Icarus Asia

Research

Rates & Macro Research

September 14, 2026

U.S. Treasuries · Term Premium · Fed Policy

Above Five, Then Back Below

The 10-year Treasury touched its highest yield since 2007 on Monday. Buyers showed up again. A term premium near 0.9 percentage point means the next test will be harder to pass.

A Familiar Level, an Unfamiliar Backdrop

Benchmark 10-year Treasury yields rose above 5% on Monday for the first time since October 2023, touching an intraday high of 5.014% before buyers pushed the yield back under the threshold.^{2 3} The move put the note at levels not otherwise seen since July 2007; a rise past roughly 5.02% would have marked a new high for the entire post-financial-crisis period.^{10 13}

The reversal is the part worth watching. In October 2023, the 10-year briefly touched about 5.02% and closed near 4.84% the same day; it fell to roughly 3.79% by year-end as investors covered short positions and bought duration.^{12 13} The 2026 test produced a smaller same-day retreat, to about 4.96% per TradingEconomics, which suggests 5% still functions as a demand trigger. It has not yet produced the kind of capitulation in bearish positioning seen three years earlier.²⁰

Figure 1 · Two Tests of Five Percent

Figure 1: Intraday high and same-day retreat are directly comparable across both episodes. The October 2023 year-end level (3.79%) reflects nearly three months of subsequent disinflation and Fed pivot expectations. There is no equivalent data point yet for 2026, since that period has not elapsed. Sources: Wolfstreet, Investopedia, TradingEconomics.^{12 13 20}

What Five Percent Is Actually Made Of

A 10-year nominal yield is, approximately, the expected average short-term policy rate over the life of the bond plus a term premium: the extra compensation investors demand for the risk of holding a fixed-rate, long-duration asset rather than rolling short-term paper.¹⁴ The Dallas Fed put the 10-year term premium at roughly 60 basis points in June; by early September, the New York Fed's ACM model estimate had risen to about 89 basis points.^{14 6}

The two estimates come from different models (the Dallas Fed's term-funding-premium framework and the New York Fed's ACM model), so the move from about 60 basis points to about 89 basis points reflects both a genuine increase in duration risk and some methodological difference between them. The direction matters more than the precise size of the move.

The distinction matters for what comes next. A yield increase driven by stronger expected growth can coexist with healthy earnings and stable credit performance. A yield increase driven by inflation or fiscal risk premium is more destabilizing: it raises discount rates without guaranteeing stronger nominal cash flows, and it reduces the odds of a near-term Fed easing cycle.

Oil, Then Inflation, Then the Fed

Energy triggered it first. Crude oil climbed above $100 a barrel in the U.S. while Brent approached the high $90s to $100 range, sparking a global bond selloff on concern that higher energy costs would feed into broader pricing expectations.^{15 1} August's CPI report reinforced the concern: headline CPI rose 0.4% month over month and 3.4% year over year, with gasoline up 3.9% for the month and the energy index up 16.3% from a year earlier. Core CPI, at 0.3% month over month and 2.4% year over year, was less alarming. But the headline acceleration was enough to make a near-term Fed hike look likely.^{4 16}

Figure 2 · August 2026 CPI, Year Over Year

Figure 2: Headline and core figures are seasonally adjusted 12-month changes; the energy index is the broader BLS energy category, not the narrower gasoline sub-index. Source: U.S. Bureau of Labor Statistics, August 2026 CPI release.¹⁶

Economists moved fast after that. In the wake of the inflation release, 85% of those surveyed by Reuters expected a 25-basis-point increase at the September 15–16 Fed meeting, and futures markets priced close to a 90% probability of a hike plus roughly four more increases by July 2027.⁹ That repriced both components of the yield at once: the expected path of short rates, and the uncertainty premium investors attach to how far and how fast the Fed will actually move.

Why the Rally Didn't Come

The fiscal backdrop explains why the 10-year did not rally decisively on approach to 5%. The Congressional Budget Office projects a $1.9 trillion federal deficit for fiscal year 2026, equal to 5.8% of GDP, with debt held by the public at 101% of GDP. Net interest outlays are projected to exceed $1 trillion in 2026 and to rise from 3.3% of GDP this year to 4.6% by 2036.^{5 17}

Figure 3 · The Fiscal Trajectory, CBO Projections

Figure 3: All figures are CBO baseline projections and subject to revision in future reports. Debt held by the public is shown on the same percent-of-GDP axis as the deficit and interest series for comparability; note the difference in scale. Source: Congressional Budget Office, The Budget and Economic Outlook: 2026 to 2036.^{5 17}

These numbers create both a supply problem and a risk-premium problem. Treasury has to place a large volume of securities even in a non-recessionary year, and investors have to judge whether higher interest costs will widen future deficits and require still more issuance. CBO projects debt held by the public reaching 120% of GDP by 2036 and the deficit widening to 6.7% of GDP despite a smaller primary deficit, a sign that interest expense itself is increasingly driving the debt trajectory.¹⁷

Auction sizes illustrate the scale of what the market has to absorb. The August refunding offered $125 billion across 3-, 10- and 30-year maturities, including a $42 billion 10-year note, and raised about $28.7 billion in new cash from private investors.^{18 19} Treasury has said nominal coupon and floating-rate-note auction sizes should hold steady for at least several quarters. But "stable" at today's elevated debt stock still means substantial recurring duration supply.¹⁸

Private borrowing adds to the competition for capital. Heavy corporate issuance tied to artificial-intelligence infrastructure is running alongside sovereign borrowing, which matters because the marginal Treasury buyer is comparing duration not just against equities and cash, but against investment-grade credit offering additional spread. Stronger private credit demand can push up the yield required to clear a Treasury auction.¹

Why 2026 Is a Harder Test Than 2023

The key difference is the policy path ahead. In 2023, 5% ultimately proved restrictive enough to slow the economy and pull in buyers as the tightening cycle matured. In 2026, the market is confronting a fresh inflation impulse before it has any confidence that policy is restrictive enough, which means 5% may generate tactical demand without yet marking the cycle high. A second difference is composition: when the term premium is elevated and rising, a weaker economy does not automatically produce a proportional bond rally, because fiscal supply and inflation uncertainty can offset a lower expected path for short rates. That raises the odds of a "bear steepening" episode, in which long yields rise relative to the front end as investors demand more compensation for duration.

