The Treasury Yield Curve Inverted Again - Should We Be Worried?

The yield curve is typically represented in a graph comparing Treasury yields from 3-month to 30-year using data from the U.S. Federal Reserve. The Fed Funds rate (interest rate at which depository institutions trade federal funds with each other overnight) is not typically included although as we will see below, it is worth examining.

Typically, analysts compare the difference between 10-year and 3-month yields to prognosticate the health of the U.S. economy. A steep yield curve, when short-term rates are well below long-term, generally implies a growing economy. It shows that borrowers are paying less for long-term instruments such as 10-year Treasuries and instead seek higher returns in riskier investments (like stocks).

FIGURE 1: DAILY CHART FROM THE FEDERAL RESERVE SHOWING THE 10-YEAR TREASURY MINUS THE 3-MONTH TREASURY YIELDS. First signal occurred in May 1989 and recession (in grey) was declared July 1990 (10 month lead), next signal was September 2000 and recession confirmed March 2001 (6 month lead), next signal September 2006 and recession began December 2007 (16 month lead).

Strident yield curve followers would have witnessed that the daily 10-year Treasury yield (T10YR) first dropped below the 3-month Treasury yield (T3MO) rate on March 22, 2019 and stayed there for five days. Thirty one trading days later on May 13, the rate again dipped into negative territory and for all but six days (up to July 5 when the data were analyzed) stayed negative.

What is significant is that the T10YR rate was below that of the T3MO for the second quarter of 2019, a signal that Professor Campbell Harvey of Duke University's Fuqua School of Business says has been accurate 100% of the time in presaging recessions. As part of his 1986 dissertation, Harvey showed that from 1960, a negative quarterly T10YR-T3MO reading presaged every recession. Since then, it has accurately predicted three more recessions retaining its perfect batting average. The most recent signal would be number four if confirmed.

So how is a recession defined? Many have said that it occurs following two quarters of negative growth in the Gross Domestic Product (GDP). However, the NBER does not adhere to this characterization and instead defines a recession as "a significant decline in economic activity spread across the economy, lasting more than a few months, normally visible in real GDP, real income, employment, industrial production, and wholesale-retail sales."

A depression is different from a recession in that it is marked by a prolonged period of economic stagnation and a significant decline in both incomes and employment. A more popular tongue-in-cheek explanation says that a recession occurs when someone else loses their job. It's a depression when you lose your job.

Unfortunately, no matter what recession definition you use, it is at best subjective. Using two quarters of negative GDP growth provides firmer parameters. However, relying on the NBER definition is anything but and so of little help for traders looking for concrete signals.

The good news is that knowing when a recession (or depression) has arrived is less important in timing trades because by that point, your portfolio has already been hit. Wouldn't it be better to have advance warning and close the barn door rather than getting warning too late and watching your horse ride into the sunset?

BUILDING THE PERFECT PLOT
The Federal Reserve plots T10YY minus T3MO back to April 1982 on a daily basis. Unfortunately, it does not publish T3MO data prior to that date so users are unable to examine the longer relationship at fred.stlouisfed.org.

However, as we see in figure 2, the Fed Funds (FF) rate (in blue) is available back to 1955 and does a reasonable job of approximating the T3MO rate (in red) albeit at the cost of greater volatility and more false signals.

FIGURE 3: DAILY CHART OF 10-YEAR TREASURY YIELDS MINUS THE FED FUNDS RATE FROM 1962. As we see, the difference proceeding the 1982 recession was the biggest on record and occurred between the short 1980 recession and longer one in 1981-1982.

Here is one final indicator that can also be used to warn of a coming U.S. recession, albeit somewhat less reliable. The main advantage is the T10YR minus 1-year Treasury yields (T1YR) is the data dates back to 1962. As we see from Figure 4, it provided two 1970 recession warnings and a one-year advance warning of the 1974 recession. And although they were less pronounced than T10YR-T3MO signals, it also provided warnings of the next three recessions.

Since its not a recession but rather a bear market that is the greater risk to traders, how accurate were T10YR-T30MO signals in getting traders out of the market in time to miss the worst draw downs since post 1982?

FIGURE 5: CHART OF THE S&P500 INDEX (MONTHLY) SHOWING CONFIRMED T10YR-T3MO YIELD CURVE SIGNALS (ODD NUMBERS) AND CONFIRMED NBER RECESSION CALLS. Chart courtesy FreeStockCharts.com.

As we see in Figure 5, showing T10YR-T30MO confirmed signals by the red arrows (odd numbers) and confirmed recessions by purple arrows (even numbers) for the last three recessions, signals were about two years early to exit the 1991 and 2008 bear markets but nearly perfectly timed for a 2000-2001 bear market exit.

FEELING LUCKY?
Before relying on any trading signal, it is crucial to develop a degree of confidence. Since the early 1960s there have been seven U.S. recessions and seven signals and seven does not provide a high degree of statistical confidence. Yes, confirmed quarterly T10YR-T3MO signals have been 100% accurate but with the large variation in lag times between a signal, a bear market and recession, signals cannot be considered robust.

A better way to use these warnings is as wake up calls to tell traders that the economy is slowing, no matter how strong GDP growth or low unemployment figures may seem. Armed with this information, we can then tighten our stops, reduce risk tolerance and become razor-focused on money management; not a bad idea no matter what the markets are doing!

• St Louis Federal Reserve Treasury and Fed Funds Yield data
• Understanding The Yield Curve, Part 1, Giorgos E. Siligardos, PhD
• Understanding The Yield Curve, Part 2, Giorgos E. Siligardos, PhD
• It's Official: The Yield Curve is Triggered. Does a Recession Loom on the Horizon?
• The Yield Curve as a Predictor of U.S. Recessions
• NBER Discussion on Recessions