30-Year Treasury Yield Hits Its Highest Since 2007, Fed on Hold

The Number Nobody Put in a Headline on Monday

Stocks in New York fell half a percent on Monday. The S&P 500 closed at 7,747, the Dow at 53,460, the Nasdaq at 26,647. For a market that had climbed 6% in twelve sessions to a record, that was an ordinary breather, and nobody called it news.

The same afternoon, the yield on the 30-year U.S. Treasury bond touched 5.31% intraday. The last time that yield was seen was June 2007 — nineteen years ago, before the global financial crisis, before zero rates, before the pandemic, before the inflation wave.

Here is what makes it interesting. The Fed's policy rate has not moved since June. The market is not pricing a hike this year. So this selloff is not a bet on what the Fed will do. Long-term yields are moving independently of the short-term policy rate, and they have been doing it for weeks.

Three of those four numbers are talking to each other. That is the story.

The Two Ends of the Curve Are Not the Same Thing

A government bond's yield is not set by one thing. Short-dated yields are largely set by the central bank; a two-year note is essentially the priced expectation of what the Fed will do over the next two years. At the long end, the central bank's grip weakens. An investor buying a 30-year bond is not pricing the next FOMC meeting — they are pricing thirty years of inflation, debt and uncertainty.

The gap between the two is the term premium: the extra yield an investor demands for tying up money for a long time.

It is the second one that is moving now. And four separate forces are pushing it higher at once.

Force One: The Deficit Grew by $200 Billion

In its August report, the Congressional Budget Office raised its fiscal 2026 deficit projection from $1.9 trillion to $2.1 trillion. The $200 billion increase came from lost revenue, not new spending.

According to the CBO, the main line item is tariffs. A Supreme Court ruling on February 20 invalidated the tariff program imposed under IEEPA authority, erasing roughly $250 billion of projected customs-duty revenue for the year. Income and payroll taxes came in $75 billion above forecast and other lines $25 billion below, leaving a net gap of about $200 billion. The administration is trying to fill the hole using other authorities under the Trade Act of 1974, but the CBO expects only "a substantial share" of the loss to be recovered.

The distinction worth drawing here: this is not a statement or an expression of intent. It is a court ruling already in force, and its consequences are already in the budget numbers. By the count of Maya MacGuineas, president of the Committee for a Responsible Federal Budget, the United States has borrowed $1.8 trillion this fiscal year — nearly $6 billion a day.

A bigger deficit means the Treasury must sell more bonds. More bonds means more supply hitting the same pool of buyers. Prices falling — which is to say yields rising — is the arithmetic consequence.

Force Two: A Fed Without Guidance

Fed Chair Kevin Warsh has held to a policy of not offering forward guidance since taking office. His stated view is that moves in market yields can substitute for policy action. A Goldman Sachs note summarized the result: the absence of guidance has pushed volatility out along the curve.

Mark Cabana of BofA Global Research was blunter: "There is literally a price to be paid for the lack of guidance that Warsh seems so set on."

That price shows up as an uncertainty premium embedded in the yield. An investor who cannot read the Fed wants to buy the long bond cheaper. Minutes from the July meeting are released Wednesday; that is the release the market is actually waiting for this week.

Force Three: The Treasury's New Competitor Is a Data Center

This is the least-told part of the story. Amazon, Alphabet, Meta and Oracle sold roughly $194 billion of bonds through July 7 this year. The figure for all of 2025 was about $108 billion — a 79% increase. Goldman Sachs expects the five largest cloud companies to reach roughly $250 billion in 2026 and $400 billion in 2027.

That money is going into data centers, chips and grid connections. But from the bond market's point of view, what the company wants the money for is irrelevant. What matters is that two hands are reaching into the same investor's wallet at the same time.

The buy side is not growing. If anything it is shrinking: traditional long-duration buyers such as pension funds and insurers have spent years cutting the weight of long bonds in their portfolios. Their place is being taken by price-sensitive private investors. A price-sensitive buyer asks for a discount.

