Equity Returns, Recessions, and the Yield Curve: The Most Insightful Thing Ever?
This might be one of the more insightful things I’ve ever realized in my life, so I will try to describe it succinctly and simply:
The yield curve is an important predictor of equity performance in the intermediate term.
I will show:
- The relationship between valuation and returns is extremely weak in the short term yet extremely strong in a 10 year time horizon.
- The spread on the 10 year to 2 year bond yield is a significant predictor of returns in the intermediate (3-5 years) timeframe.
- I propose the business cycle and the dynamic nature of recessions bridges the gap between short-run and long-run discrepancies.
- Current projected market returns based on valuations and the yield curve are poor.
Returns vs Valuation: The Shiller Effect
Robert Shiller won the 2013 Nobel Prize in economics for proving that valuation matters in long-run equity returns, so this concept is not new. He coined the concept of CAPE (cyclically adjusted price to earnings) to account for his measure of valuation, and showed that this metric was negatively correlated with future returns.
Rather than using CAPE, I use Warren Buffett’s alleged favorite indicator for market valuation, which is Total Market Cap / GDP. Throughout this article, I refer to it simply as the Buffett Indicator (BI). It illustrates the current price environment is rather pricey.
Over the last 30 years, a plot of 10 year forward equity returns with starting valuation clearly illustrates Shiller’s point. VALUATION MATTERS!
However, conducting the same approach on 1 year returns shows a very weak association between equity returns and beginning valuation, consistent with the work of Fama, Malkiel, and others. The proverbial ‘random walk’ suggests that markets are reasonably efficient in the short run and CANNOT BE PREDICTED.
The longer the investment horizon, the more the starting valuation impacts eventual returns. We cannot predict the moves day-to-day or even over the next couple of years of the stock market, but 90% of your future return in the decade-plus time frame is dictated by the price of earnings at entry point.
The Yield Curve Matters
The yield curve contains important, statistically relevant information to starting valuation in explaining intermediate term returns.
This can be seen graphically by plotting the percentage of the equity return explained by valuation and yield spread metrics over time:
This graph is amazingly important, so I want to be sure its meaning is captured.
#1 on the graph (the upward slope of the line) is the representation of Shiller’s Nobel Prize. Returns are pretty much impossible to predict in the short run, but over a 9-10 year period, valuation does quite a good job at explaining those returns. For 1 year returns, we predict about 10% of the market’s return with valuation whereas 85% of the returns are explained by valuation in the 10 year time frame.
#2 on the graph (the gap between red and black lines) is the realization from this article that in the intermediate time frame of 2-5 years, the yield curve spread is an important component of equity return. Between them, the yield spread and valuation explains over 50% of the 4 year annualized return. As the time frame lengthens to 8 or 9 years, the predictive power of the yield curve spread evaporates, and valuation is the key determinant. Full estimates of the regressions are at the bottom of the article.
Reconciling short-term vs long-term dynamics
Robert Shiller attributed the divide between short-run efficiency and long-run inefficiency to investor exuberance and sentiment. I believe there is certainly an element to that, but ultimately I believe it is the business cycle that collapses valuations, and my belief is the specific time of recessions cannot be forecast because they are themselves dynamic.
I think people become exuberant not because they are stupid, uneducated, or otherwise simply financially unsavvy, but because they think they have the secret sauce of figuring out when a recession will hit and that they will be able to exit equities before that time period. The difficulty of observing a recession in real time is discussed here for the 2000 recession.
To repeat, I don’t think people are stupid, but I think they think they are stupid enough to believe they will see a recession coming. However, recessions are stochastic events that CANNOT be seen coming.
- The economy is an extremely dynamic and interrelated system, balancing millions of different markets with their own lead and lag times and adapting at different rates.
- The economy is run and controlled by humans. Humans are subject to basic emotions of greed, fear, and a particularly nasty herd mentality.
- Capital markets are too large of a part of life for individuals, companies, and governments to ignore. CEOs and families alike act differently when they have just watched their worth get blown up or interest rates spike by 200 bps.
- Therefore, capital markets themselves are feeding into the decisions made that impact the real economy, while receiving and reacting themselves dynamically to the activity in the real economy. This can be seen in that interest rates clearly impact both equities and the real economy in a feedback loop.
- But what ultimately causes a recession? Why does one specific bank failure tilt the wagon over when in another time and place things might be just fine?
- The underlying decay of the economy (in debt loads, bad investment decisions, etc) reaches a point where it is extremely susceptible to a bank failure or a Fed rate hike. Perhaps it takes a mere 10 percent equity meltdown to set off a crisis. The herd starts to panic.
- Projects are delayed, workers are gradually furloughed. The effect snowballs and is only seen in the economic data ex-post, much as we see with the fall 2000 and early 2001 data.
- Therefore, the reason we observe a peak in the stock market 6 months before a recession is because the snowballing effect of equity sell-offs and loss of confidence is itself becoming a recession in a self-fulfilling prophecy. Economic data is only observable ex-post of the severity of the crisis.
Outlook: Its Not Great
I therefore view the yield curve as a wonderful indicator of how at risk the economy is. No one (not even the Fed) can be absolutely certain they will be able to spot a recession in real time, but the yield curve provides useful intelligence of the aggregate level of risk in the macroeconomy. I liken an inverted yield curve to counting cards and knowing the deck is stacked against you… it’s probably time to walk away from the table.
With that said, the projected ten year returns from this point are likely to be meager:
Importantly, the yield curve suggests intermediate term equity returns are also likely to be poor. While the model fit is much poorer than the 10Y, they are nevertheless suggestive of an economy which is late cycle and an equity market that is extremely expensive from a valuation perspective.
Please follow me at @fattailedhappy for more content like this.