The $38 Trillion Treasury Market and the Search for Its Marginal Buyer

The size of the U.S. national debt gets most of the attention. The harder question is how a debt stock approaching $38 trillion is continuously financed.

Every year, Treasury securities mature and must be refinanced. Federal deficits require additional borrowing on top of those maturities. Interest expense adds to future spending, and higher market rates gradually enter the government’s borrowing cost as older securities mature.

Together, these forces create a continuous financing cycle. Existing debt matures and is refinanced. Deficits create new issuance. That issuance adds to the debt stock, while its interest cost becomes part of future federal spending and financing needs.

Ultimately, every Treasury security issued through that process has to land on a balance sheet.

That brings us to the central question:

Who absorbs the next dollar of Treasury issuance, at what maturity, and at what yield?

The answer connects federal deficits with money-market funds, banks, foreign reserve managers, primary dealers, the Federal Reserve, repo markets, collateral rules and eventually the regulatory changes reshaping parts of the financial system.

Understanding that connection begins with the debt itself.

The scale of the financing system

Federal debt can be measured several ways. For market analysis, debt held by the public provides a useful starting point because it captures Treasury obligations held outside federal government accounts.

The Congressional Budget Office projects debt held by the public at approximately 101% of GDP in 2026. Under its February 2026 baseline, that ratio reaches approximately 120% of GDP by 2036. CBO also projects a federal deficit of approximately $1.9 trillion in 2026, equal to about 5.8% of GDP.

Those figures describe a financing system operating at a much larger scale relative to the economy than it did at the beginning of the century.

Inside that system sits the marketable Treasury market.

Federal Reserve Financial Accounts data put marketable Treasury liabilities at approximately $30.6 trillion in the first quarter of 2026. Roughly $6.8 trillion consisted of Treasury bills, while approximately $23.8 trillion consisted of notes, bonds and TIPS.

That maturity composition determines which investors can comfortably absorb different portions of the government’s financing requirement.

Treasury bills mature quickly and expose holders to little duration risk. Longer notes and bonds commit investors across years or decades of changing inflation, interest rates and fiscal conditions.

A three-month bill and a thirty-year bond therefore represent different financing propositions even though both are obligations of the same government.

The Treasury has to find demand across the maturity spectrum.

A large debt stock creates a large refinancing machine

The federal government continuously refinances securities as they mature, while deficits create additional borrowing needs. Treasury’s August 2026 quarterly refunding shows how these two flows operate together.

Treasury announced $125 billion of new 3-year, 10-year and 30-year securities. Approximately $96.3 billion replaced privately held notes and bonds reaching maturity, leaving approximately $28.7 billion of new cash raised

This is the basic rhythm of Treasury financing. Existing debt reaches maturity and new securities replace it, while federal deficits add incremental borrowing on top of the refinancing requirement. As the outstanding debt stock grows, the volume passing through this process also grows.

The Treasury market therefore has to absorb both the replacement of existing claims and the additional claims created by new borrowing. As the outstanding debt stock grows, so does the recurring financing requirement before future deficits add another dollar of debt.

The financing flow is already enormous

Treasury expects $739 billion of privately held net marketable borrowing from July through September 2026 and another $628 billion from October through December, bringing expected net borrowing during the second half of 2026 to approximately $1.37 trillion.

The scale becomes easier to understand when placed beside entire national economies. Using 2024 nominal GDP, Türkiye ranked roughly 17th in the world with about $1.36 trillion of annual economic output, while Saudi Arabia ranked roughly 18th with about $1.24 trillion. Treasury therefore expects to raise, in six months, an amount of net marketable financing roughly equivalent to the annual economic output of one of the world’s twenty largest economies.

And $1.37 trillion represents net borrowing. Gross Treasury issuance is considerably larger because securities reaching maturity must also be refinanced.

The financing requirement therefore combines the replacement of existing debt with the additional borrowing created by fiscal deficits. Absorbing that supply requires balance-sheet capacity across foreign reserve managers, money-market funds, banks, pension funds, insurers, asset managers, hedge funds and other investors, with primary dealers intermediating auctions and secondary-market activity.

