MARK RIEPE: During pre-virus times, when people were still at the office, many would hit the vending machines at around 3 p.m. every day.
Others might order French fries from a drive-through or just consume an entire pint of ice cream right in the front of the freezer at midnight.
Carb-heavy and sugar-intensive comfort food is built into the daily routines of many Americans, the assumption being that eating it will make you feel better. But it turns out that this assumption may not be true: Scientific studies now show there is no correlation between sugar consumption and mood.1 More specifically, a meta-study of 31 other studies investigated the effects of sugar on anger, alertness, depression, and fatigue. The aggregated results indicated that there was no positive effect, and this result held even as the volume of sugar consumed increased.
If anything, sugar tended to make people more tired and less alert in the hour after consumption.
Now, I’m not here to harp about nutrition, because I am certainly no role model when it comes to that.
The point of this example is that we run into trouble when we make decisions based on an assumed relationship that doesn’t actually exist. This is known as “illusory correlation,” and it isn’t a new phenomenon.
Will Rogers was an American humorist, actor, and storyteller who was popular in the 1930s.
He was all over this problem and said, “What gets us into trouble is not what we don’t know. It’s what we know for sure that just ain’t so.”
This idea that our strongly held beliefs can get us into trouble is our subject today. We’re going to discuss several notions that investors have that aren’t supported by data.
I’m Mark Riepe, and this is Financial Decoder, an original podcast from Charles Schwab. It’s a show about financial decision making and the cognitive and emotional biases that can cloud our judgment.
Every conscious decision we make is a product of our underlying assumptions about how the world works and how we interpret the information we have available to us. Many of the biases we talk about on this show deal with misinterpreting information or systematically failing to obtain a complete set of information.
But there’s another problem we face when making decisions: What if our information or our assumptions are just plain wrong? We can be the poster children for rational decision-making, free of cognitive and emotional decision-making biases. But none of that helps if we’re basing our decisions on incorrect data or assumptions.
This leads me to a conversation Liz Ann Sonders and I recently had about the correlation between the stock market and the economy. Liz Ann is a senior vice president here at Schwab, and she’s our chief investment strategist.
In addition to that, last year she was named as one of the “100 Most Influential Women in Finance” by Barron’s magazine.
We started talking about the stock market and the economy but ended up covering not only that topic but many other relationships that matter to investors—but are more complex than you might think.
MARK: So Liz Ann, the premise of this episode is that we’re talking about the concept of illusory correlations. By that, I mean people think a correlation between two variables exists, and that correlation, in reality, either isn’t there at all, or it’s much weaker than is commonly believed. And I think a great example of this right now is the correlation or the lack thereof between the performance of the economy and the performance of the stock market. So why do you think people believe that there really should be a nice tight correlation between those two things?
LIZ ANN SONDERS: So I guess I’d answer by saying there is somewhat, obviously, a relationship between what’s going on in the economy and what’s going on in the stock market. I think what often trips investors up, and it’s certainly in this environment, is maybe the time differential. So the stock market is a leading economic indicator. It tends to move in advance of turns in the economy. It sort of is an anticipatory mechanism for what’s happening in the economy. In fact, if you actually do a correlation analysis, and we did one all the way back to the inception of the S&P 500®, which was in the late 1920s, and if you just do a simple correlation between annual changes in real GDP, which is inflation-adjusted GDP, and annual real total returns for the S&P, depending on what inflation metric you use, it’s somewhere between 0.09 and 0.11. Now, remember, zero means there’s no correlation. You would have to get all the way up to 1.0 to have a perfect correlation, so that’s extremely low. Interestingly, though, if you lag the economy by a year or so, the correlation actually jumps up to about close to 0.4, still not a perfect correlation. So I think it’s the understanding that the market tends to move in advance of changes in the economy and that’s why at times like this, you see the stock market booming and wondering how that can be the case in advance of similarly strong economic data.
