Can recessions be predicted?

The answer to that depends on what you mean—by both what a recession is and what it means to predict one. Broadly speaking, a recession is when the economy is shrinking. I could be reasonably bold and say we are going to have a recession again sometime because they’re an enduring feature of the US economy, but that doesn’t do much more than remind people that there is a business cycle.

The questions that people tend to ask more often are, “Is a recession coming soon?” and “When exactly are we going to have one?” Those questions are tricky to answer because it’s hard to predict a recession definitively. The National Bureau of Economic Research officially designates the business cycles, and they decide we’ve experienced a recession only after it’s been going for a while. So, it is not easy to predict a recession before it happens.

Still, economists can offer a bit more insight into the probability of recession. We can more easily answer questions along the lines of, “Is a recession more likely in this economy?” Currently, the Federal Reserve is tightening monetary policy and reducing its balance sheet; historically, these activities have been followed by a recession. But it’s important to consider that these activities do not necessarily cause one another. Economists look at several factors when assessing whether a recession could be coming in the next year or two. The one I watch most closely is called the yield curve.

Here’s a brief explanation of the yield curve. The US government borrows money, and it does this by selling Treasury bonds. Some of these bonds are long-term and some are short-term, just like people can borrow using 30-year mortgages, 5-year car loans, and credit cards. People in the finance industry trade Treasury bonds, and the yield curve examines the yield, or return, on long-term Treasury debt versus the return on short-term Treasury debt. On average, the yield curve slopes upward, meaning longer-term bonds have higher interest rates than the shorter-term bonds, reflecting that investors expect to be paid more because they’re less certain of how bonds will perform over a longer period of time. But sometimes short-term rates end up higher than long-term rates, and that’s when we say the yield curve inverts. Yield curve inversion has been a fairly good predictor of having a recession in the following 12 to 18 months.

One example of when an inverted yield curve predicted a recession is in 2007. At that time, some analysts thought the idea of an imminent recession was silly because the economy was amazingly strong. But shortly thereafter, there was the great financial crisis and a very serious recession. There are times when the yield curve simply confirms what economists are seeing in the other ways they use to monitor the likelihood of recession, but this was one time when it definitely surprised people. Another example of an inverted yield curve predicting a recession was quite recent. The yield curve inverted in 2019 before the COVID-19 pandemic. Again, many said, “This is going to be a case of the yield curve missing what’s going on.” But then the virus appeared, and the economy shut down. Maybe this example happened because of coincidence, but the most recent recession was in 2020.

The yield curve is the indicator that I watch most closely, but I’m always aware that it is not the only indicator, and it is not infallible. We are not currently seeing an inversion (other people look at a different spread, which is closer to inversion). Our latest yield curve update placed the probability of recession in one year at 2.6 percent. Is that prediction right or wrong? We’ll have to wait to see. I know that people would often rather have a simple answer. But the yield curve’s real value is in predicting how likely a recession is coming and how certain policies affect that likelihood.