The simple version

The term premium is the extra yield investors demand for locking money into a long-term bond instead of rolling over short-term ones. It cannot be seen directly. It has to be estimated with a model, and different models give different answers.

It matters now because the 10-year Treasury yield recently closed at 5.29%, its highest close since May 2002 by our check of Treasury's daily data. The term premium is one explanation for a move like that. One widely used New York Fed model says the premium did rise, to its highest level since 2014, but it accounts for only about a quarter of the past year's climb.

The numbers

• The 10-year Treasury par yield closed at 5.29% on September 30, 2026, 5.24% on October 1, and 5.28% on October 2 (U.S. Department of the Treasury, Daily Treasury Par Yield Curve Rates)
• The September 30 close was the highest since May 14, 2002, when it was 5.32%, by our search of Treasury's daily data back to 2000 (U.S. Department of the Treasury, Daily Treasury Par Yield Curve Rates)
• A year earlier, on October 1, 2025, the 10-year closed at 4.12% (U.S. Department of the Treasury)
• The New York Fed's ACM model estimated the 10-year term premium at 0.91 percentage point on October 1, 2026, up from 0.61 a year earlier (Federal Reserve Bank of New York, ACM term premium data)
• The same model estimated the expected average short-term rate over the next ten years at 4.31% on October 1, 2026, up from 3.55% a year earlier (Federal Reserve Bank of New York, ACM term premium data)
• The last time the model's 10-year term premium was at or above its October 1 level was July 3, 2014, by our check of the daily series (Federal Reserve Bank of New York, ACM term premium data)
• The 30-year fixed mortgage rate averaged 7.28% as of October 1, 2026, up from 6.34% a year earlier (Freddie Mac, Primary Mortgage Market Survey)

Two pieces inside one yield

The New York Fed describes a Treasury yield as two parts. The first is what investors expect short-term rates to average over the life of the bond. The second is the term premium, the extra pay investors want for the risk that rates change over that time.

Neither part has a price tag. It estimates them with a statistical model known as ACM, after the three economists who built it, and says its figures are not official estimates of the Federal Reserve. A different model can split the same yield differently.

On its latest reading, for October 1, the model split the 10-year yield into 4.31% of expected short-term rates and 0.91 percentage point of term premium. Compared with a year earlier, the expected-rate piece rose about 0.76 point and the term premium about 0.29 point. By this model, roughly three quarters of the year's increase came from investors expecting higher short-term rates, not from extra pay for risk.

The model works from its own fitted yields, which differ slightly from Treasury's published closes. The split still answers a useful question. A rise in expected short rates is a view on where the Federal Reserve (the Fed) takes its policy rate, while a rise in the term premium is a demand for more pay to own long bonds at all.

The Real Cost lens on a long-rate climb

Mortgage rates are not set by the Treasury market. Lenders price them, and they rose over the same year. Here is that change on one loan, using Freddie Mac's weekly averages and stated assumptions.

• Assumption: a $300,000, 30-year fixed loan, principal and interest only, with no taxes, insurance, or points.
• At 7.28%, the Freddie Mac average as of October 1, 2026, the payment is about $2,053 a month.
• At 6.34%, the average a year earlier, the payment is about $1,865 a month.
• The difference is about $188 a month, about $2,255 a year, or about $67,640 over 30 years if the rate never changed.

The payment does not care which piece of the yield did the moving. The split still tells a reader what kind of change is being priced: expected Fed policy, or extra pay for lending long.

What this means

When long-term rates jump, the explanation usually comes in one of two forms. Either the market expects short-term rates to stay higher, or investors want more pay for lending long. The term premium is the name for the second.

Because it is estimated, any single figure for it belongs to a model. The durable habit is to ask which model, and what share of the move it assigns to expectations versus the premium.

What this is NOT

This article explains what the term premium is and how one model divides a Treasury yield. It does not forecast Treasury yields, mortgage rates, or Federal Reserve decisions, and it is not advice on buying or selling bonds, refinancing, or timing a home purchase. The term premium and expected-rate figures are estimates from the New York Fed's ACM model, which the New York Fed says are not official Federal Reserve estimates, and other models produce different numbers. The highest-since-2002 and highest-since-2014 comparisons are our searches of Treasury's and the New York Fed's published daily data. The mortgage payments are illustrations with stated assumptions, not quotes. This is not a political endorsement or criticism of anyone, including fiscal or monetary policymakers.

Sources

• U.S. Department of the Treasury, Daily Treasury Par Yield Curve Rates, 2026: https://home.treasury.gov/resource-center/data-chart-center/interest-rates/TextView?type=daily_treasury_yield_curve&field_tdr_date_value=2026
• U.S. Department of the Treasury, Daily Treasury Par Yield Curve Rates, 2002 data file: https://home.treasury.gov/resource-center/data-chart-center/interest-rates/daily-treasury-rates.csv/2002/all?type=daily_treasury_yield_curve&field_tdr_date_value=2002&page&_format=csv
• Federal Reserve Bank of New York, Treasury Term Premia: https://www.newyorkfed.org/research/data_indicators/term-premia-tabs
• Federal Reserve Bank of New York, ACM term premium data file: https://www.newyorkfed.org/medialibrary/media/research/data_indicators/ACMTermPremium.xls
• Freddie Mac, Primary Mortgage Market Survey, October 1, 2026: https://www.freddiemac.com/pmms