Who will buy America’s debt?

The United States has never had much difficulty finding buyers for its debt. The Treasury market is the largest and most liquid government bond market in the world, the dollar remains the world’s principal reserve currency, and institutions ranging from American pension funds to foreign central banks have enormous structural demand for U.S. government securities.

None of that is likely to change abruptly, but there are signs of stress. The federal government needs to finance an increasingly large stock of debt at the same time as interest expense on that debt has risen dramatically. In addition, two important sources of foreign demand may be fatigued: central banks have been gradually diversifying their reserves away from the dollar, while Japanese investors, who are the largest foreign holders of U.S. Treasuries, have shown that their appetite for American bonds will be affected by relative interest rates and currency-hedging costs.

No one thinks a funding crisis is imminent, but it does portend increased scrutiny on rates, and that can mean a lot for bond investors.

For decades, federal interest expense was remarkably well behaved. From the mid-1980s through roughly 2020, the rolling twelve-month total of federal interest payments generally remained between approximately $150 billion and $250 billion. But things have changed recently. As interest rates rose after 2021, the government began refinancing an enormous stock of debt at substantially higher rates. By 2025-26, annualized interest payments had climbed above $1 trillion, which is roughly four times their pre-pandemic level!

The increase has obvious proximate effects on the federal budget; higher interest expense itself has to be financed. Unless it is offset by higher tax revenues or lower spending elsewhere, additional interest payments contribute to larger deficits, which require additional Treasury issuance, which adds to the stock of debt on which future interest must be paid.

On its own, that does not make the process inherently unsustainable. Nominal economic growth matters, as do inflation, tax receipts, spending policy and the level of interest rates. The United States also enjoys borrowing advantages that few other countries possess. But the scale has changed in a big way. The Treasury market increasingly needs to absorb substantial new issuance simply to finance the government’s existing commitments—including the cost of servicing debt accumulated in earlier years. That naturally raises the question of who will buy it.

Japan is particularly important because it remains the most reliable and largest foreign holder of U.S. Treasury securities. Japanese Treasury holdings increased from roughly $950 billion in early 2019 to approximately $1.2 trillion in 2022. They then reversed sharply, falling by about $150 billion over the following two years.

(Interestingly, Japanese holdings of U.S. corporate bonds behaved differently. Corporate holdings rose from approximately $225 billion in 2019 to roughly $350 billion in 2023 and have subsequently remained relatively stable.)

There is an important explanation for at least some of this divergence: Japanese investors do not evaluate U.S. Treasury yields in isolation. For an investor whose liabilities are denominated in yen, the relevant return may be the yield on a Treasury after hedging the dollar exposure back into yen. As the Federal Reserve raised rates aggressively while the Bank of Japan maintained much lower rates, those hedging costs became unusually expensive.

The apparent yield advantage of Treasuries could therefore disappear—or even become a disadvantage—for a hedged Japanese investor.

That helps explain why Japan’s declining Treasury holdings in 2022-24 should not necessarily be interpreted as a loss of confidence in the United States. It may have been something considerably more mundane: the economics of owning hedged Treasuries temporarily became unattractive.

There is also evidence for that interpretation in what happened next. Japanese Treasury holdings recovered to roughly $1.15 trillion by May 2026, retracing a meaningful portion of their earlier decline. So, one can’t really say that Japan is abandoning the Treasury market.

However, the episode demonstrates something potentially more important. Japanese demand is price sensitive. If Japanese government bond yields become more attractive, Bank of Japan policy changes further, or the cost of hedging U.S. assets rises again, some capital that might otherwise finance the United States might be deployed elsewhere.

There is another, much slower change taking place: at the beginning of this century, the U.S. dollar represented roughly 72% of disclosed global foreign-exchange reserves. Today its share is approximately 57-58%.