MARK: So a lot of people think that in a given 12-month period, the performance of the stock market and the performance of the economy in that year is going to be tightly correlated. And what you’re saying is the data shows that the stock market is tracking next year’s economic performance much higher?
LIZ ANN: Correct. More closely. It’s still not a perfect correlation because there’s so many other factors that drive what the market is doing, not least being things like monetary policy, what the Fed is doing, earnings growth rates, which aren’t necessarily correlated to what’s going on in the economy, sentiment conditions. So there are so many other factors aside from just the economy that affect the stock market, but it’s also that timing difference that’s important.
MARK: And I’m glad you mentioned those other factors because I think this issue of the economy and the market, it seems to be on the mind of investors much more now because that disconnect between the two seems much greater than it used to be. And my sense is that all those variables you just mentioned are much more in flux now than maybe they’ve been over the past several years.
LIZ ANN: They’re not only in flux, but I think what makes the current environment so unique is we’ve never seen a full stop in the economy by virtue of a government-mandated shutdown. That had the effect, of course, of compressing the economy and all the data associated with the economy to a level unlike anything we have seen, save for maybe the Great Depression.
Now, the fact that the market rallied in advance of the pickup off those extreme lows is in keeping with what has happened historically. But what I think is still hard to get your hands around is just how depressed the level of the economic data has been and how long a path it is going to be, regardless of how big the sort of month-over-month percentage increases in the data are to get back to anything resembling pre-COVID normalcy. So I think it’s the depth of the weakness that is unique relative to past cycles.
MARK: Obviously, these are things that professional economists and professional investors think about a lot. There are, in fact, indexes created, the leading economic indicators index, the lagging economic indicators index. Maybe talk a little bit about what constitutes leading, what constitutes lagging, and if you could focus a little bit on the unemployment rate, because I think that’s the one economic statistic that people seem to pay an awful lot of attention to.
LIZ ANN: Absolutely. I think it’s probably one of the more popular economic statistics and probably one, if you did, you know, man-on-the-street type interviews and you asked just anybody walking by if they understood the unemployment rate and about what it was, it’s probably one of those statistics that more often than not people have a general sense of. The problem is, it’s one of those lagging economic indicators. So as the labels would suggest, there’s sort of a subset of leading economic indicators. Those are indicators that tend to move first in advance of overall changes in the economy. Then there’s coincident economic indicators that tend to move in line with things like GDP, and then, of course, the lagging indicator.
So before I get to the unemployment rate, we can think of labor market data broadly as also just having leading coincident and lagging components. So the most leading indicator of all those labor market indicators is initial unemployment claims. That’s the number we get every Thursday at 8:30 Eastern Time in the morning—it’s how many people have initially filed for unemployment insurance. That tends to lead broader changes in the economy. The payroll number that we get as one of the headline numbers in the jobs report every fourth Friday, that is a coincident indicator and then the unemployment rate is one of those lagging indicators. Because of its popularity, and it’s one of those indicators that really sort of gets to the heart of what’s going on in the economy, we feel it as consumers and as individuals, but the reality is it’s a lagging indicator. And when you’re comparing a leading indicator like the stock market to a lagging indicator like the unemployment rate, it’s why the data I’m about to share has held true in the post–World War II era.
So if you break the unemployment rates since 1948 into quartiles, so the lowest unemployment rate quartile, which is when it’s under 4½ percent, all the way to the highest quartile, where it’s more than 6.8%, the best performance for the stock market, up 19%, has come when the unemployment rate is in its highest quartile. If you leave quartiles aside and just look at periods when the unemployment rate was above 8%, which clearly we are now, the annual return for the S&P historically was 25% because the stock market moves first in anticipation of the eventual turn down in the unemployment rate. But there’s no question it trips investors up.
MARK: Yeah, so to kind of summarize that, maybe grossly simplify it, if you want to know where the labor market has been, look at the unemployment rate. If you want to know where it’s going, look at the initial unemployment claims. Does that sound right?
