab, ©Copyright 2020 Ned Davis Research, Inc. Further distribution prohibited without prior permission. All Rights Reserved. See NDR Disclaimer at . For data vendor disclaimers refer to , as of 1/21/2020.
As you can see, optimism is currently in the extreme zone; during which historically the market has struggled on an annualized return basis. Data like this is not foolproof by any means—case in point would be 2017, which was a year of healthy returns and low volatility in spite of sentiment staying mostly in the extreme optimism zone. Do note though that the ultimate peak occurred at the end of that year—setting up the correction that ensued in early 2018. (Conversely, the plunge in sentiment to the depths of the extreme pessimism zone at the end of December 2018 acted a near-perfect contrarian set up for the huge rally that ensued.)
Two of the components of NDR’s CSP are the weekly survey from the American Association of Individual Investors (AAII) and Investors Intelligence (II), which is a measure of advisor/newsletter sentiment. As you can see in the first chart below, AAII’s bulls have jumped from a low of only 20% last October to more than 45% last week. Clearly, this is not yet near the extreme bullishness registered in early January 2018 (which was “perfectly” ill-timed), but worth watching. The second chart shows II’s “bull ratio” (bulls/bulls+bears). The last time newsletter writers were this bullish was in early fall 2018—optimism that was “rewarded” by the near-bear market that hit bottom on Christmas Eve 2018.
Individual Investors Bullish, But Not at Extremes
Source: Charles Schwab, Bloomberg, as of 1/23/2020. AAII= American Association of Individual Investors.
Newsletter Writers Very Bullish
Source: Charles Schwab, FactSet, as of 1/17/2020. Bull ratio=bulls/(bulls+bears).
In reference to II, the bull ratio has been above 75% for 13 weeks—tying for the eighth longest streak in 50 years. According to ST, the S&P 500s annualized return was 5.4% when the II bull ratio was above 75% versus nearly double that when the ratio was below 75%.
“Smart money” not so smart lately
My favorite of ST’s behavioral sentiment indicators are their Smart Money and Dumb Money Confidence Indexes, which I show often in these reports as well as during client presentations. ST is tracking what these two cohorts of investors are doing (i.e., how they’re positioned); with the “smart money” being the large commercial hedgers/institutional position traders/speculators and the “dumb money” being the smaller odd-lot traders/speculators. There are no opinions represented in these indexes—they are real money gauges of how the cohorts are positioned. The “smart money” has historically been mostly right at extremes; while it’s the opposite for the “dumb money.”
But in keeping with the Keynes quote I cited at the beginning of this report, for the past couple of months, the “smart money” has been less-smartly positioned than the “dumb money,” which has successfully ridden the rally (see chart below). ST concedes this is a longer-than-usual streak; but it doesn’t diminish the risk that pullback likelihood is high. As of early last week, the 50-day average of the SM/DM spread neared -60% for only the second time in the history of the data (since 1998). Historically, even after the 50-day average was approaching -55%, upside momentum was challenged in the subsequent few months.
“Dumb Money” Has Been Right (So Far)
Source: Charles Schwab, SentimenTrader, as of 1/24/2020. Confidence Indexes are presented on a scale of 0% to 100%. When the Smart Money Confidence Index is at 100%, it means that those most correct on market direction are 100% confident of a rising market. When it is at 0%, it means good market timers are 0% confident in a rally. The Dumb Money Confidence Index works in the opposite manner.
Volatility (or lack thereof)
Last week brought the first -1% weekly decline (Friday’s close relative to the prior Friday’s close) in 16 weeks. Historically, when previous streaks of low volatility ended, the S&P typically saw more weakness over the next couple of months. According to ST, this has been particularly true over the past 50 years, since most of the bullish cases occurred prior to the 1970s.
Source: Charles Schwab, Bloomberg, as of 1/24/2020.
Mean reversion coming?
With stretched sentiment has come some stretched technical indicators as well. The S&P 500 has spent about six weeks above its upper Bollinger Band (BB), using a 50-week moving average, as you can see in the chart below. Over the full history of the S&P 500, when stocks became similarly overstretched, returns over the next month or so were on the weaker side: for the 31 prior occurrences, the percentage positive was 48%, with a median return of -0.7%, according to ST data.
Bollinger Band Breakout
Source: Charles Schwab, Bloomberg, as of 1/24/2020. For more info information on Bollinger Bands®, see .
We’ve been getting an increasing number of questions with regard to the leadership of the technology sector (notably, some of its biggest names) and whether we are seeing shades of circa 1999-2000—which of course ended in spectacular fashion. I agree there are shades of similarity; but they’re not (yet) casting the same ominous shadow. In the lead-in to the peak in 2000, tech stocks’ outrageous valuations brought up the S&P 500’s trailing price/earnings (P/E) to about 45; while today it’s about 22. In addition, there is much stronger earnings support for the sector than was the case two decades ago.
There are important macro differences between the 1999-2000 era and today—including the fact that the Fed was raising rates in 1999, while lowering them in 2019; and the fact that oil prices were surging in 1999, while range-bound over the past year.
Like in the late-1990s, into the 2000 peak, technology stocks have the dominant market players—rallying in 14 of the past 17 weeks. Since the mid-1990s, streaks like that have been relatively rare. Based on 25 years of ST data, streaks like that lead to some large losses for tech stocks over the ensuing one-to-three months (especially in 2000 and 2012); it was actually more negative for the S&P 500 overall. Adding some technical flavor, as of last Thursday’s close (before Friday’s weakness), nearly 50% of the S&P 500 Technology Sector’s members were overbought. ST’s notes that a percentage that high has only been matched a handful of times since 1989; with a downward bias for tech stocks over the subsequent three-to-six months.
Defensives, and playing defense
There is an unusual tenor to the recent stage of the bull run in stocks. It’s not just tech stocks that have attracted investors’ buying interest; but highly-defensive sectors as well. Sector breadth statistics from ST show that utilities, staples and health care have recently seen a tremendous spike in the percentage of stocks not only hitting new highs, but jumping outside of their volatility bands and becoming overbought at the same time. It’s one of the strongest defensive sector breadth surges in three decades. As ST points out, a sector like technology is better able to handle speculative activity—or at least extremely eager buying pressure—than sectors like utilities.
In keeping with Keynes’ admonition about market irrationality, we are maintaining our oft-expressed view that in riskier times, tried-and-true disciplines should be dusted off, if not already being utilized. We continue to recommend that investors remain at their long-term strategic equity allocations—but use swings in either direction to rebalance back toward those targets. For those investors who have enjoyed the ride to recent new highs for U.S. stocks—but don’t view success as perfectly top- or bottom-ticking inflection points—make sure your portfolio has not gotten out of whack relative to your long-term targets.