The 10-year just broke 5% while the CBO projects 4.3% — that gap now adds roughly $2 trillion to the US interest bill

The Congressional Budget Office's February 2026 outlook — still the working Washington baseline — projected the 10-year yield at 4.1% in 2026, 4.2% in 2027, hovering near 4.3% from 2028 to 2031, and edging to 4.4% from 2032 to 2036. The market, as of Sept 16, 2026, is paying 5.02% on the same benchmark — its highest closing level since 2007. The roughly 90bp gap is not a quarterly wobble; it has been widening for months, accelerated by the February start of the U.S.–Iran war, and shows no sign of closing.

The Committee for a Responsible Federal Budget ran the math on what even a smaller overshoot means. If yields stay just 55bp above CBO's baseline for the next decade, federal debt reaches 125% of GDP by 2036 instead of 120%, and net interest climbs to $2.5 trillion — 5.3% of GDP — versus the $2.1 trillion the CBO already penciled in. Push the gap to 80bp, and the CRFB estimates annual interest hits $2.7 trillion by the end of the decade, more than Medicare or Social Security retirement benefits on a stand-alone basis.

For years the deficit was treated as a slow-moving structural problem — a long-dated worry, not a market-moving one. That posture is breaking. Ed Yardeni, who popularized the \"bond vigilantes\" concept, spent2025 telling clients that 4–5% yields were simply a normal range. In a note last week he reversed course: \"We are starting to worry now that the 10-year US Treasury bond yield may be on the verge of breaking out above 5.00%.\" Yardeni added that the 10-year has jumped a full percentage point since the Iran war began in late February, with half of that move in the past two months alone.

Jared Bernstein, who chaired the Council of Economic Advisers under President Biden, told the same outlet he is \"getting closer\" to alarm. \"Even this nonalarmist finds alarming,\" he said, pointing to rising rates, a roughly $2 trillion annual deficit, and zero willingness from either party to address it. Maya MacGuineas, head of the CRFB, escalated the rhetoric further: \"A fiscal crisis, once unthinkable, is now a distinct possibility.\"

Treasury Secretary Scott Bessent has been the administration's primary market-calming tool. On Aug 19, 2026 the Treasury announced it would at least double the maximum size of its long-dated liquidity support buybacks — from $2 billion to at least $4 billion per operation, effective Sept 9. Three weeks later, on Sept 9, Treasury announced a single $6 billion buyback operation, triple the prior normal level. The intended signal was that the long end had a backstop.

The market heard the opposite. Long-end yields rose after the announcement. Barron's headline after the $6 billion plan was blunt: \"Yields Rise to New Highs After $6 Billion Treasury Buyback Announcement.\" The Financial Times reported the same pattern. Underlying demand was already fragile — the Aug 13 30-year bond auction cleared at a 5.216% yield, above the when-issued level, with a bid-to-cover ratio of 2.39 (below the 12-month average) and primary dealers absorbing 11.5% of issuance.

On Sept 16, 2026, the Federal Reserve under Chairman Kevin Warsh delivered a unanimous 25bp hike — its first rate increase since 2023 — lifting the target range to 3.75–4.00%. The dot plot now implies another quarter-point move by year-end (median fed funds ~4.00–4.25% in 2026) and shows the policy rate staying around4% through 2027, with no easing in the forecast horizon. The hike came with a Warsh framing of \"inflation too high,\" a posture that markets had begun pricing only after his Aug 28 Jackson Hole speech.

For the long end, a Fed that is hiking into a $2 trillion annual deficit is the wrong-footed combination. The 10-year yield topped 5% on the day of the decision — the highest closing level since 2007. Investors' Business Daily noted on the close: \"The Fed's hawkish pivot didn't fix the 10-year yield,\" pointing out that the long bond barely budged on the hawkish news, evidence the market sees the move as already priced and is more worried about the fiscal backdrop.

