Yield Curve

Ten-year money sometimes costs less than three-month money. When that happens in the U.S. Treasury market, the bond market is saying today's short rates are too high to last. That shape, the yield curve, matters because a simple spread inside it has forecast U.S. recessions well enough that the New York Fed still publishes a monthly recession-probability model from it. The curve is not prophecy; it is the price of waiting, inflation, and policy credibility drawn as one line.

How the line is built

A yield curve is a snapshot of borrowing costs across time. In a normal expansion, 3-month Treasury bills yield less than 2-year notes, 2-year less than 10-year notes, and 30-year bonds often sit highest of all. Longer lending usually demands extra compensation because inflation, policy, and opportunity cost have more time to surprise you.

The curve turns strange when that order flips. If the 3-month bill yields 5 percent and the 10-year note yields 4 percent, lenders are accepting less return for a much longer commitment. That only makes sense if they expect future short-term rates, inflation, or growth to fall enough that locking in 4 percent for a decade looks attractive.

There is no single canonical spread. Television prefers 10-year minus 2-year. The New York Fed's recession model uses 10-year minus 3-month. Campbell Harvey's early work leaned on 5-year minus 90-day bills. Those choices are not cosmetic. Each maturity pair mixes policy expectations and risk compensation in a slightly different way.

Why investors care

In 1996, Arturo Estrella and Frederic Mishkin found that the 10-year minus 3-month Treasury spread outperformed many other financial and macro indicators at predicting U.S. recessions two to six quarters ahead. That result stuck. The New York Fed still updates a term-spread recession model each month because the signal survived enough cycles to stay useful.

The mechanism has two layers. First, the curve reflects expectations: if investors think the Federal Reserve will have to cut rates next year, long yields can fall below short yields today. Second, the curve can help transmit the slowdown. Banks often fund short and lend long; when that margin compresses or turns negative, credit creation gets less attractive.

The sharp framing line is this: the curve does not see the future. It prices the odds that the present can survive.

What bends the signal

The clean story gets messy once central banks start buying duration on purpose. The Federal Reserve's first post-2008 large-scale asset purchase wave bought about $1.7 trillion of longer-term Treasuries, agency debt, and agency mortgage-backed securities between late 2008 and early 2010. That was designed to push down long-term yields. A flatter curve after such purchases does not mean exactly what a flatter curve meant in 1988.

Japan made the intervention explicit. In September 2016, the Bank of Japan adopted yield curve control and targeted the 10-year Japanese government bond yield at around 0 percent while keeping the short policy rate at -0.1 percent. At that point the curve was no longer just a market output; it was partly a policy instrument.

This is why economists split long yields into two pieces: expected future short rates and term premium. If long yields fall because investors expect weak growth, the curve is sending one message. If they fall because central banks, insurers, or pension funds need long bonds regardless of price, the message is harder to read.

What's contested

The first dispute is causal. Does an inverted curve help cause recessions by tightening bank incentives and credit conditions, or does it merely forecast the same slowdown that other actors are already sensing? Both stories can be true at once, but they imply different policy lessons.

The second dispute is about contamination. The more central banks guide the path of rates, the more the curve reflects official signaling as much as private belief. The New York Fed now publishes a separate Treasury term-premia series for exactly this reason: the curve mixes expectation and compensation, and the mix changes over time.

The third dispute is about confidence, not direction. On April 15, 2025, a simple New York Fed term-spread model put the probability of recession within a year at about 51 percent, with a 68 percent confidence interval of 39 to 64 percent. A modified version of the model, adding the 3-month yield's deviation from its 3-year average, gave about 75 percent with a 67 to 82 percent interval. Same family of signal, different warning level. The curve is useful, but it is not a verdict.

Why this has to do with other realms

The yield curve belongs to money, but its logic is close to concept fermi paradox. That page asks why a galaxy old enough for intelligence looks so quiet. The yield curve asks why a market that usually charges more for time sometimes charges less. In both cases, the interesting part is not the visible object. It is the missing thing that should have been there.

That is why the better companion page may be concept signal vs noise. A curve inversion can be a deep warning, a policy artifact, or a crowded trade reacting to the same public data. Reading it well means deciding whether absence is evidence or just a measurement problem.

An open question

If balance-sheet policy, forward guidance, and global demand can all flatten the same curve, what replaces inversion as the clean early-warning signal for recession? The next page worth writing may be concept term premium, because that is where the curve's simple story starts to fracture.

Key Sources

Arturo Estrella and Frederic S. Mishkin, The Yield Curve as a Predictor of U.S. Recessions (Federal Reserve Bank of New York, 1996) — the classic 10-year minus 3-month result.
Arturo Estrella and Mary R. Trubin, The Yield Curve as a Leading Indicator: Some Practical Issues (Federal Reserve Bank of New York, 2006) — how to build and interpret the signal in real time.
Campbell R. Harvey, The Real Term Structure and Consumption Growth (Journal of Financial Economics, 1988) — early academic foundation for using the term structure to infer future growth.
Board of Governors of the Federal Reserve System, Monetary Policy Report (February 26, 2013) — official account of the Fed's first large-scale asset purchases and their scale.
Bank of Japan, How have the Bank's guidelines for market operations changed? — official summary of 2016 yield curve control and the 10-year JGB target.
Richard Crump and Nikolay Gospodinov, How Uncertain Is the Estimated Probability of a Future Recession? (Liberty Street Economics, 2025) — useful corrective against treating one recession model as a certainty machine.

Abhishek's take

I feel this first through a purchase order, not a chart. When short money gets expensive, a 100-day lead time turns into a funding choice: hold the option open, cut depth, or ask the vendor to share the wait.

Abhishek