LIZ ANN: Absolutely. And if you want to know where the labor market is, look at payrolls.
MARK: That’s right. And then when it comes to at least the relationship between the stock market and the labor market, based on that data you were just saying, it literally is the case where it’s often darkest before the dawn, and investing when a lot of the news is bad, historically, on average, has paid off.
LIZ ANN: Absolutely. And it also helps to highlight something that I think is misperceived by investors. People think, “Well, how could a recession start with the unemployment rate low?” or, conversely, “How could a recession be over with the unemployment rate higher, still rising?” The unemployment rate, the moves in the unemployment rate, don’t cause recessions. Recessions happen, and then they cause an increase in the unemployment rate, and then the opposite occurs when you move out of recovery. So there’s also that sort of order of things that comes into play with just economic data, leaving the stock market aside.
MARK: Another example of this disconnect is the relationship between the money supply and inflation. When I was in school taking econ and investing classes, and you probably were taught this, as well, that the view was when the money supply goes way up, then inflation is sure to follow. Why was that the prevailing wisdom? What’s the sort of kernel of truth behind that?
LIZ ANN: So there is a kernel of truth looking over the long history because, traditionally, rapid money supply growth tended to correspond to booming economic growth environments, which also tended to correspond to rising inflation periods. The difference, not just in this environment, but really the last cycle, as well, coming out of the great financial crisis, is that we saw the Fed step in with massive amounts of quantitative easing following the ’08 financial crisis, and of course they’ve done the same thing in this environment, so they’ve been pumping massive liquidity into the financial system through the ... because of these last two crises that we’ve had. And then especially recently, you add what’s been done on the fiscal side with what Congress has done with stimulus checks being given directly to individuals, as well as enhanced unemployment insurance, that’s a massive increase, record-breaking increase in the money supply. But the only way that increased money supply becomes inflationary is if it leaves the hands of consumers, leaves the financial system through the lending channels, through the borrowing channels, and picks up what’s called velocity. It basically is money that’s put to work in the economy. That’s when you develop an inflation problem. But when you’ve got businesses and/or consumers sort of hoarding a lot of that liquidity that they’ve received, you don’t get that velocity, and, in turn, you don’t get inflation. High inflation will likely only materialize if this so-called high-powered money created by the Fed, brought into the picture by Congress, generates a sustained acceleration in economic activity that, ultimately, results in much higher demand relative to productive capacity. That hasn’t happened now, and it didn’t happen in the last cycle either.
MARK: Yeah, if the money is just sitting in the bank’s vault, so to speak, that’s not going to be generating inflation. It’s got to be, as you said, it’s got to be out in the economy being used for, you know, productive purposes.
LIZ ANN: Right, and we have seen a massive surge in personal income. In fact, what’s interesting is everything that’s been done by the Fed and Congress has actually added more in the pockets of many individuals than they would have otherwise gotten. But the savings rate at its recent high went up to 32%. It’s settled back to 23%, but although there’s been some spending of that additional money that individuals have gotten, a lot of it has been put into savings, so it’s not working its way into the economy.
MARK: So you mentioned the great financial crisis of 2008 and 2009 as kind of the last time the Fed really put a lot of money into the financial system. And I think that’s when this relationship between money supply and inflation started to become very apparent that it was breaking down a little bit. Why do you think that happened?
LIZ ANN: Well, you know, in the financial crisis era, I think it happened for two reasons. So the Fed did, basically, reliquefy the financial system. Remember, the crisis was within the financial system. It was a collapse in the global financial system. So what the Fed did was directly sort of easing the strains in the financial system. But the problem was, even though they pumped massive amounts of liquidity into the financial system, the financial system was in the process of deleveraging, largely because of massive leverage being at the heart of the crisis, not to mention regulatory changes that required banks to just hold more capital. So it was sort of forced deleveraging on the part of the financial system. And then, of course, households, because they were burned by the bursting of the housing debt bubble, also went through a period of deleveraging. So all this money sloshing around in the system, but an unwillingness on the part of banks to lend because they were deleveraging, and an unwillingness or lack of desire on the part of households to borrow because they were deleveraging. So basically, the Fed filled up the trough, but if no one is coming to the trough to drink, and the owner of the trough has put a lid on the trough, that liquidity doesn’t get out into the economy.
