What The Marshmallow Test Means For Today’s Market

By Nicholas Colas of DataTrek Research

The Marshmallow Test is a legendary 1970s experiment in child development which seemed to show that some people have an innate ability to defer gratification, a key life skill. More recent research shows that all it really measures is how reliable a child’s home environment may be. For Story Time Thursday we apply the same idea to capital markets. Investors are constantly assessing environmental conditions when deciding to allocate capital. When they see uncertainty increasing – as now – they pull back. Peak unpredictability creates market bottoms.

Today we’ll discuss the “Marshmallow Test”, a topic we touched on twice in 2020 but not since then. How can almost-May 2022 remind us of 2020 so much that we are revisiting a theme from that year? Read on …

The setup: back in the early 1970s, research psychologists at Stanford University developed a simple test to measure a young child’s ability to delay gratification. Here is the basic structure of their experiment and its core findings:

• Offer a child (aged 3 to 5 years old in the original studies) their choice of common treats: a cookie, or a piece of hard candy, for example, or (yes) a marshmallow.

• Then tell the child that they will be left alone for about 15 minutes and, if they don’t eat the treat, you’ll give them another one when you return.

• Some children in the study waited the allotted time and received the second treat, but others did not.

• That outcome is hardly surprising, but it was never the point of the experiment.

• Here’s what was: the researchers tracked their subjects’ academic and personal progress over subsequent years. The ones who “passed” the Marshmallow Test tended to have higher SAT scores and do better in school. They were also less likely to develop substance abuse problems later in life or be incarcerated.

Given the Marshmallow Test’s seeming ability to predict academic success in even very young children, many American private schools started using it to screen potential applicants in the 1980s and thereafter. And, as you can imagine, parents started prepping their 3- and 4-year-olds to pass. I (Nick) have many friends here in New York City who spent hours with their toddlers doing exactly that.

The twist: 40 years after the original study, a doctoral candidate working at a homeless shelter had the observation that few of the children she worked with every day in that setting would pass the Marshmallow Test. More importantly, this had nothing to do with their innate abilities. Growing up with persistent food and shelter insecurity, they would eat the first treat because experience had taught them this was the right strategy.

She created a preamble to the Marshmallow Test, where the child subjects would be promised arts and crafts material but in some cases the researcher would then tell them that there were none left. Those children receiving the materials tended to “pass” the subsequent Marshmallow Test. Those who were disappointed by the researcher were more likely to “fail”. If the adult couldn’t deliver on crayons and paper, they probably weren’t going to come across with the second treat either.

The upshot: the Marshmallow Test says more about a child’s environment than it does about anything innate in their personalities. If they are growing up in a household where adults’ promises are predictably kept, they are more likely to believe the researcher’s promise of a second treat. If they do not have that baseline confidence that the world is a predictable place, then eating the first treat right away is perfectly logical.

How all this relates to investing and trading, in 3 points:

#1: The perceived reliability of an environment strongly influences human decision-making. The capital in a portfolio is analogous to the first treat in the Marshmallow Test. Expected future returns are the second one, the “reward” for patience. It is simply human nature to constantly assess the environment that is supposed to deliver the next “marshmallow” and respond accordingly to signals that might change the probabilities of its eventual appearance.

The current global investment climate is, ironically, less predictable than when the world was in the throes of the pandemic crisis and its immediate aftermath. Monetary policy is likely to tighten at an aggressive rate, rather than remain very accommodative. Inflation pressures (nonexistent in 2020) remain a wild card, both with respect to future central bank policy and corporate profit margins. And, of course, the Russia-Ukraine war puts fresh uncertainty into both the future growth of the European economy and global oil/food prices.

#2: “Peak unpredictability” makes for sustainable stock market bottoms. The one question we ask ourselves every morning is whether US/global equities have already made their lows for the year. The market’s litany of woes, outlined in the prior point, are well known. Absent a new catalyst, it is tempting to believe their impact is already incorporated into asset prices.

We remain cautious, however, because the range of possible economic and corporate profit outcomes remains so wide. The market is not in a position like March 2009 or March 2020, where changes in fiscal and monetary policy created incremental certainty around economic outcomes. Moreover, corporate earnings are at all-time highs, unemployment is near all-time lows, and wage growth hasn’t been this hot in almost 40 years. The Fed may well be able to thread the needle and engineer a soft landing as it fights inflation, but we won’t know if they’ve accomplished that for at least a few quarters. Since daily market fluctuations are a function of anticipated economic conditions 6 months out, it is hard to argue that the final bottom is in. We’ll keep asking the question, though … It will come.

#3: Relatively high unpredictability makes for tradeable lows. We’ve spent a lot of time this year analyzing the CBOE VIX Index and, last night, added the NASDAQ 100 Volatility Index (VXN) to our toolbox. In our mental model of the market, these measure relative uncertainty:

• We know that the long run average of the VIX and VXN are 20 and 25, respectively.

• We also know that the market faces above-average uncertainty right now, so the VIX and VXN should track at or more often above their long run means.

• When the VIX and VXN get above 1 standard deviation from the mean (28 and 37, respectively), markets are signaling unusually high uncertainty. Once the VIX/VXN get to 36 and 49 respectively (2 standard deviations), you know markets are truly fearful. Those are spots where traders may want to take a shot long in anticipation of a near term bounce.

Takeaway: while US equities always eventually deliver the second marshmallow, stock prices over the shorter term are a function of investor confidence in the current environment to deliver it.