Treasury Yield Curves Split, Raising Questions About What Comes Next for U.S. Economy
Executive Summary
The Treasury market is delivering an unusually divided assessment of the U.S. economy. The 10-year/2-year spread narrowed to 39 basis points last week, down sharply from 50 basis points a week earlier and approaching its 2026 low of 22 basis points. At the same time, the 10-year/three-month spread stood at 83 basis points, leaving a 44-basis-point gap between the two measures. The divergence reflects dramatically different pricing at the short end of the curve: the two-year Treasury has risen as investors price greater odds of a rate increase, while three-month rates remain closely anchored to the Fed’s current policy setting. Similar disconnects occurred in 1998, 2019 and at the beginning of the 2022-2024 inversion cycle. History suggests the split itself does not predict recession, but it can signal a transition in expectations about monetary policy and growth.
Two Curves, Two Economic Messages
Developing a clear read on the U.S. economic outlook is challenging these days, a reality underscored by the widening split between two of the Treasury market’s most closely watched yield curves.
The 10-year Treasury ended Tuesday at 4.73%, compared with 4.34% for the two-year, producing a 39-basis-point 2s10s spread. The 10-year/three-month spread, however, finished at 83 basis points, according to Federal Reserve Bank of St. Louis data.
That 44-basis-point difference between the two curve measures is important. It essentially represents the unusually large premium investors are demanding to hold two-year Treasuries rather than three-month securities; a reflection of expectations that monetary policy may become tighter before it becomes easier.
The 2s10s spread has flattened substantially since the Iran war began on Feb. 28 and has moved back toward its June low near 22 basis points. That is a classic bearish flattening: short-term yields rise faster than long-term yields as investors price a more restrictive Fed. The 10-year/three-month curve is moving differently. At 83 basis points, it remains substantially steeper and well above levels around the beginning of the Iran conflict. The message from two-year Treasuries is increasingly hawkish.
History Shows Yield Curves Don’t Always Agree
There is a precedent for this kind of disagreement. In 1998, the 2s10s curve briefly inverted while the 10-year/three-month curve remained positive. No recession followed. The episode coincided with the Asian financial crisis, Russia’s default and the collapse of Long-Term Capital Management, demonstrating that market stress and shifting Fed expectations can temporarily distort one part of the curve without producing an economic contraction.
The reverse happened in 2019. The 10-year/three-month curve inverted while the 2s10s remained positive for much of the period. Economic growth was slowing, but the recession that ultimately arrived in February 2020 resulted from the COVID-19 shock rather than a conventional late-cycle contraction.
Another important divergence emerged in 2022. The 2s10s inverted in July as markets anticipated aggressive Fed tightening, while the 10-year/three-month curve remained positive until later that year. Eventually both inverted deeply. The 2s10s inversion lasted from July 2022 until August 2024 and reached roughly negative 108 basis points in July 2023.
The lesson is that the curves often move at different speeds around major changes in monetary policy.
Why the 3-Month Curve Matters
There is also an important distinction between the curves’ analytical pedigrees. Wall Street frequently focuses on 2s10s, but the Federal Reserve Bank of New York’s recession model uses the spread between the 10-year Treasury and three-month Treasury rate.
Research by economists Arturo Estrella and Frederic Mishkin found that the yield curve became a particularly useful predictor of recessions beyond one quarter. Their work helped establish the 10-year/three-month spread as one of the most closely studied forward-looking economic indicators.
That matters today because the 83-basis-point 10-year/three-month spread is nowhere near inversion. Viewed strictly through the New York Fed framework, therefore, the Treasury curve is not currently sending the classic recession warning.
Warsh Pushes Markets Toward a September Hike
The latest catalyst for the 2s10s flattening came from Federal Reserve Chair Kevin Warsh.
“While this summer’s [inflation] readings were better than expected, they do not tell me that underlying trends have meaningfully improved,” Warsh said at Jackson Hole last Friday
“We must be confident that underlying inflation is moving to our objective, clearly and at sufficient speed,” he added. “Otherwise, we have work to do.”
Fed-funds futures responded quickly. The probability of a 25-basis-point September increase jumped to roughly 57%-60% following Warsh’s remarks from about 35% a day earlier.
Inflation gives policymakers reason for caution. Twelve-month PCE inflation was 3.7% in July, while the six-month annualized rate reached approximately 4.1%. Longer-term inflation expectations are considerably better behaved: the 10-year Treasury breakeven inflation rate stood at 2.31% as of this writing and the five-year, five-year forward inflation rate was 2.32%.
That combination suggests investors are distinguishing between stubborn near-term inflation and longer-run inflation expectations that remain relatively anchored.
The Long End Adds Another Complication
The 10-year Treasury itself is not purely an economic growth signal. At 4.73%, its yield is supported by a combination of real rates, inflation compensation, and term premium. The 10-year TIPS yield has recently been above 2%, meaning a substantial portion of today’s nominal Treasury yield reflects historically elevated real borrowing costs.
Fiscal concerns and heavy Treasury issuance can also push long-term yields higher independently of economic growth. The government sold roughly $797 billion of Treasury securities during the final week of August, including $235 billion of notes, adding another supply consideration for investors.
That means a steep 10-year/three-month curve cannot automatically be interpreted as a bullish growth signal. Some of that steepness may represent compensation investors demand for inflation, fiscal uncertainty and duration risk.
Resilience Now, Greater Risk Later
The most defensible interpretation of the split curves is therefore less dramatic than either “recession” or “no recession.” The Treasury market appears to be pricing an economy resilient enough to tolerate Fed tightening while simultaneously acknowledging that higher rates increase the risk of weaker growth farther down the road. The 2s10s spread is flashing greater medium-term caution. The 10-year/three-month spread, meanwhile, says the classic recession signal has not arrived.
History shows that those messages can coexist. In 1998, the divergence resolved without recession. In 2022, it eventually resolved through a broad and historically deep inversion. In 2019, the signals were complicated by an external shock that no yield curve could reasonably have forecast.
The critical question now is which direction the curves converge. If 2s10s falls through zero while 10-year/three-month remains strongly positive, the market would primarily be signaling expectations for tighter Fed policy rather than an imminent downturn. If the three-month curve subsequently follows 2s10s into inversion, however, the historical recession signal would become considerably harder to dismiss.
For now, the Treasury market isn’t pricing the end of the economic cycle. But compared with earlier this year, it is putting a noticeably higher price on the possibility that today’s resilience will eventually give way to something weaker.
We want to hear your views.
What is the Treasury market really pricing: higher-for-longer rates, slower economic growth, or both?
Please share your comments below and click here for prior editions of “Treasury & Rates”
