Bessent vs. the Bond Market: The Sleight of Hand That Gold Has Already Figured Out

The movements currently being observed in the markets appear contradictory. U.S. banks are rising sharply, yet long-term bonds are refusing to follow suit decisively; gold remains in high demand, and the dollar appears to be the most fragile variable of the bunch. These movements, however, become much more coherent when we simultaneously observe the actions of the U.S. Treasury, the Fed, the banks, and major foreign holders of U.S. debt.

The first change stems from foreign demand. China, the third-largest foreign holder of Treasuries, has reduced its holdings by approximately $60 billion since February. Data on T-bills also show significant sales by the official foreign sector in 2026. This is therefore not necessarily a simple shift from long-term to short-term maturities: some investors are reducing their exposure to U.S. debt itself. Moreover, this trend extends beyond central banks. ABP, the massive Dutch pension fund that manages more than €500 billion for approximately three million civil servants and teachers, reduced its Treasury holdings from about €29 billion at the end of 2024 to just €4.6 billion at the end of March — a reduction of nearly 84%.

This development comes at the worst possible time for Washington, as U.S. financing needs remain enormous. The more central banks and major foreign institutions stop rolling over their Treasuries — or sell them — the more the market must find new price-sensitive buyers. The problem is particularly acute for long-term maturities, where investors must agree to bear the risk of inflation, deficits, and currency depreciation for years on end. Yields must therefore rise to the level necessary to attract new buyers.

Paradoxically, the Fed finds itself on the other side of the trend. Its holdings of Treasuries rose from approximately $4.205 trillion in August 2025 to $4.542 trillion in mid-August 2026, an increase of $337 billion over the course of a year. Officially, this is not a new round of QE. Quantitative tightening has ended, and the Fed is now primarily purchasing T-bills to maintain bank reserves at a sufficiently high level, while MBS redemptions are also being reinvested in Treasuries. But the result is the same in one key respect: just as some major investors are reducing their exposure to U.S. debt, the Fed is increasing its own.

It already holds an extraordinary proportion of certain maturities. For Treasuries with ten to fifteen years remaining to maturity, its share exceeds 50 percent. The situation is therefore quite unusual: the Fed does not even need to buy more long-term debt to gradually become the holder of last resort for certain Treasuries maturities. All that is needed is for other holders to continue reducing their positions while the Fed maintains its own.

It is in this context that Scott Bessent’s new strategy must be interpreted. The Treasury is increasing its bond buybacks and is now considering using the Treasury General Account (TGA) to finance them. This account holds nearly $1 trillion in cash already raised and deposited with the Fed. Bessent therefore has considerable firepower at his disposal without needing to immediately issue new bonds to finance each buyback.

The mechanics are important. When the Treasury uses $100 billion from its TGA to repurchase Treasuries from the private sector, its account with the Fed decreases by $100 billion and bank reserves increase by the same amount. The Fed did not create an additional $100 billion: it simply converted one of its liabilities — the TGA — into another — bank reserves. Cash that had previously been tied up in the government’s account thus returns to the banking system.

For banks, this transaction is particularly attractive. A bank that sells a long-term bond to the Treasury can replace an interest-rate-sensitive asset on its balance sheet with reserves held at the Fed. It reduces its duration risk at the very moment its liquidity increases. In the wake of the 2023 banking crises — triggered in particular by accumulated losses on bond portfolios — the value of such a safety net is clear. This helps explain why bank stocks can rise sharply even as 20- or 30-year U.S. Treasury bonds have yet to see a genuine buying surge.

The banking sector does not, in fact, need to believe that Bessent will succeed in driving down long-term rates. It simply needs to understand that banks now have a better opportunity to transfer part of their duration risk to the public sector.

The bond market, however, is more demanding. Before buying 30-year bonds on a massive scale, it wants to know how many hundreds of billions will actually be mobilized, which maturities will be repurchased, and for how long. The prospect of intervention is enough to reduce the extreme risk for banks, but not yet enough to significantly alter the price of long-term debt.

The strategy becomes even more interesting when we look at what the Fed is doing at the same time. The Treasury is sharply increasing its issuance of T-bills and accumulating cash in its TGA. At the same time, the Fed is buying more short-term debt to maintain ample bank reserves. The notion that a rise in the TGA would automatically trigger a massive drain on liquidity is therefore incomplete: if increased T-bill issuance threatens to excessively reduce reserves, the Fed can offset this by purchasing T-bills on the secondary market.

A much more profound shift then emerges. The Treasury is financing the government more through short-term debt and accumulating cash. The Fed absorbs a portion of this short-term debt. The Treasury can then use its cash to remove long-term debt from the market. Banks and investors thus shed part of their duration, while bank reserves remain abundant. Legally, this is neither classic QE nor direct financing of the Treasury by the Fed, but the net result is comparable in one essential respect: an increasing share of the risk that the private sector no longer wishes to bear is shifting to public balance sheets.

That is also why the notion that Bessent could never influence the bond market deserves to be qualified. Obviously, he cannot dictate the price of the 30-year bond. However, he can alter the amount of duration the market must absorb. If investors demand much higher yields to hold long-term U.S. debt, the Treasury can respond by issuing more short-term debt and redeeming a portion of the long-term debt. It does not eliminate the debt; it changes its maturity and shifts the risk to someone else.

The problem is that this move does absolutely nothing to resolve the U.S. budget imbalance. It merely shifts it elsewhere. And that’s probably where gold becomes interesting.

If Washington succeeds in preventing part of the adjustment from taking place through a surge in long-term yields, while the Fed prevents this from translating into a scarcity of bank reserves, the market will have to look elsewhere for compensation for the risk created by the accumulation of debt. The dollar then becomes an obvious candidate.

This monetary issue ties into another major development: central banks are buying gold on a massive scale, while Washington is increasingly explicit in reminding the world that access to the Western financial system is a geopolitical weapon. Bessent’s recent remarks about countries that continue to serve as a “financial artery” for Iran are particularly revealing. In essence, he is warning them that they risk sharing Iran’s isolation. Following US sanctions against foreign banks, including the spectacular precedent set by BNP Paribas, and then the freezing of Russian assets in 2022, central banks have fully understood that a financial claim held abroad also carries jurisdictional risk.

Moving from the dollar to the euro, moreover, only solves part of the problem. It diversifies the currency and the issuer, but not necessarily the risk of being dependent on Western infrastructure and jurisdiction. Sovereign-held physical gold is different: it is no one’s liability and does not require the authorization of a foreign intermediary to exist. Central banks therefore do not need to anticipate the disappearance of the dollar to buy more gold. They simply need to want to marginally reduce their dependence on a financial system that has simultaneously become more indebted and increasingly used as a political instrument.

This is ultimately what makes current market movements far less contradictory than they may appear. Banks are rising because part of their duration exposure could be transferred to the public sector. The bond market remains cautious because it is waiting to assess the true scale of the intervention. And the gold price is rising because the market understands that stabilizing US debt could gradually shift the adjustment in yields toward the currency. Ultimately, the dollar could be the one to bear the risk of the debt.

Bessent may therefore very well succeed in calming the bond market. The real question is who will pay for that success. If long-term yields eventually stabilize while gold continues to rise and the dollar weakens, the risk will not have disappeared. It will simply have shifted.