As the Treasury market grows, the capacity and composition of this buyer base become increasingly important. Treasury must continuously place enormous quantities of securities across investors with different mandates, maturity preferences, funding constraints and sensitivities to yield.

Who absorbs the next dollar?

Treasury ownership data tell us where decades of accumulated securities currently reside. Financing conditions depend on which balance sheets are willing to expand next.

That difference separates the stock of Treasury ownership from the flow of Treasury demand. A foreign government, pension fund or bank can hold hundreds of billions of dollars of Treasuries without adding another dollar to its portfolio. Meanwhile, an investor with a much smaller existing position can become an important source of demand by rapidly increasing its purchases.

This becomes consequential when Treasury brings additional securities to market. Existing demand absorbs some of the issuance, while the remaining supply must attract investors comparing Treasuries with cash, corporate bonds, equities, mortgages, foreign sovereign debt and other assets.

Yield helps determine where that additional demand appears. As Treasury prices adjust, the expected return changes until enough investors are willing to provide the balance-sheet capacity necessary for the market to clear.

The financing question therefore reaches beyond who owns Treasuries today. It increasingly depends on who is adding Treasuries tomorrow, how much capacity they have, which maturities they want, and what yield attracts them.

That is where the marginal buyer appears.

Yield travels back into the federal budget

The yield Treasury pays to attract buyers eventually becomes part of the federal government’s financing cost. The effect arrives gradually because the debt stock reprices over time.

A Treasury security issued several years ago continues paying its existing coupon until maturity. When that security matures, Treasury may need to refinance it at the rates available in the market at that time. New borrowing enters at current rates as well. The maturity structure of the federal debt therefore determines how quickly prevailing yields work their way into the government’s average borrowing cost.

CBO projects federal net interest outlays of more than $1 trillion in 2026, equivalent to 3.3% of GDP. By 2036, net interest expense is projected to reach approximately $2.1 trillion, or 4.6% of GDP. CBO estimates that net interest costs will grow at an average annual rate of about 7.5% over the decade, driven by the combination of a larger debt stock and higher average interest rates.

By 2036, CBO projects net interest expense will account for nearly one-fifth of all federal spending. At approximately $2.1 trillion, annual interest expense would also nearly equal the federal government’s entire projected discretionary budget of roughly $2.2 trillion.

The comparison with defense spending makes the shift visible even earlier. Net interest expense surpassed defense spending in 2024, and CBO’s projections show the gap widening over the following decade. Visual Capitalist has already built a strong historical visualization of this crossover using White House historical data and CBO projections.

As older Treasury securities continue to mature, prevailing interest rates gradually enter the government’s average borrowing cost. The consequences accumulate across a debt stock that CBO projects will rise from 101% of GDP in 2026 to 120% in 2036. The rate at which that repricing occurs depends heavily on when the debt matures and how Treasury chooses to finance the next round of borrowing.

That brings us back to the marginal buyer. The yield required to clear Treasury issuance today can eventually become part of the fiscal burden that determines how much Treasury must finance tomorrow.

Maturity determines where financing pressure appears

Treasury has considerable flexibility in distributing borrowing across the yield curve, and each maturity shifts risk to a different part of the financial system.

Bills provide short-term financing and require frequent refinancing, so changes in short-term interest rates pass into Treasury’s borrowing costs relatively quickly. Notes and bonds lock in financing costs for longer periods, but investors taking that duration risk generally require greater compensation.

Demand also varies by maturity. Money-market funds naturally concentrate at the short end, while pensions and insurers have stronger reasons to hold long-duration assets. Foreign reserve managers may allocate across the curve, banks respond to liquidity, capital, regulation and relative returns, and hedge funds connect different parts of the market through Treasury cash-futures and repo trades.

Treasury debt management therefore involves distributing an enormous financing requirement across maturities, investor bases and forms of balance-sheet capacity. Treasury’s recent language makes this increasingly visible.

