Treasury Is Buying Its Own Bonds. Where Is The Money Coming From?

Treasury Is Buying Its Own Bonds. Where Is The Money Coming From?

By James Broughel

On August 19, the Treasury Department announced that it will at least double the size of its “liquidity support” buybacks of long-dated government bonds. The maximum size of each operation rises from $2 billion to at least $4 billion, while the purchases target securities with 10 to 30 years left to maturity. The larger operations run from September 9 through November 4. The announcement landed in a jittery market. The 30-year Treasury yield reached 5.31 percent on August 17, a level last seen in 2007, and it fell to 5.19 percent the day of Treasury Secretary Scott Bessent’s announcement.

Most of the coverage so far has focused on the politics. A Treasury secretary visibly leaning against long-term interest rates puts pressure on Kevin Warsh, who was sworn in as Federal Reserve chair in May and gives his first Jackson Hole speech at the Kansas City Fed’s symposium, taking place on August 27 through 29. But the politics rests on an economic question the coverage has mostly skipped. When the government buys back its own bonds, where does the money come from?

Two Ways To Buy A Bond

Start with what a bond purchase does when the Federal Reserve makes it. Under quantitative easing, the Fed buys Treasury securities and pays for them by crediting the reserve accounts of banks. Reserves are deposits that banks themselves hold at the central bank, and the Fed can create them at will with a keystroke. The Fed’s own explainer resists the phrase “money printing” while conceding that the substance is not very different, noting that an increase in its Treasury holdings “is matched by a corresponding increase in reserve balances held by the banking system.” Base money, the sum of currency and those reserve balances, expands one for one with the purchase.

Treasury has no such keystroke ability. It spends out of the Treasury General Account, its checking account at the Fed, and every dollar in that account got there through taxes or borrowing.* So when Treasury repurchases a 30-year bond, the operation must be funded the way any other outlay is funded. Josh Frost, who ran Treasury’s financial markets office when the current buyback program was first designed in 2023, said as much before it launched. “We plan to treat the increase in our borrowing needs due to buybacks the same way that we treat other outlays,” he explained. A buyback, in other words, is a swap of one government liability for another.

Follow The Financing

Which liability replaces the repurchased bonds? The August 19 release does not say. Two weeks before the announcement, the department’s quarterly refunding statement said Treasury anticipates “maintaining nominal coupon and FRN auction sizes for at least the next several quarters.” Coupon securities are the notes and bonds that carry semiannual interest payments, and holding their auction sizes fixed means any extra borrowing this quarter cannot come from selling more of them. The same statement says the balance of financing will come through regular weekly bill auctions, cash management bills, and the monthly coupon auctions whose sizes are already fixed. Since the coupon amounts cannot change, the instrument that does is the bill, a Treasury security that matures in a year or less. The minutes of Treasury’s borrowing advisory committee meeting on August 4 record the same expectation among primary dealers, with changes in bill supply “likely being adequate to address any changes” in borrowing needs.

Put the pieces together and the mechanics are fairly clear. Treasury expects to borrow $739 billion in privately held marketable debt this quarter. When a buyback operation retires $4 billion of 30-year bonds, that borrowing need grows by $4 billion, and with coupon auction sizes frozen, the marginal instrument is the bill. The buyback removes a 30-year bond from the market and, at the margin, a short-term bill takes its place. Bank reserves are roughly unchanged once the dust settles, because the reserves Treasury pays out to the bond seller are pulled back in when the bill buyer pays Treasury. If money market funds buy the bills with cash they had parked overnight at the Fed, reserves can drift somewhat, but either way no new deposits are created for anyone to spend. What changes is the maturity of the debt, which is why the operation resembles the Fed’s 1961 and 2011 “twist” programs more than it resembles QE, a comparison drawn within hours of the announcement.

Bills Are Near-Money

Circumstances are further complicated by the fact that bills are near-money, meaning they are safe and liquid enough that markets treat them almost like cash. Money market funds hold them against shares that households treat like checking balances. So while a bill-financed buyback creates no new money, it does make the outstanding stock of government debt more money-like, one $4 billion operation at a time.

