Few indicators carry the reputation of the yield curve. When short-term Treasury yields exceed long-term yields — an inversion — commentators reach for the word recession, and the historical record gives them reason to. Every U.S. recession since the 1960s has been preceded by an inversion of the spread between ten-year and three-month Treasury yields. The trouble is what people do with that fact.
Why inversion carries information
Long yields reflect the expected average of future short rates. If investors expect the central bank to be cutting rates two years from now, long yields fall below current short rates. Cuts are expected when growth is expected to weaken. Inversion therefore is not a cause of recession; it is the bond market pricing a slowdown that it believes will force policy easing. It is a survey of the most heavily capitalised forecasters in the world, expressed as a price.
Read moreWhat a 10-K Really Tells You About a Company
The New York Federal Reserve maintains a recession probability model built on this spread, and the Federal Reserve Bank of New York publishes both the current reading and the methodology. The historical hit rate is strong enough that dismissing the signal has been expensive.
The three problems
First, timing. The lag between inversion and the start of a recession has ranged from roughly six to twenty-four months. An investor who moved to cash on the day of inversion has historically missed a substantial portion of the late-cycle equity rally, and in several episodes the opportunity cost exceeded the drawdown avoided.
Second, the sample is small. There have been fewer than a dozen clean observations in the modern era, which is not enough to distinguish a robust relationship from an impressive coincidence with a plausible story attached. Statisticians would be reluctant to draw strong conclusions from eight or nine data points in any other context.
Read moreHow Credit Scores Are Actually Calculated
Third, the structure of the bond market has changed. Sustained central bank bond purchases, regulatory demand from pension funds and insurers, and global demand for dollar-denominated safe assets all compress the term premium — the extra yield investors demand for holding longer maturities. A compressed term premium makes inversion easier to achieve for reasons that have nothing to do with growth expectations. Official statistics on recession dating, incidentally, come not from the curve but from the National Bureau of Economic Research, and they are published well after the fact.
What to watch alongside it
Practitioners who take the curve seriously treat it as one input in a dashboard rather than a switch. Initial jobless claims, credit spreads, bank lending standards and new orders in manufacturing surveys all deteriorate before output does, and they update weekly or monthly rather than in a single dramatic crossing. Comparative work by the Organisation for Economic Co-operation and Development on composite leading indicators shows how much more reliable a basket is than any single series.
There is also a purely mechanical consequence of inversion worth understanding. When short rates exceed long rates, borrowing short to lend long — the core business model of a commercial bank — stops being profitable. Sustained inversion squeezes bank net interest margins and tightens credit availability independently of any forecast, which is one way the signal becomes partly self-fulfilling.
Read moreWhy State Tax Competition Reshapes Business
The honest summary is that an inverted curve is a credible warning about the direction of the economy over the following year or two, a poor instruction about what to own next quarter, and a reminder that the bond market is a forecasting institution whose views are worth reading even when they turn out to be wrong.
What the signal is and is not
The yield curve is not a forecasting model; it is a summary of what a very large number of participants currently expect about the path of short-term interest rates, plus a term premium that compensates lenders for uncertainty. When the curve inverts, the market is saying that it expects policy rates to be lower in two years than they are today, and the most common reason for rates to fall is a weakening economy. That is the whole of the mechanism. The curve has no independent knowledge; it aggregates expectations that can be wrong.
Daily curve data published by the U.S. Treasury shows why the historical record looks so impressive and why it is easy to over-read. The sample of postwar recessions is small, the lead time between inversion and downturn has ranged from several months to nearly two years, and the depth of inversion carries little information about the severity of what follows. Treated as one input among several it is genuinely useful. Treated as a trigger for portfolio changes it has misled investors repeatedly.