MARK: So the million-dollar question would seem to me is, do you think this is a permanent change, or is there going to be something about what we’re going through right now which will trigger that bout of inflation that some people are worried about?
LIZ ANN: So I do think that there is a risk that the trajectory for inflation looking, you know, years ahead could be a little sharper on the upside than what is currently built into expectations. I think the consensus has got to the point where many, many people view inflation as dead and buried forever. And I’m always more intrigued by the story few are telling than I am the story everyone is telling. And I do think that there is a risk, not of hyperinflation 1970s-style, but a bit of a pickup in inflation, for a variety of reasons. One is tied to massive budget deficits and the cumulative effect of running budget deficits, which is rising debt levels. But maybe also separate from that is this whole notion of deglobalization. As we move to sort of bring production back closer to home, if you’re a believer as I am that globalization over the last 30 years has been one of the reasons why inflation has come down so much, it’s hard to then argue that deglobalization won’t cause a little bit of a pickup in inflation. So I think near-term, that’s more of the risk than some of these other factors like the increase in money supply.
MARK: Let’s move on to talk a little bit about the relationship between the valuation of the stock market overall and short-term performance. It is a question I get—I know it’s a question you get—whenever people look at, say, the price-earnings ratio of the S&P 500, and they see that number getting into the 20s, the upper 20s, they start getting very concerned that the stock market is overvalued and it’s going to correct over the next six months. Talk to me a little bit about, first of all, you know, how is valuation measured, and then, again, what does the data really show about the correlation between valuation and short-term performance?
LIZ ANN: So Mark, as you know, there are myriad ways to value either the market or an individual security. In fact, from a market valuation perspective, I keep a running table that’s got 12 to 15 valuation metrics of every variety, your standard P/E ratio, and even there, are you looking at trailing earnings, earnings that have already gone in the books, or forward earnings? If you’re looking at trailing, how far back are you going? You can look at equity risk premiums relative to either Treasuries or corporate.
But no matter which metric you use, and even if you just hone in on your standard P/E ratio, there is very little, if any, correlation between where valuation is at any point in time and near-term market performance, meaning, say, subsequent one-year performance, there’s just no relationship. And it makes me think of that famous Benjamin Graham line, that in the short run, the market is a voting machine, but in the long run, it is a weighing machine. And I think there are so many other factors that drive short-term market performance, valuation just doesn’t happen to be one of them, even if it is a much better guide for longer-term performance.
MARK: Yeah, and so why do you think that is? And I love that Ben Graham quote. Why do you think valuation starts to matter more and more as you go out two years, three years, 10 years? What’s driving that?
LIZ ANN: I’ll get to why valuations matter in the long term. But in the short-term, prices, whether it’s of a stock or an overall index, like the S&P, are typically set via, you know, reflexive forces, whether it’s the information we’re getting, headlines on a day-to-day basis, herding in momentum-chasing, feedback loops, you know, animal spirits kind of stuff. And that actually ties to valuation. We think of valuation as this quantifiable metric, because we always know what the price is of an index or the stock. Even on a forward earnings basis, there’s a number, typically, out there. There’s an estimate for a company. There’s consensus estimates for, say, the S&P 500. So you think, OK, we’ve got these two quantifiable components of the equation, it should mean something. But the reality is that valuation is as much a sentiment indicator as it is a fundamental indicator. There are times like circa 2000, where investors were willing to pay, you know, nosebleed valuations for stocks that really had no prospects for earnings, and other times, like early ’09, that investors didn’t want to pay anything. So I think it’s that sentiment factor in the short-term.