The 30-year fixed mortgage averaged 6.95% in the Freddie Mac survey for the week of Sept 17, 2026, up from 6.76% the prior week and 6.26% a year earlier. The CRFB estimated that each 55bp rise in mortgage rates adds nearly $200/month on a $500,00030-year mortgage and almost $130,000 in lifetime interest on a $1 million loan. That math is no longer a stress test — it is the current rate sheet, and it is what D.R. Horton and Lennar are selling into.

On the long bond itself, iShares 20+ Year Treasury Bond ETF sat near $81.26 on Sept 18, 2026 — down roughly 3.3% year-to-date on a total-return basis — even as its SEC yield climbed to about 5.23%. For banks, a steeper curve helps net interest margins in the short run but increases the embedded loss on the long-duration Treasury book carried on bank holding companies and at the regionals. For equities, the S&P 500's earnings yield now sits at roughly 10.2%, leaving the equity risk premium versus the 10-year at about 5.3% — wider than the 4.4% reading earlier in 2026 but still historically compressed.

The marginal buyer of US Treasuries is no longer the marginal buyer of US Treasuries. Foreign holdings totaled $9.29 trillion as of June 2026, with Japan at $1.116 trillion (down 2.3% from May), the UK at $939.9 billion, and mainland China at $633.4 billion. June saw a broad pullback across the top three holders, and the pattern matters because foreign central banks were once the structural absorber of every incremental Treasury supply.

On the auction side, the August 30-year sale cleared at 5.216% with a bid-to-cover of 2.39 and primary dealers taking 11.5% of the issue — both worse than 12-month averages. The August 19 20-year auction came in at a 2.53x bid-to-cover, described in coverage as \"softer.\" A failed auction remains extremely unlikely given the depth of the market, but the marginal concession required to clear a $25 billion-plus long-bond sale has been creeping higher for two quarters.

Over the next one to three quarters, three dates matter more than any Fed commentary. The November 4, 2026 quarterly refunding is when Treasury announces whether the $6 billion buyback ceiling sticks and how it sizes coupon auctions for early 2027. Any hint that buybacks are scaling back because of cost is a long-end negative. The Dec 15–16 FOMC meeting is the next opportunity for the Fed to either deliver the dot-plot-implied 25bp hike or pause; another hike locks in a4.00–4.25% policy rate through 2027. And the next 30-year auction, scheduled in the refunding quarter, will be the cleanest read on whether the 5.216% August clearing was a one-off or a new floor.

Over one to three years, the structural story takes over. CRFB's math implies net interest climbs to roughly 30% of federal revenue by 2036 if yields stay 55bp above baseline, from about 19% in 2025. Net interest per household rises from $7,900 today to $17,000 in 2036. Interest exceeds the average interest rate over the economy's growth rate (r > g) starting in 2029. These are the CBO's own numbers under a stress case, not a doomsday scenario. If the CBO's February baseline is revised in its next outlook — typically January2027 — the $2.7 trillion annual interest figure is what markets should be prepared to hear from a nonpartisan office that, until this year, was treated as the authoritative voice.

The fact pattern is unusually clean. The CBO's February baseline projects 10-year yields around 4.3% through 2031. The market is paying 5%. That ~90bp spread, if it persists, costs the federal government on the order of $2 trillion in additional interest over a decade and reshapes the trajectory of US debt from \"manageable at120% of GDP\" to \"structurally unstable at 125%.\" The experts who used to dismiss the deficit as a slow-burn issue are now publicly flagging it. The Treasury's buyback program is a liquidity tool, not a fiscal fix. And the Fed is hiking into the squeeze, not easing through it.

The investor takeaway is that the CBO-versus-market credibility gap is now the lead variable in US macro. Banks benefit in the short run from a steeper curve; long-duration Treasury holders, homebuilders, and any equity whose earnings yield is being discounted against a higher risk-free rate are the first casualties. Over a one-to-three-year horizon, the question is whether Washington treats the bond market's verdict as a binding constraint — and that question is now the most important fiscal variable in the US economy.