It would be easy to overstate what this means; the dollar remains overwhelmingly the world’s most important reserve currency. The euro has not displaced it: its share of reserves has remained broadly in the low-to-mid-20% range for much of the period. Nor has another individual currency emerged as an obvious replacement. What has happened instead is that central banks have diversified gradually across a collection of alternatives. The share represented by currencies other than the dollar and euro has risen from roughly 12% in 2000 to around 20% today.

That is quite different from the frequently discussed idea of sudden “de-dollarization.” There has been no wholesale flight from dollar assets. The decline has instead taken place over a quarter century and through numerous changes in U.S. administrations, monetary policies, exchange rates and economic cycles.

In some ways, that makes the trend more interesting. A gradual decline from 72% to 57-58% does not require a crisis or a new dominant reserve currency. Central banks simply need to allocate a little less of each incremental reserve dollar to U.S. assets.

For most of the past 25 years, that has been relatively easy for the Treasury market to absorb, but it could become more consequential when Treasury financing requirements are simultaneously becoming much larger.

Taken separately, none of these developments is particularly alarming.

Federal interest expense is high, but the United States has an enormous economy and an extraordinary capacity to borrow. Japan reduced its Treasury holdings, but has subsequently increased them again. And the dollar has lost reserve share, but remains comfortably dominant.

The interesting part is the interaction among the three: the U.S. government is asking investors to absorb an increasing supply of securities. Foreign reserve managers appear to have a gradually declining structural preference for dollar assets. And Japan has provided a real-world demonstration that even one of the Treasury market’s largest and most reliable foreign buyers will reduce its holdings when relative returns become unattractive.

Someone will still own the securities, but at what price? Markets generally solve excess supply through price adjustment. In the Treasury market, a lower bond price means a higher yield. Higher yields, in turn, make Treasuries more attractive to pension funds, insurers, households, banks, foreign investors and other potential buyers.

That mechanism is precisely why a Treasury “buyers’ strike” is probably the wrong framework for thinking about the issue. There is almost certainly a yield at which sufficient demand emerges. The real biting question is whether that yield is higher than investors, borrowers—or the federal government—have become accustomed to. And that could matter well beyond the Treasury market.

Treasury yields occupy a special place in financial markets. They are not merely the government’s borrowing cost; they are also a reference rate against which enormous portions of the financial system are priced. Mortgage rates, corporate bonds, private credit, commercial real estate financing and many equity valuation models are all influenced, directly or indirectly, by Treasury yields. A structurally higher yield required to clear an increasingly large Treasury market could therefore propagate across asset classes.

That does not necessarily mean poor investment returns. Higher starting yields improve prospective returns for bond investors, and strong economic growth could simultaneously support both higher rates and corporate earnings. Markets also adjust: capital moves toward securities offering better compensation. But it could mean that the extremely low interest-rate environment that prevailed for much of the period following the Global Financial Crisis was truly atypical.

For portfolios, the implication is not that long-duration Treasuries should be avoided. If economic growth weakens materially or inflation declines, longer-term government bonds could still perform very well. The recent recovery in Japanese Treasury holdings also argues against extrapolating the most pessimistic interpretation of foreign demand. Rather, investors may need greater compensation for assuming duration risk when fiscal borrowing requirements are large and some historically important sources of demand are becoming more price sensitive.

There is a tendency for discussions about U.S. government debt to migrate quickly toward extremes: either deficits do not matter because the United States issues the world’s reserve currency, or mounting debt inevitably culminates in a fiscal crisis. Both are too stylized to be particularly useful. Another outcome may be less dramatic: the United States continues to borrow. Treasuries continue to serve as the world’s principal safe and liquid asset. The dollar remains the dominant reserve currency. Foreign investors continue to participate. But increasingly, they may need to be paid to do so.

If that is right, the most important consequence of America’s fiscal trajectory may not be an inability to finance the debt. It may simply be a higher equilibrium cost of capital—first for the federal government, and ultimately for everyone else.

Important disclosures