The evidence that discounts are being asked for is in the numbers. Cover ratios on these four companies' bond sales — orders received per bond sold — fell from roughly 5 times in February to below 2 times in July.

The new-issue concession has widened too: the extra yield issuers must offer over their existing bonds to bring investors in rose from 2.25 basis points in 2025 to 12 basis points in 2026. In the words of Colby Stilson of Brown Advisory, "We're already seeing fatigue within credit markets in supporting this massive debt issuance."

The $510 Billion Ceiling

Amanda Lynam, head of credit strategy research at Goldman Sachs, drew a distinction between two numbers in analysis published on August 5. Amazon, Google and Microsoft could collectively add roughly $2 trillion of debt to their balance sheets and still keep investment-grade ratings. But the U.S. investment-grade bond market can comfortably absorb only about $510 billion of it before investors start pushing back.

That is the heart of the mechanism. A company's debt can hit a wall in the market without its credit ever deteriorating. When it does, the issuer offers a higher yield. A higher corporate yield forces the government bond of the same maturity to offer more as well. That is how the chain closes.

Duration: Same News, Eight Times the Damage

One calculation is enough to explain why the long end moves so violently.

The reason is this: a bond's price is the present value of the cash flows you are owed. A two-year note has four coupon payments and a near-term principal left; when the discount rate moves, the present value of those flows barely changes. A 30-year bond has sixty coupons and a principal payment three decades out; the smallest change in the discount rate hits those distant flows far harder. This sensitivity is called duration, and at the long end it is roughly eight times what it is at the short end.

Which is why "government bonds are safe" is an incomplete sentence. There is no default risk in a Treasury; there is price risk, and it grows quickly with maturity. Interest Rates and Bonds covers the underlying relationship in more detail.

Force Four, and a Selloff That Spread Overnight

The fourth force is oil. Brent holding above $90 on U.S.-Iran tension and uncertainty over the Strait of Hormuz feeds the worry that inflation stays sticky and the Fed's hands stay tied.

And this selloff is not an American phenomenon. As of Tuesday morning, long-dated government yields are at multi-decade highs across the world.

Japan's 30-year yield, meanwhile, is close to an all-time record. Four countries with very different politics, budgets and central banks are repricing the same maturity at the same time — which says the cause is structural rather than local.

Why Equities Did Not Move

The most instructive data point in this piece may be a non-move. Stocks fell half a percent on Monday — on a day that produced the highest long-term yield in nineteen years, with indices just below records.

Sources do not tie that decline to the bond market. The agenda that day was anticipation of retail earnings, and most market commentary described the move as a breather with no single cause. SanDisk rose 9% the same session while Meta fell 3.5%. In other words, sector news drove equities, not the long bond.

That is worth reading as a finding: the market does not currently regard the rise in long-term yields as binding on equity valuations. Yet the same discount rate operates in both places. Most of a technology company's present value comes from cash flows a decade out; when the discount rate used in a valuation rises, those distant flows lose value exactly as the principal of a 30-year bond does.

The index most sensitive to the discount rate is the Nasdaq 100, whose weight is concentrated in companies expecting the bulk of their earnings in distant years.

What Is Left

There is no single crisis here. A court ruling widened the deficit, a Fed chair chose not to give guidance, four technology companies decided to fund the largest capital-spending program in history with debt, and a strait closed and left oil above $90. None of these caused the others. All of them are pressing on the same maturity at the same time.

It is no accident that neither chart in this piece shows the bond market. The repricing happened in a market most equity investors never look at. The chart above shows what the equity side did over the past three months; over those same three months the 30-year yield climbed to a 19-year high. That is not a forecast — it is a description of the prices so far.

Yield and index levels in this piece come from CNBC, Bloomberg, Anadolu Agency and Motley Fool market reports dated August 17-18. Budget figures are from the Congressional Budget Office's August report as reported by Fortune, and auction details from the Committee for a Responsible Federal Budget's analysis. Corporate bond issuance and demand data come from a Reuters compilation; the $510 billion ceiling calculation comes from Goldman Sachs credit strategy analysis published August 5. Bond price calculations are our own, using the standard present-value formula, and are approximate.