Ahead of the curve?

The yield curve has attained almost sacred status in the economic analyst’s medical kit. It is consulted for recession warnings, interrogated for clues about the Federal Reserve and occasionally treated as though it’s as dispositive as the crop report in Trading Places. If it inverts? Supposedly stormy seas ahead. Then a return to an upward slope invites relief. Maybe we are getting back to normal? That depends on what we mean by normal.

When you look at charts or history, you can often see echoes of the present. Our history of US Treasury yields, stretching back to 1962, offers plenty of familiar-looking shapes. But, we argue, it also shows why appearances can deceive and how history never exactly repeats itself. Two curves can slope in the same direction while describing very different economic circumstances. The current curve, a snapshot from 1995 and another from 2008 make that point well. Then, for more robust conclusions, we will want to take a closer look and understand the absolute level of yields, which maturities have moved by how much, and what might explain the movement.

A Chorus Line?

A yield curve plots the yields available on bonds of different maturities at a particular moment. Here, those bonds are US Treasuries, ranging from bills maturing within months to bonds extending thirty years into the future. It is a cross section of today’s market, rather than a forecast showing where one interest rate will go over time. In some respects, it is like the VIX index, reflecting the price a large number of market participants demand to bear risk.

The short end is most closely influenced by the Federal Reserve’s policy rate and expectations for its next decisions. Further along the curve, investors must consider the likely path of interest rates over many years. They also require compensation for the uncertainty involved in committing money for longer. Economists call that component the ‘term premium.’ The term premium can change; it can even be negative. The term premium is estimated/inferred—it cannot be directly measured.¹

An upward slope means longer maturities offer higher yields than shorter ones. Inversions, with longer maturities commanding lower yields, are seen as foreboding: historically, inversions have provided useful warnings about recession risk. But a warning doesn’t mean you will actually experience a weather event in your own specific neighborhood. It’s not a timetable nor a diagnosis of the entire economy, and different analysts/forecasters focus on different things. The New York Fed’s recession framework, for example, uses the difference between ten-year and three-month Treasury yields. The widely followed ten-year/two-year spread is a different measure. (LucenWorks, the analytics platform used by Belvest, allows us to view each of these.)²

You can see that in action in the lower panel of our charts, which plots both spreads. Above zero, the longer yield is higher. Below zero, that part of the curve is inverted. Shading shows historical recessions (determined ex post).

Today’s Curve: Higher Borrowing Costs

On October 1, 2026, the Treasury’s published three-month yield was 4.17%, the two-year 4.78%, the ten-year 5.24% and the thirty-year 5.61%. That puts the ten-year yield 1.07% above the three-month yield, but only 0.46% above the two-year. The curve is broadly upward-sloping, although it levels off at the longest maturities.³

Chart 1

An Upward Slope, at Higher Yields

US Treasury yield curve, October 1, 2026, with earlier curves from October 2024

The dark curve lies above the dashed October 31, 2025 comparison across most maturities, particularly from two years onward. The defining feature is higher medium- and long-term borrowing costs. The lower panel shows the ten-year yield less the three-month and two-year yields, in percentage points. Current nominal yield figures cross-checked against the US Treasury’s Daily Treasury Par Yield Curve Rates. Source: Federal Reserve H.15 via FRED; NBER recession dates; prepared with LucenWorks.

The earlier curves show how both the level and shape of the yield curve have changed. A ‘parallel shift’ would move yields at every maturity by the same amount, leaving the slope unchanged. But here, yields have risen unevenly, with larger increases from two years onward than at the shortest maturities. Today’s upward slope therefore reflects both higher yields and a change in shape. Further, acclaiming it “normal” glosses over the painful rise in borrowing costs at most maturities.

We also show a pale blue curve showing yields on Treasury Inflation-Protected Securities, or TIPS. These are real yields, reflecting returns above the inflation adjustment built into those securities. The gap between nominal and real yields provides a measure of inflation compensation. It is not a pure inflation forecast, because inflation risk and differences in liquidity also influence the gap. But it’s another important piece of the puzzle.⁴ Belvest and other investors are most concerned with what is producing the higher yields. There are a number of contenders: stronger expected growth, a higher expected path for policy rates, and increased risk aversion are among the ‘usual suspects’. And there are likely other contributors.

1995: Bond-Market Pain, Economic Resilience

A keen reader will note that The Usual Suspects was actually released in 1995. But, we digress. It’s also useful as another echo of today’s yield curve. March 1995 was at the end of a sharp tightening of monetary policy in 1994. The Federal Reserve had moved to restrain emerging inflation pressures, and bond markets sold off in sympathy. Despite the tightening of financial conditions, the ensuing economic slowdown became a soft landing rather than a recession. Growth continued and inflation remained contained.⁵

Chart 2

Higher Rates Did Not End the Expansion

US Treasury yield curve, March 31, 1995, with earlier curves from March 1993

Relative to the dashed March 1994 curve, short-term yields had risen much more than long-term yields, leaving a flatter curve at higher rates. The subsequent soft landing illustrates why losses in bonds need not imply an approaching recession. TIPS were first issued in 1997, so no real-yield curve appears for 1995. Source: Federal Reserve H.15 via FRED; NBER recession dates; prepared with LucenWorks.

