The Yield Curve Inverted (But Don't Fear a Recession)

The yield curve is back in the headlines after inverting last week. Yields on the three-month Treasury bill are higher than those of the 10-year Treasury note.

A bit of background for those of you not familiar with the yield curve. The yield curve plots the interest rates for Treasuries of different maturities. Under normal circumstances, the curve is upward-sloping. The upward slope reflects demands from investors who want higher rates of return for parting with their money over longer periods of time. If you expect inflation to be higher in the future than it is now, you want your longer-dated bonds [or certificates of deposit (CDs)] to pay higher interest rates rather than shorter-term ones.

Sometimes, the curve inverts. This occurs when shorter-term rates are higher than longer-term rates. This is what happened last Friday: The three-month Treasury bill closed with a yield of 2.46% whereas the 10-year note closed with a yield of 2.44%. As of yesterday’s close, the yield on the three-month bill was 2.44% versus 2.39% for the 10-year note. When I last wrote about the yield curve in early December (An Updated Look at the Yield Curve and Stock Market Volatility), the three-month note yielded 2.43% while the 10-year note yielded 2.91%. Last May, when I first discussed shorter-term and longer-term rates starting to converge (The Yield Curve Is Getting Flatter), the three-year bill yielded 1.92% and the 10-year note yielded 3.11%. As you can see, longer-term yields have fallen significantly, while shorter-term yields have stabilized after initially rising.

Many observers are watching the yield curve because historically it has preceded recessions. Before every recession since the mid-1960s, the yield curve has inverted. The time between inversion and recession has varied widely. Recessions started as soon as five months after inversion to as long as 16 months later according to data from Charles Schwab chief investment strategist Liz Ann Sonders. So, while the average is a little under a year, the actual timing of when a recession starts is often anything but average.

There is some debate about which Treasury maturities signify an inversion. Duke professor Campbell Harvey identified the link between inverted three-month/10-year rates and recessions in his 1986 University of Chicago dissertation. He also looked at the three-month bill and the five-year note. In both cases, Harvey’s model uses quarterly yield data to match the reporting cycle of GDP. A more temporary yield curve inversion—such as one lasting just a few days—does not signal a recession. (I spoke with him earlier today; an edited transcript of our conversation will appear in an upcoming issue of the AAII Journal.) Tufts University professor Michael Klein says every recession since the 1970s has been preceded by inverted two-year and 10-year rates. By Klein’s measure, the yield curve is not inverted since the two-year note’s yield of 2.22% is 17 basis points lower than that of the 10-year note. Adding to the discussion is a Federal Reserve paper published last year using expectations of easing monetary rates as measured by the forward spread (the difference between the current yield and the future implied yield). When the forward spread between 10-year notes and one- or two-year notes is negative, economic weakness follows. Sonders noted earlier this week that this curve is currently inverted.

All of these reflect the expectations of traders, who in turn price in observations about current and projected economic and market conditions. The timing of the next recession remains unknown. There will eventually be one simply because the economy is cyclical, but nobody knows with any certainty as to when.

Here’s what we do know. Expectations for a rate cut are growing. The CME’s FedWatch Tool now shows an approximate 70% chance of at least one quarter-point (0.25%) rate cut being announced before year’s end. The recently released National Association for Business Economics (NABE) survey shows expectations about the economy having “reached an inflection point, with the consensus forecasting real GDP growth to slow from 2.9% in 2018 to 2.4% in 2019, and to 2.0% in 2020.” NABE survey chair Gregory Daco added, “recession risks are still perceived to be low in the near term. Panelists put the odds of a recession starting in 2019 at around 20%, and the odds of a recession by the end of 2020 at 35%,” in part because of “the Federal Reserve’s dovish policy U-turn in January.”

A different indicator of recessions, housing starts, is not flashing a warning sign. Seven out of the last eight recessions were preceded or accompanied by a drop of at least 30% in housing starts according to Sam Stovall, chief investment strategist at CFRA Research. Though the housing market has weakened, it’s nowhere near recessionary levels. Housing starts totaled 1.162 million in February 2019 versus 1.290 million a year prior, a 9.9% drop.

There’s a general consensus that the pace of economic growth will slow this year. Slowing and contracting are two different things. Even if a recession does occur, its timing and severity are unknowns. More to the point, the S&P 500 index has averaged a 2.8% gain from inversion to recession, according to Sonders. (The range of returns is –14.6% to +16.5%.) Thus, fear about what might happen shouldn’t drive your portfolio decisions. While it’s helpful to stay informed, caution always has to be used to ensure short-term headlines aren’t driving your long-term portfolio decisions.