You know, ultimately, high valuations set up a condition sometimes called criticality, and this ties into how it’s a sentiment indicator. As they get more stretched, even relatively small shocks to the system can trigger a change of state. So I still very much believe in their applicability to long-term market returns, but in the short term, it’s really a function of market sentiment.
MARK: Let’s talk a little bit about earnings, because earnings showed up quite a bit in some of the different valuation metrics that you were listing earlier. What’s the relationship between, say, an earnings growth rate and subsequent market performance?
LIZ ANN: Yeah, this always surprises investors when I share these statistics, because even if somebody understands what we touched on in the beginning of your show about the relationship between the stock market and the economy, people do very directly tie earnings to stock prices. But, again, you’ve got the leading nature of the stock market. And, of course, when earnings are released for a company collectively, they’re rearview-mirror earnings, and the market tends to anticipate that, which is why—and this is similar analysis to what I showed with the unemployment rate—if you break earnings growth into various sort of ranges, now the worst performance for the stock market has come when earnings growth is worse than minus 20%, so sort of, you know, recession-type Armageddon scenario. The average return for the market or the average annual return for the market was negative almost 14%. But interestingly, the best zone for the stock market in terms of earnings growth is when earnings were somewhere between negative 20% and plus 5%. And that tends to reflect that the stock market usually has a tough time when earnings are imploding, but at that turn when they start getting better, they start moving into less bad territory, that’s usually the launch point for the stock market.
By the way, the worst performance for the stock market, at only a little more than 2% annualized return, is when earnings growth is more than plus 20%. So again, it’s understanding the relationship between economic and earnings data and the leading nature of the stock market.
MARK: Liz Ann, a lot of interesting things in that answer, but one of those you mentioned was stocks are pricing based on the expected earnings. And another question we get periodically is company X will report earnings, they lost, say, a billion dollars during a quarter, and its price goes up because, frankly, the market was expecting it only to lose maybe $2 billion or $3 billion during the quarter. So could you talk a little bit about how investors sort of collectively should be thinking about those expectations, and how they’re formed, and what happens when those expectations are either exceeded or disappointed?
LIZ ANN: Yeah, so I think this is also a very important concept about where the expectations bar is, whether it’s for economic data, or an earnings release by a company or earnings more broadly, say, for a sector or for an index overall is the market in the short term tends to care more about better or worse than it does about good or bad. And I think that’s an important concept to think about. Better or worse tends to matter than good or bad. It’s human nature to think of data, whether it’s earnings or economic data, in more level-based terms, good versus bad, strong versus weak. But the market tends to key off better or worse inflection points, and that comes into play with earnings, as well. Where the bar has been set and whether the company is exceeding that bar or underperforming that bar, in the short term, can drive a lot of the market performance.
MARK: Thanks, Liz Ann. I’ve got one more question and then we’ll wrap up here. When we think about these correlations or lack thereof, it’s really an assumption that investors are making that a relationship exists, and that relationship isn’t there or it’s not as strong as they thought. But people make assumptions all the time that influence their investing behavior. And one statement I get is people say, particularly during periods of uneconomic uncertainty, they will say, “Well, I’m just going to stop investing. I’m going to turn into a saver, and I’ll become an investor again once everything has settled down.” I think implicit in that kind of statement are a lot of strong assumptions. What do you think?
LIZ ANN: Oh, definitely, a lot of strong assumptions. And often, when that question is asked or the comment is made, the terminology used is, “Why wouldn’t I just, you know, get out for a while and then just get back in when things settle down?” And the problem with that is how does one make that judgement? We know that market timing is really an impossible task. And when you’re talking about making all-or-nothing decisions, you’ve got to be right at both times in order to sort of nail it. And what’s going to drive that decision? Is it just sort of maximum emotional pain?
