What an inverted yield curve is and why it matters

What the yield curve is

The US Treasury issues debt at different maturities: three months, two years, ten years, thirty years. Each maturity pays a different yield, and if you plot all those yields on a chart you get the yield curve.

The normal shape slopes upward: the longer you lend your money, the more you are compensated for the risk that inflation erodes it or that you need it back sooner. A ten-year bond should pay more than a two-year one, just as a long-term deposit pays more than an instant-access account.

What inversion means

When the two-year bond pays more than the ten-year, the curve has inverted. Translated: the market is saying it expects interest rates to be lower a few years from now than they are today. And rates fall, basically, when the central bank needs to stimulate an economy that is slowing down.

Inversion does not cause the recession, it anticipates it. It is the aggregate price at which thousands of investors are willing to lend money to the US Treasury at each maturity, and that price embeds their expectations for growth and inflation. When those expectations deteriorate enough, the curve flips.

There is also a direct channel into the real economy. Banks make money borrowing short and lending long. With the curve inverted that business stops being profitable, so they tighten lending. Less credit means less investment and less consumption, which is precisely the mechanism by which a recession arrives.

The two spreads people watch

The most quoted is the 10-year minus 2-year. It is the one that shows up in the press and the one with the longest track record. When it drops below zero, the curve is inverted in that segment.

The second is the 10-year minus 3-month, and several Federal Reserve studies consider it the better predictor. New York Fed economists build their recession probability model on it. The practical difference is that the three-month leg tracks the policy rate closely, so this spread captures the clash between current monetary policy and market expectations more sharply.

When both invert at once, the signal is more consistent. Semavor tracks both and colours severity by how far they stray from their usual range, not by the sign of the day’s move.

The lag: why it cannot date anything

This is where the signal turns slippery. Since 1955 the curve has preceded every US recession, but the lag between inversion and the start of the recession has ranged from six to twenty-four months. A signal that can be two years early is useless for timing anything.

And it has produced at least one false positive: it inverted in mid-1998 with no recession following. The 2022-2024 inversion, the longest in modern history, was not followed by a recession within the usual window either, which has reopened the debate over whether the indicator has lost power after years of massive central bank bond buying.

A nuance that rarely gets explained: historically the recession has not arrived while the curve is inverted, but shortly after it normalises again. The steepening — when the long end recovers its premium over the short end, usually because the central bank starts cutting rates fast — has been the immediate prelude to several recessions.