Where a Sustained 5% Actually Bites

Housing is the fastest domestic channel. As the 10-year moved toward 5%, the average 30-year fixed mortgage rate crossed 7%.²¹ Persistently high mortgage rates suppress transaction volume, weaken affordability, and reinforce the "lock-in" effect for homeowners still holding older, low-coupon mortgages.^{22 23}

Equity sensitivity depends on earnings. The S&P 500 has stayed resilient even as the 10-year approached 5%, with strong profit growth offsetting some valuation compression. That dynamic can persist while nominal growth and earnings hold up. It turns unstable if higher yields coincide with weaker earnings, wider credit spreads, or margin compression from the oil shock itself.^{11 24} The valuation arithmetic is already demanding: at the end of June, one estimate placed the S&P 500's trailing earnings yield near 4.11% against a 4.38% 10-year yield, a negative gap that puts the burden of proof on equity valuations to justify a lower current earnings yield with durable growth.²⁵

Buybacks Without Shock and Awe

Treasury raised the size of its long-end liquidity-support buybacks from a maximum of $2 billion to at least $4 billion per operation, effective September 9, and Secretary Scott Bessent later signaled purchases could go higher. A September operation as large as $6 billion was announced.^{8 26} The market's muted reaction is itself informative.²⁷

Buybacks are not quantitative easing. They exchange one government liability for another and are meant to improve market liquidity and functioning, not to eliminate the underlying supply that private investors ultimately have to finance. Absent lower net borrowing or a shift in the maturity profile, their effect falls mainly on liquidity and auction dynamics, not on the long-run equilibrium yield. If markets came to see larger buybacks as an attempt to target a specific yield rather than to smooth liquidity, that could increase the risk premium rather than reduce it, by raising doubts about the consistency of debt-management policy.

Four Paths From Here

What This Means for Portfolios

At roughly 5%, the 10-year offers real nominal carry and a credible hedge against disinflation or recession, but it is not automatically low-risk. Using a modified duration near eight years as a rough market convention, a 50-basis-point rise in yield implies an approximate 4% price decline before carry and convexity, and a comparable decline in yield produces a similar-sized gain — illustrative math based on standard duration arithmetic, not a specific security.

A staged entry into duration is more defensible than treating 5% as a single binary signal. An initial allocation captures improved carry and potential convexity; further additions can be conditioned on oil, inflation breakevens, auction performance, and evidence that the term premium has stopped climbing. Investors focused on income may prefer intermediate maturities or a barbell structure until the long-end supply premium stabilizes.

For equities, the more useful screen is balance-sheet and cash-flow duration rather than a blanket call on the asset class. Companies generating current free cash flow, with low refinancing needs and real pricing power, should tolerate a higher sovereign discount rate better than growth stories funded with external capital, provided earnings revisions stay positive.²⁴ For emerging markets, a 5% Treasury benchmark raises the required local-currency carry and sharpens sensitivity to oil, external balances and reserve adequacy; commodity importers with weak fiscal positions face the toughest combination of higher energy bills, imported inflation and tighter dollar liquidity.

Indicators to Monitor

• Fed reaction
  Size of the September move, forward guidance, and whether the Committee treats the oil shock as temporary or as a risk of second-round inflation.⁹
• Oil persistence
  A reversal in crude removes the immediate catalyst; sustained triple-digit oil keeps both headline inflation and breakeven risk elevated.¹⁵
• Inflation composition
  Headline is 3.4%, core is 2.4%: a broadening from energy into services and wages would be more consequential than gasoline alone.^{4 16}
• Term premium
  The early-September estimate near 0.89 percentage point is the key gauge of whether the selloff is structural or largely a Fed-path repricing.⁶
• Auction quality
  Tails, bid-to-cover ratios, dealer takedown and indirect-bidder participation at 10-, 20- and 30-year auctions will test real-money absorption.
• Curve shape
  A front-end-led selloff signals Fed repricing; long-end underperformance signals term-premium and fiscal stress.
• Cross-asset confirmation
  A stronger dollar with resilient credit points to a growth/policy move; a weaker dollar with wider credit and lower equities points to a confidence shock.
• Mortgage spread
  Even if the 10-year stabilizes, elevated volatility and prepayment uncertainty can keep mortgage rates well above the Treasury benchmark.²³

A Test, Not Yet a Breakout

The first move through 5% was a test, not a breakout. Buyers defended the threshold, market functioning stayed orderly, and the yield pulled back from its intraday high. But the combination of an oil-driven inflation shock, a likely Fed hike, a term premium near 0.9 percentage point and a deficit close to 6% of GDP makes this episode more structurally challenging than the October 2023 spike.^{2 3 5 6 9}

Repeated closes above the 2023 high, especially alongside a bear-steepening curve and weak auctions, would signal that 5% is no longer resistance, but the floor of a new, higher-yield regime.