Two-Year Treasury Yields Jumped Almost 60 Basis Points in September

Two-year U.S. Treasury yields rose almost 60 basis points in September 2026, on track for the biggest monthly jump since early 2023, according to Reuters. A basis point is one-hundredth of a percentage point, so that move is about 0.60 points. Yield is the annual return from holding the bond. When yields rise that quickly, prices of existing bonds fall and short-term borrowing costs rise.

U.S. Treasuries have often lost value at this point in the calendar. Over the past decade, the median September result was a loss of 0.9% and the median October result was a loss of 0.7%, according to data compiled by Bloomberg. Bloomberg

The Federal Reserve publishes H.15 Selected Interest Rates (Daily), with an edition dated September 29, 2026. Federal Reserve The H.15 Data Download Program, last released on Monday, September 28, 2026, offers a Treasury Constant Maturities dataset. Federal Reserve The program tracks series RIFLGFCM03_N.B, market yield on U.S. Treasury securities at 3-month constant maturity quoted on investment basis from September 1, 1981 to September 25, 2026, and includes a series for market yield at 5-year constant maturity. Federal Reserve Federal Reserve

The broader context here is what a fast repricing at the short end tells you. Two-year constant maturities sit where funding costs, expectations for Fed policy and hedging demand meet. A shift of almost 60 basis points squeezes carry, the profit from holding the bond, raises variation margin calls that demand extra cash against derivatives positions, and forces a new valuation on anything tied to short-term benchmarks. Liquidity thins. Positions across the yield curve get reviewed.

In my view, the seasonal pattern adds to the technical pressure rather than explaining it. Back-to-back median losses for September and October do not predict returns, but they shape how desks hold bonds into month-end. When bonds are already falling, dealers cut back their balance sheets. Managers matching assets to liabilities delay buying longer bonds. That can push prices in the same direction even without a change in fundamentals.

Looking at what this means for measurement, the H.15 constant-maturity build matters more in months like this than headline price indexes. Like using a fixed ruler, it estimates par yields at exact tenors. That allows clean comparison over time when specific bonds age and carry different coupons. The 3-month series quoted on investment basis and the 5-year series sit on the same curve structure as the 2-year point that drove September. That consistency is why trading, risk and accounting teams anchor to it.

In my view, the date of the data deserves attention from anyone valuing portfolios or explaining performance. A download dated September 28 with last observations through September 25 leaves month-end trading out of the published history at release. The September 29 daily release narrows that gap but does not remove month-end effects. For performance measurement, that lag opens a gap between internal prices and published fixings. For risk, it argues for using live price feeds alongside H.15 to check limits.

Looking at October positioning, the market starts short on duration and heavy on caution. Duration measures sensitivity to rate moves. Past losses are already known. What is priced for the next roll of maturing bonds, for term premium and for digesting new supply is less visible. Discipline means keeping past volatility separate from expected return, and using seasonal medians to plan for liquidity rather than as a signal.