You may have heard someone claim that the stock market has detached from the “Real Economy.” Generally this is stated in a disdainful way and is meant to indicate that, sadly, the stock market is no longer a reliable indicator of economic health. I find this to be an ignorant and misleading claim for three reasons. First, the stock market has not historically been an indicator of economic health. Second, I take issue with the expression “real economy.” Third, I don’t see why stock market valuations should be a straightforward indicator of economic health.
Reason #1: A Healthy Stock Market Has Never Indicated a Healthy Economy
A healthy stock market, from the perspective of an investor in stocks, is a stock market which can deliver high returns. Rate of return can be basically calculated through price growth in a stock and the dividend payout rate (more adjustments need to be made to deal with stock-splits). The idea that the health of a stock market should relate to the health of the economy is based on the notion that corporate profits should drive returns on the stock market. In theory, a healthy economy means good corporate profits, which should correspond to good returns on the stock market.
The chart below based on data from Robert Shiller disproves the idea that stock market returns were ever tied to corporate profits in the medium term over at least the past 100 years. His data gives a historical record of the dividends, corporate earnings and price of a hypothetical S and P 500 index dating back to January 1871. It also provides a monthly CPI index starting in 1871. Using monthly data I have constructed annualized inflation-adjusted rates of return for stocks. I have additionally constructed a data series for the amount of stock returns which can be explainable in terms of corporate profits. If the annualized profit level of the S&P 500 is 5% of its value, then it would rationally explain a 5% stock-return on the S&P 500. However, the chart below shows that stock returns are not explainable in terms of the profitability of their companies. Instead, as the chart below shows, the profit level of firms relative to their price stays reasonably stable but stock market returns fluctuate wildly around them. The chart used 5-year averages of monthly data.
The Pearson R value between real returns and real returns explainable by earnings in 5-year averages is -.01, or -.12 if we exclude data from the 21st Century. I exclude 21st Century data because one may argue the weak correlation is attributable to recent events. In fact, over 5-year intervals the overall performance of stocks was unbounded from the overall performance of publicly traded companies even prior to the 21st Century. Looking at monthly data, the Pearson R for returns and justifiable returns is -.01 overall, and -.04 excluding data from this century. Overall this data shows no strong statistical relationship between justifiable returns and actual returns.
Hypothetically, if we let the economy and stock market continue largely unchanged for a few centuries, we may see some long-term connection between corporate profit level and stock market returns. But, from the data we can observe, we have to conclude that if returns on the stock market are very good for about a decade even though companies aren’t very profitable, this is nothing new.
Given that stock returns don’t strongly correspond to corporate profitability, there is even less reason they should necessarily strongly correspond to other economic indicators. The financial health of corporations should directly affect the health of stocks in a rational market which puts a price on future income. However, even a rational stock market has no reason to reflect the well-being of ordinary people, except that the economic health of ordinary people might relate to the economic health of corporations. Why should a largely irrational stock market represent the economic health of normal people?
While there is no intuition to explain the idea that the stock market ever would indicate the economic well being of the “little guy,” let’s also demonstrate it with data. The chart below, which also utilizes 5-year averages, plots the unemployment rate (left) against the annual log-returns on the stock market (right). If stock market returns ever did correspond to economic health, there should be a noticeable negative relationship between unemployment, which means a bad economy, and stock returns, which mean a good stock market return for stock investors. No relationship can be observed to exist. So, since the early 1950s (when the chart starts), there is no evidence to suggest that high unemployment should correspond to bad returns on the stock market.
The Pearson R value for 5-year averages of unemployment and stock returns is -.17 overall and -.23 excluding data from this century. Over 1-month intervals the Pearson R is .12 overall and .10 excluding 21st Century data. Not only are statistical relationships between unemployment rates and stock returns weak, but the statistical relationship between changes in unemployment and stock returns are weak. Over a 1-month basis, the Pearson R between the change in unemployment and the real return on stocks is .05 overall, and .07 excluding data from the 21st Century.
Reason #2: “Real Economy” is a Bad Term
People like to use the term “Real Economy” to distinguish financial markets as being “fake.” I would challenge anyone who uses such a term to explain what authenticity means in the context of a market and why it matters. I think the “realness” of the “real economy” versus the stock market is meant to connote “real” economic health as opposed to “fake” economic health. In this framing, the economy is “really” healthy if unemployment is low, for instance, but if the stock market is making impressive returns, then this is “fake” health. However, this way of talking is convoluted. It would be much simpler to say that the health of the economy has nothing to do with the level of profits in the stock market, or other financial markets.
Obviously, there is nothing particularly “fake” about financial markets. One really does exchange real money in order to purchase a real thing in financial markets. The difference between the stock market and the market for agricultural goods (for example), is not one of realness and fakeness.
Furthermore, this terminology overlooks how financial markets can affect non-financial markets and alter employment and wages. Some commodity producing firms, for instance, engage in futures trading for the goods they produce. They bet against the future value of the thing they will produce to minimize risk. Higher futures prices incentivize those commodity-producing firms to expand production. To them, a futures contract is a way for them to get income right away for goods they put money into producing in the future. So clearly, financial parts of the economy can impact nonfinancial parts. When we normally distinguish “real” and “fake” things, we understand, or at least imply, that fake things cannot affect the real things.
The term “real economy” seems to be popular among people who want to express the idea that the stock market is unimportant (or no longer important). While I agree that strong stock market returns should not be a public policy goal, I also don’t think cheap tractor prices should be a goal on its own. Should I argue that both the stock market and the market for tractors are outside of the “real economy?”
Reason #3: Stocks Don’t Have to Indicate the Economic Health
I am sympathetic to those who would prefer the stock market does badly owing to a view towards reducing inequality. However, this in of itself doesn’t mean that we should want stock market returns to reflect something like unemployment. Shouldn’t we want the stock market to be doing terribly even if the economy does well? After all, don’t high returns on the stock market contribute to inequality regardless of the state of the economy?
Why do we need an indicator, like the stock market, to tell us about overall economic health? In what way could one ever use an aggregate measure of goodness to productive ends? The Federal Reserve, for instance, uses unemployment not just as a measure of overall goodness to tell them how good things are doing, but as an indication of the relative strength of demand, labor market tightness and inflationary pressure. I am perfectly happy with looking at the S&P 500 as a measure of equity valuations. On its own, higher equity valuations should mean higher spending on investment and consumption and could be useful in a multivariable econometric model. However, looking at something like real GDP or the S&P 500 and asking: “Is this a useful measure of ‘goodness’ in the ‘real economy'?” Is a fruitless task.