Look at the left side. Short yields have moved up considerably relative to the year-earlier comparison, while the long end has moved less. The ghosts of yield curves past show how the adjustment rippled through the intervening period.

This episode does much to establish the tension between an asset’s price and the economic backdrop. And the importance of your starting point. Obviously, existing bondholders will suffer when markets demand higher yields, even if the economy continues to expand. On the other hand, for a new purchaser, those same higher yields mean superior returns for new bonds purchased and held to maturity. That makes 1994–95 another informative antecedent.

2008: A False Dawn… The Curve Improved Before the Economy Did

The Global Financial Crisis was such a seminal event that it resists facile comparisons. But it does provide a strong example that the yield curve can change much more rapidly than the real economy (although it did foreshadow it well). In March 2008 the economy was already in a dark recession. The Federal Reserve had been cutting interest rates as conditions deteriorated, but that was only pronounced at the shorter end of the curve.⁶

Chart 3

A Positive Slope During a Recession

US Treasury yield curve, March 31, 2008, with earlier curves from March 2006

The nearly flat March 2007 comparison gave way to an upward-sloping curve as short-term yields fell much further than long-term yields. The return to a positive slope accompanied economic deterioration. It was not an all-clear signal. Source: Federal Reserve H.15 via FRED; NBER recession dates; prepared with LucenWorks.

Here, the yield curve ghosts tell a different story: the short end has dropped sharply below earlier curves. Longer yields have also declined, but by less. The curve has become steeper through falling yields, especially at shorter maturities.

That is a genuine contrast with today’s picture. Both episodes have positive slopes, but the movement behind them differs. In March 2008, a beguilingly conventional shape coexisted with worsening conditions. The economy was bad and it was getting worse. So, you had the reversal of an inversion without coming close to the end of the economic pain. This goes to show that a facile heuristic—upward slope, good; inversion, bad—is incomplete and often misleading.

The Speed Matters, Too

The Financial Times recently discussed several interesting implications for today’s bond markets (Unhedged discussion, “Bond bombshell”). Robert Armstrong argued that higher yields can make sense alongside strong nominal growth, while Katie Martin and Ian Smith emphasized that the speed of the adjustment and the potential for forced selling matter, too. Both observations can be true: yields can rise for understandable economic reasons while putting financial institutions and leveraged investors under pressure. For our purposes, the discussion adds another dimension to the historical comparisons: the curve’s level and slope matter, but so does how quickly markets move between them.⁷

For such insights we might want to consider the lessons of Britain’s 2022 gilt crisis. Then, rising yields precipitated collateral calls on leveraged pension investment strategies. As many things do, it looked good at the time. But soon forced sales worsened market liquidity and accelerated price declines.⁸ Again, that episode is not necessarily a strict analogue for today, but it drives home the point that the pace of a move deserves attention alongside its direction. Liquidity is king: a portfolio that can withstand an upheaval spread across months may struggle if it must produce cash in days.

Our snapshots show levels and changes in shape and slope. On their own they can’t establish whether trading is disorderly or investors are being forced to sell. Those questions require evidence about liquidity, financing, and positioning.

What We Would Watch

What additional evidence would we need to see to put today’s curve in the right context? We will be watching for a confluence of factors. For instance, falling short-term yields alongside weaker employment and wider corporate credit spreads would make downturn comparisons more relevant. Conversely, higher long yields alongside persistent inflation or increasing financing pressure would raise different questions about the cost of capital. Stable yields accompanied by moderating inflation and resilient activity would offer a more constructive combination. As the FT discussion makes clear, an economy can remain strong enough to justify higher yields while parts of the financial system still struggle to absorb the adjustment. The UK’s 2022 experience shows how losses and collateral demands can turn a repricing into forced selling.

Our reading is that today’s curve deserves more attention as evidence of higher borrowing costs than as reassurance that recession risk has passed. The difference from March 2008 is important: today’s positive slope accompanies higher yields across much of the curve, rather than a collapse in short-term rates. Borrowers refinancing at higher yields have a higher hurdle. Investors buying bonds receive more income, but accepting additional maturity still means accepting additional exposure to interest-rate changes.

For investors, the conclusion is more complex than noticing “bonds now pay more.” The investment implications depend on the investor and his or her balance sheet and objectives. The additional yield must justify the maturity risk, and portfolios need enough liquidity to avoid selling under pressure. The soft landing of 1995 shows that a bond sell-off need not end in recession. March 2008 shows that an upward slope offers no protection against one. Today’s curve offers better starting income alongside higher financing costs.

Appendix

The Yield Curve in Motion: 1962–2026

Source: Federal Reserve H.15 via FRED; NBER recession dates; prepared with LucenWorks.

Appendix

Endnotes

Important disclosures