I think we sometimes have to differentiate between financial risk tolerance and emotional risk tolerance. But it requires that really accurate timing in both directions. And the reality is, to be a successful long-term investor, you don’t have to pinpoint tops and bottoms with any kind of precision. Investing should always be a process over time, a disciplined, stay-in-gear process over time using things like diversification and rebalancing. It should never really be about any moment in time, because that is what trips investors up, especially if they’re dealing more with the emotional side of investing and not the sort of intellectual side of investing. So again, you don’t have to pick tops and bottoms, I think, to be a successful long-term investor.
MARK: Liz Ann, a lot of timeless wisdom, as always. Thanks for stopping by.
LIZ ANN: My pleasure. Thanks for having me.
MARK: At the beginning of this episode I mentioned that Will Rogers once said, “What gets us into trouble is not what we don’t know. It’s what we know for sure that just ain’t so.”
You may have heard this quote before because it opens the 2015 film The Big Short.
I always planned to use that quote in the introduction, but just a few days before we were going to actually record this episode, I decided to fact check that quote to make sure I got the wording right. I’m glad I did. It turns out that my assumption that Will Rogers was the originator is disputed. Other sources attribute it to Mark Twain, but still others provide evidence that he didn’t originate it either and cite humorist Josh Billings or comedian Artemus Ward2 as the true origin.
So one lesson from this episode is to make sure you fact check your podcast.
More importantly, how do you avoid the trap of the illusory correlation or making unfounded assumptions? Here are some suggestions.
First, recognize that the human brain tends to find causation, even when it doesn’t exist. This is often because we don’t cast a wide enough net and consider all instances of a phenomenon. This is why the data Liz Ann discussed looking at correlations went back to the inception of the S&P 500.
Including a larger data set often reveals that a correlation in one period doesn’t show up in other periods which casts doubt as to whether the correlation is truly universal or just a fluke.
Second, don’t be fooled by averages. Going back to 1926 the average annual return to the S&P 500 is about 10%. However, the return in any given year is almost never 10%. Instead, there’s tremendous variability year-to-year. The same applies to many supposed cause-and-affect relationships when it comes to investing. On average many relationships appear to exist, but if you look closely at all of the individual instances when they occurred, you find that some might work only 60% of the time, and others far less than that.
Third, look for the alternate explanation. If you think event A causes event B, take a moment and think about what other event could also cause event B. One example of this is the inversion of the yield curve in October 2019. At the time, many analysts pointed out that recessions often follow several months after the yield curve inverts. But this recession wasn’t caused by the yield curve inverting. The recession we’re in as I’m recording this was caused by efforts to contain the COVID-19 virus. The yield curve inversion had nothing to do with it. It’s another example of a correlation that isn’t as strong as people think—and in this one instance—is an illusion.
If you want to know more about what the markets are doing and why, check out schwab.com/volatility. And you can get all kinds of market insight if you follow Liz Ann on Twitter: @LizAnnSonders. L-I-Z-A-N-N-S-O-N-D-E-R-S.
You can follow me on Twitter: @MarkRiepe. M-A-R-K-R-I-E-P-E.
That’s it for today. Thanks for listening to the final episode of Season 5. Season 6 will start up later this year, and we just might throw in a bonus episode between now and then.
If this is the first episode that you’ve listened to, there are 30 more available for downloading.
If you’ve enjoyed any of the episodes, please consider leaving us a rating or review on Apple Podcasts or your favorite listening app. Those reviews really do help others discover the show.
For important disclosures, see the show notes and schwab.com/financialdecoder.
1 Konstantinos Mantantzis, Friederike Schlaghecken, Sandra I. Sunram-Lee, and Elizabeth Maylor, “Sugar Rush or Sugar Crash? A Meta-Analysis of Carbohydrate Effects on Mood,” Neuroscience & Biobehavioral Reviews, June 2019.
2 https://quoteinvestigator.com/2018/11/18/know-trouble/
