What the 10-Year Treasury Yield Is, and Why It Sets Your Mortgage Rate

The interest rate the government pays to borrow for a decade. The Fed doesn't set it. Bond buyers do, by guessing what the Fed will do for the next ten years.

In one sentence

The 10-year yield is the annual return an investor gets for lending to the U.S. government for ten years at today's price, and because that is the safest ten-year loan in the world, nearly every other long-term rate in the country is priced as a markup on it.

A bond pays fixed dollars, so when its price falls, its yield rises. "Yields rose today" means bond prices fell, which means somebody sold.

New articles and charts arrive in one short email on Monday mornings. Get the weekly digest

The analogy

What lenders charge for time. The Fed sets the price of money overnight. The 10-year is the price of money for a decade, and nobody sets it: it is wherever buyers and sellers of Treasury bonds settle, minute by minute. Think of it as a pooled bet by thousands of investors on three things at once: where the Fed's rate will average over the next ten years, how much inflation will eat along the way, and how much extra they want for locking money up that long, which economists call the term premium. The Fed moves the first. The bond market moves all three, and it moves them before the Fed acts, on what it expects.

How it reaches your kitchen table

Mortgage rates: within days. A 30-year mortgage rate is the 10-year yield plus a spread of roughly 1.5 to 2 points, and lenders reprice daily. In October 2026 the 10-year averaged 5.1%, its highest since May 2002, and the 30-year mortgage rate averaged 7.3% in October 2026. When the Fed cut three times in late 2024 and mortgage rates rose a full point, this was the line that went up: the market decided inflation and deficits would keep long rates high regardless. Mortgage rates has the spread.

Car loans, student loans, business borrowing: weeks. Priced off Treasuries of matching length. A company's bond pays the Treasury yield plus a markup for its own risk.

Your retirement account: the same day. Bond funds fall when yields rise. In 2022 the 10-year went from 1.5% to 4% and bond funds had their worst year on record. Stocks feel it too: when a safe bond pays 5%, a risky one has to promise more.

The federal budget: as the debt rolls over. The government refinances constantly, so each point on the 10-year works into the interest bill over several years. That bill is set by the level of the debt, not the deficit, which rate vs. level explains.

Your savings: finally. Certificates of deposit and Treasury bonds pay something again. From 2009 to 2021 the 10-year spent most of its time under 3%, and savers in long bonds earned less than inflation for a decade.

What moves it

Expectations for the Fed, first: a hotter inflation report raises the 10-year the same morning because the market pencils in higher rates for longer. Inflation expectations themselves. The term premium, which rises when investors are uncertain or when there is more debt to absorb. Supply, meaning deficits, and demand, meaning foreign central banks, pension funds, insurers and the Fed itself: the Fed's bond-buying pushed the 10-year to 0.6% in 2020, and its runoff from 2022 to December 2025, when it stopped shrinking its holdings, took that buyer away. And fear: in a panic money runs into Treasuries, prices rise and the yield falls, which is why the 10-year can drop on bad news.

How to read this month's number

Against the Fed funds rate. The slope of the curve, the 10-year minus the Fed's rate, is on the chart as a companion, along with the 10-year minus the 3-month bill, the version the New York Fed uses to estimate the odds of a recession. Normal is one to two points above. Below zero, an inverted curve, means the market expects cuts, usually because it expects a slowdown. The Fed funds version went below zero before each of the eight recessions since 1969. It also went below zero with no recession following in 1966–67 and 1998, and for two years in 2022–24, the deepest inversion since 1981 (minus 1.5 points in May 2023), with no recession so far; it dipped below zero again briefly in 2025.

Against inflation. A 5% yield with inflation expected near 2.5% is a real return of about 2.5%, high by the standards of the 2010s and ordinary by the standards of the 1990s.

Against its own history. Five percent is where the 10-year spent most of the 1990s and 2000s. What was unusual was 2009–2021, under 3% nearly the whole time, which is the period everyone's memory of mortgage rates comes from.

The 10-year yield, the slope of the curve (10-year minus Fed funds), and the Fed's hiking and cutting cycles. Where the slope dips below zero, the market expected cuts. It was right before 2001, 2008 and 2020, and wrong in 2022–24. Full chart and context →

Where the analogy breaks

A pooled bet assumes the bettors are betting. The biggest buyers of Treasuries often aren't: the Fed during QE, foreign central banks managing their currencies, pension funds and insurers matching liabilities. They buy for reasons unrelated to any forecast, which is how the term premium went negative from 2016 to 2020, meaning investors paid for the privilege of locking money up. So the 10-year is a forecast plus a great deal of plumbing, and the plumbing changes.

What people argue about

Deficits and yields. One side says a 10-year at 5% with the Fed near 4% is the bond market charging for deficits of around 6% of GDP in a full-employment economy: the term premium, near zero in 2021, is back. The other says long yields mostly track expected Fed policy and inflation, that the term premium's rise is as much about inflation uncertainty as about debt, and that Japan carried far larger debts at near-zero yields for decades. The New York Fed's estimates show the term premium rising since 2021. Why it rose is the argument.

Whether the bond market disciplines governments. The case for: Britain in September 2022, when a budget of unfunded tax cuts sent gilt yields up a point in days, forced the Bank of England to intervene, and ended a prime minister's term in seven weeks. The case against: the United States ran large deficits through 2017–2024 and the market, with the world's reserve currency, mostly didn't charge for them until recently, if it is charging now. Both sides agree the discipline exists in principle. They disagree about how far from it the United States stands.

The inverted curve. After 2022–24 one camp says the signal is broken, distorted by the Fed's balance sheet and by foreign demand. The other, Campbell Harvey among them, points to the measure he studies, the 10-year minus the 3-month bill, which by his count had called eight out of eight recessions since the 1960s without a false signal, and argues that the warning itself softened the slowdown as companies cut back early. The Fed funds version on this chart has a looser record, with misses in 1966–67 and 1998 as well. The next recession will reopen the question either way.

See also

Sources

Corrections

October 29, 2026: An earlier version implied that 2022–24 was the only time the 10-year minus Fed funds spread had gone below zero without a recession following. It also did in 1966–67 and 1998.

October 30, 2026: An earlier version described the Fed's runoff of its bond holdings as still under way. It ended on December 1, 2025.

The weekly digest

One email on Monday mornings: the week's new articles and charts, the numbers that moved most, and what comes out in the week ahead. No ads, and no tracking of opens or clicks.

Past issuesHow addresses are kept