Inflation is spiking, that is certain. But there’s a huge debate brewing among economy-watchers. Is this a “transitory” increase? Or something more long-term and structural?

In testimony before Congress the week of June 21, Federal Reserve chair Jerome Powell characterized the big new wave as “transitory.” For Powell, what’s causing the jump is the confluence of rising demand as the economy roars back and popular products are temporarily in short supply. Powell noted that depleted stocks of and heavy orders for computer chips and used cars and trucks have sent their prices soaring. But he predicted that production of such scarce goods will ramp up quickly, reversing the spikes. He also posits that the famous worker shortage will soon ease, restraining today’s big wage increases. “If you look behind the headlines and look at the categories where these prices are really going up,” Powell testified, “you’ll see that it tends to be in areas that are directly affected by the reopening.”

But another lesser-watched inflation indicator—the Treasury yield curve—may be signaling something different. Put simply, the best road map to where rates are headed isn’t the outlook from Wall Street or the Fed. It’s what’s baked into the array of yields set by the galaxy of funds and individuals wagering money for their clients and themselves. “The yield curve is the summary of the market’s best estimates,” says Bruce Sherrick, an agricultural economist at the University of Illinois.

The yield curve is simply the line linking the dots that mark the current yields on Treasuries of all maturities. As of June 22, the slope starts low and super-flat at the one-month bill, paying a minuscule 0.04%, rising to the one-year at 0.09%, then to the two-year (0.26%), three-year (0.44%), five-year (0.90%), on to the benchmark 10-year at 1.48%, and finishing at the 30-year (2.10%).

What’s mostly ignored is the forecast embedded in those numbers. They contain a guide for where the one-, two-, three-year, and other maturities will go in the years ahead. If, for example, you buy a five-year today, at 0.90%, you’ll be getting a compound return of 4.6% total by mid-2026. Now let’s run the numbers. The math says that the total gain on the five-year has to equal the sum of what you’d get from buying five one-year bonds in each 12-month period from mid-2021 to June 2026, reinvesting the interest each time. So if you’re starting this year at just 0.09%, one-year yields in future years must ramp up fast to get you a total, cumulative return over those five years of 4.6%.

Today, the curve is predicting much higher rates in a few years for exactly that reason: Even though yields on, say, the three- and the five-year are extremely low in historical terms, the one-year is so incredibly depressed by the Fed’s stimulative stance that future one-years need to jump, and keep jumping, to reach what you’d get holding the three- or five-year to maturity.

With Sherrick’s assistance, I assembled the “forward” numbers showing where the market is betting yields will go in future years. The curve forecasts that the one-year will jump from that 0.09% today, to 0.82% in 2023, 1.3% in 2024, and 2.32% in 2027. The “through the windshield” view shows the two-year and three-year going from 0.26% and 0.44% to 1.29% and 1.45%, respectively, in 2023. The five-year? It’s expected to rise from 0.90% to 1.62% over the next 24 months.

So although the yield curve predicts that future rates will be low in historical terms, it also augurs that huge increases are coming.

Here’s why that shift matters: Since the yield curve forecasts future rates, a shift in the curve changes the forecast. Three weeks ago, bondholders reckoned the one-year would reach 0.64% in 2023; now, the projection is 0.82%. On May 7, the market put the two-year in mid-2022 at 0.45%; now, the “baked in” number is 0.62%. Likewise, the 12-month forecast on the three-year has gone from 0.73% to 0.84%.

“The yield curve shift is signaling something that is really important,” says John Cochrane, an economist at the Hoover Institution. A plausible explanation, he says, is that inflation will be higher in the next several years than previously thought. Adds Sherrick, “We’re seeing higher expectations for inflation than a few weeks ago.”

That the market is predicting higher inflation than just a few weeks ago is notable—and at odds with what chairman Powell has been predicting. But there’s one final twist: Over this time yields on the 10-year Treasury went the opposite way. Since May 7, the long bond yield fell from 1.60% to 1.48%.

Hence, in just a couple of weeks, the bond market made two conflicting judgments about future inflation. First, that prices would rise faster than anticipated over the next three to five years. And second, that inflation would be lower than it predicted from 2026 to 2031. “That long yields went down while short yields rose is a real puzzle,” says Cochrane.

One thing’s for sure, however: Over the next one to five years, the pros are betting inflation will run hotter than the Fed is currently expecting. Meaning our era of ultralow rates may be the thing that’s transitory—and inflation is with us for the short-to-medium term.