Bessent's TGA Buyback Gambit Reveals Thin Debt-Refinancing Margins

The Cash That Isn’t Really Cash

The US Treasury has nearly a trillion dollars sitting in its General Account at the Federal Reserve. It has been sitting there for months—$950 billion, to be precise, roughly $350 to $400 billion above where the Biden Treasury considered appropriate. It is funded with tax receipts that came in. It is, in the parlance of modern government finance, money that was already yours.

Treasury Secretary Scott Bessent may now decide to spend it on something that looks, to market observers, a great deal like money creation.

The math is simple but consequential. Last week, the Treasury surprised everyone by doubling its off-the-run bond buyback floor from $2 billion to at least $4 billion per quarter. Bessent told CNBC the minimum “could be even larger than the new higher minimum.” When asked about funding, he offered a single phrase: a “Treasury Twist.” That is a direct lift from 2011, when the Fed and Treasury coordinated to swap short-dated paper for longer tenors in a quiet attempt to flatten the yield curve. This time, the coordination would not involve the Fed at all. It would involve the Treasury’s own checking account.

What the TGA represents is a choice about where leverage lives.

Who Loses When the Account Gets Drained

The first order of business: understand the mechanics. The Treasury uses the TGA to smooth cash flows—paying bills when receipts dip, rebuilding when they surge. A drop to zero would not trigger a default. The next debt ceiling does not arrive until winter, possibly spring. But it would mean the government would be carrying far less operating cushion. In a normal administration, the TGA is described as covering “a week ahead of cash needs.” The current level is not described that way.

The bigger loss falls on the market.

Buybacks target off-the-run securities—treasuries that have left the primary market and trade in quieter conditions. Their yields tend to be higher than on-the-run equivalents, which creates an inefficiency the Treasury can exploit. But every dollar spent on buybacks is a dollar not held in reserve. Every off-the-run trade reduces secondary-market liquidity precisely where liquidity is already thin. That is what spooked some analysts after the initial announcement: bonds retreated from their early rally. The market’s answer was not skepticism about the idea. It was skepticism about the budget.

The question has always been whether $4 billion a quarter buys enough to move the long end. With the TGA, the number becomes less abstract. $10 billion? $50 billion? Even $200 billion would be meaningful, and it would not require issuing a single new security to finance it.

Who Wins When the Twist Returns

The immediate winner is whoever holds long-duration treasuries right now. A credible TGA-funded buyback program puts a bid under the market that did not exist before. Off-the-run spreads compress. Yield volatility, especially on the 30-year, softens. That is exactly the kind of quiet equilibrium Bessent said he was after when he talked about keeping the market from “trading the headlines during a quiet period in a thin market.”

The longer-term winner is the Treasury itself—if it can sell back into the demand it creates. Here is the trick: using the TGA does not require issuing short-term bills. That makes the operation look less like refinancing and more like spending. If the market reads it that way, it will price accordingly. Yields could fall on the long end and rise on the short end, exactly the inversion a twist is supposed to engineer. But the signal matters as much as the mechanics. If investors conclude the Treasury is using its war chest because regular issuance is no longer enough to manage the curve, the premium on risk rises everywhere.

The third winner is the president. Whatever else the TGA drawdown represents, it is evidence that someone is trying to fix a broken market without asking the Federal Reserve for help. That framing lets Bessent push back against the criticism that the Treasury abandoned its old discipline of being “regular and predictable.” He has a point. The Aug. 19 announcement came nearly three weeks before the first operation on Sept. 9, and the quarter-wide plan was laid out at the same time. The auction schedule itself did not change. But the real question is not whether the Treasury followed its own rules. It is whether the rules still serve the country.

What Happens Next

Three scenarios are plausible.

In the base case, the Treasury taps a modest portion of the TGA—say $30 to $50 billion—to supplement the buyback program through year-end. Off-the-run spreads narrow. The 10-year yield dips a few basis points. Bessent announces it as a technical adjustment and moves on to tariff negotiations, which he says will replace refunded revenue and improve the deficit.

In the aggressive case, the Treasury spends closer to $150 billion. That is large enough to reshape the secondary market and small enough to avoid looking like outright monetization. The yield curve flattens noticeably. The Fed, which had been told it would not be involved, finds itself reacting rather than acting. Market participants who were worried about a fiscal corner begin to believe the corner has been avoided—for now.

In the cautionary case, the Treasury does not touch the TGA at all. It funds the expanded buybacks through short-term issuance, as most market participants expected. The program proceeds as announced. Yields move little. The lesson is that the market does not trust the tool as much as it trusts the signal, and the signal was wrong.

The most likely outcome sits between the first two. The TGA is discretionary. The Treasury has explicitly said it is available. Bessent has signaled his appetite for action. A partial drawdown, announced alongside the first few operations, is the move that preserves optionality while demonstrating capability. The question is how much capability the market believes it can muster before demand for the paper exceeds whatever firepower is left in the account.

The Real Story Is the Deficit

Bessent said he expects progress on the deficit once tariff revenue replaces court-mandated refunds. That is an assumption, not a fact. Tariff policy is being challenged in court. Revenue projections depend on trade volumes that shift with the global cycle. The Treasury officials would not commit to a timeline for rebuilding the TGA even if some of it is spent.

What is clear is that the existence of a $950 billion buffer changes the calculus of every decision that follows. It gives the Treasury room to act without immediately raising new debt. It also gives the Treasury a reason to act, because the alternative—letting the account sit idle while off-the-run spreads blow out—looks like surrender. The question is whether this is a sustainable way to manage a debt market or merely a clever postponement of the problem that produced the buyback program in the first place.

The answer will show up in the spreads, the size of the draws, and the speed with which the account refills. It will not show up in any press release.