Treasury is watching structural demand

In its August 2026 Quarterly Refunding Statement, Treasury discussed growing private-sector demand for Treasury bills and said future issuance decisions would consider trends in structural demand, alongside the costs and risks of different issuance profiles.

That phrase captures an important part of the Treasury market. Demand does not come only from investors deciding that a particular yield is attractive. It also comes from the functions Treasury securities perform within the financial system.

Money-market funds need short-duration assets. Banks need liquidity. Foreign central banks hold reserve assets. Dealers use Treasuries for market making and financing. Repo markets rely on them as collateral. Financial institutions also hold government securities for regulatory, liquidity and operational purposes.

These functions create a layer of Treasury demand embedded in the architecture of the financial system, where demand can arise from what Treasuries enable institutions to do, not simply from the yield they offer.

At its August meeting, the median primary-dealer forecast implied roughly a $1.45 trillion financing shortfall across fiscal years 2027 and 2028 if current coupon auction sizes and privately held bill supply remained unchanged. Dealers generally expected Treasury eventually to increase nominal coupon auction sizes during 2027.

The problem is therefore larger than simply raising a given amount of money. Treasury must determine which maturities have sufficient capacity, which investors can absorb additional supply, how durable that demand is, and what price the government must pay to access it.

The buyer base is changing

Foreign investors remain a major source of Treasury demand, but the composition of that demand matters as much as the headline holdings. Treasury International Capital data for June 2026 show Japan holding about $1.12 trillion of Treasury securities, the United Kingdom about $940 billion, and China about $633 billion.

Those positions are large, but future financing depends on whether they are growing, shrinking or simply being maintained. China illustrates the difference. Its Treasury holdings have fallen from a peak of roughly $1.32 trillion in 2013 to about $633 billion in June 2026. A large existing position therefore does not necessarily translate into additional demand for new issuance.

The same distinction appears in the domestic market. Federal Reserve Financial Accounts data show the rest of the world holding roughly $9.3 trillion of Treasury securities in the first quarter of 2026, while money-market funds held roughly $3.4 trillion. About $2.5 trillion of the money-fund total was concentrated in Treasury bills, which helps explain why that sector matters so much at the short end of the curve.

Money funds can also change their Treasury exposure materially from one quarter to the next as cash balances, relative yields and funding conditions move. That makes the marginal buyer dynamic. The investor group absorbing the next increment of Treasury supply can shift even when the largest existing holders barely change.

The Treasury financing problem therefore cannot be reduced to the size of the debt alone. A growing stock of government liabilities must continually find balance-sheet capacity across investors with different mandates, maturity preferences and sensitivities to yield.

The next question is whether that capacity is fixed, or whether the financial system itself can create new reasons to hold Treasury debt.

As Treasury financing needs grow, the constraint increasingly becomes the financial system’s capacity and willingness to absorb that debt at an acceptable cost.

What I’m tracking

The architecture of Treasury demand is not static. Regulation can change which institutions hold government debt, market infrastructure can change how that debt is financed and used, and new forms of dollar liabilities can create new channels through which Treasury securities enter the financial system.

I’m tracking those changes as they happen through Architecture, a live research build following Treasury issuance and ownership, stablecoin reserve rails, GENIUS implementation, Treasury clearing, institutional conversion pathways and sovereign reserve assets.

The research separates what is legally permitted, what is operationally available and what can actually be observed in the data.

Sources

U.S. Department of the Treasury, Monthly Statement of the Public Debt

U.S. Department of the Treasury, Quarterly Refunding Statements and Treasury Borrowing Advisory Committee Materials

U.S. Department of the Treasury, Treasury International Capital System

Federal Reserve Board, Financial Accounts of the United States

Federal Reserve Bank of New York, SOMA Holdings and Treasury Market Operations

International Monetary Fund, Global Financial Stability Report and Fiscal Monitor

Securities and Exchange Commission, Treasury Clearing Rules and Releases

Primary-source data and regulatory materials are incorporated throughout the analysis. Figures are rounded where appropriate for readability.

Kirandeep Kaur

Part 2: How regulation can create new sources of Treasury demand