This is precisely the mechanism Stephen Miran and Nouriel Roubini analyzed in a 2024 Hudson Bay Capital paper that coined the term “activist Treasury issuance” for the Yellen-era version of the same practice. Whereas bills “are economically similar to the base money created by central banks,” they wrote, long-term coupons carry significant interest rate risk, so shifting issuance from coupons to bills stimulates the economy through the same channels as QE. They estimated that the Yellen Treasury’s tilt toward bills had displaced more than $800 billion of coupon issuance and delivered stimulus similar to a one percentage point cut in the federal funds rate.

The scale of the change is a matter of public record. Treasury’s own presentation to its advisory committee shows about $7.0 trillion of bills outstanding as of late July against $31.4 trillion of marketable debt, a bill share of 22.2 percent. The Treasury Borrowing Advisory Committee’s recommended range is 15 to 20 percent. Every bill-financed buyback pushes the share further above the band. Nobody knows how large the bill share can get before it threatens stability. What history shows is that a debt stock that keeps tilting toward shorter maturities is usually a sign that borrowing costs have begun to dictate policy.

The Sound Of Fiscal Dominance

Fiscal dominance describes the situation in which the government’s financing needs, rather than the inflation outlook, start to steer monetary conditions. Joseph Brusuelas, chief economist at RSM, called the new policy “what fiscal dominance looks like as the fiscal authority leans on the central bank to subordinate its goal of price stability to the government’s borrowing and political needs.” Steps like the buyback expansion, he added, “will prove to be a temporary salve to an open financial wound of our own making.” The wound he has in mind is a deficit trajectory that requires $739 billion of new borrowing in a single quarter while consumer prices rose 3.4 percent over the year through July, well above the Fed’s 2 percent target.

When the fiscal authority absorbs duration to push long-term rates down, it eases financial conditions that the Fed may be trying to keep tight, and it does so through a channel the Fed does not control. In other words, Treasury has begun conducting open market operations, the buying and selling of securities to influence financial conditions, which is the central bank’s traditional tool. Brusuelas notes that Warsh has long argued for market-determined rates free of official direction. Eight days before the new chair’s Jackson Hole debut, the fiscal authority demonstrated that it has a long-term rate tool of its own.

The Fed Is Drifting The Other Way

Treasury edging into the Fed’s business is only half of the story. Monetary policy has been edging toward fiscal policy for longer. Since 2008 the Fed has paid interest on the reserves banks hold with it, and since QE it has kept the banking system saturated with them. A reserve balance that earns a rate near what a Treasury bill pays is, from a bank’s point of view, just another overnight government asset. That changes what a central bank purchase does. When the Fed buys a bill by creating reserves, it swaps one short-term interest-bearing government liability for another, and banks feel little pressure to lend the new reserves into circulation, so the effect on the money supply is subdued. George Selgin has argued that this “floor system,” the arrangement in which interest on reserves rather than reserve scarcity sets short rates, weakened the old link between reserve creation and growth in money and lending.

Once creating reserves no longer reliably creates money, the economic content of a central bank purchase lies mostly in what gets bought. Buy long bonds and the Fed is pulling duration out of the market, which is debt management, historically Treasury’s work. Buy mortgage securities and it is steering credit toward housing, a choice with clear winners and losers that Congress never voted on. Marvin Goodfriend warned in 2011 that purchases of this kind are fiscal policy conducted by a central bank, noting that when the Fed swaps Treasuries for other assets the result is just as if the Treasury itself had borrowed from the public to buy them. The two institutions are meeting in the middle. Treasury is managing yields, which looks like monetary policy, while the Fed’s balance sheet works mainly through the composition of government debt, which looks like fiscal policy.

The Case That This Is Nothing

One argument on the other side is that alarm over the new buyback program is misplaced because the operations are too small to matter. Even doubled, a $4 billion operation is a rounding error against $31.4 trillion of marketable debt, and the up to $38 billion of liquidity support buybacks Treasury had already planned for this quarter is little more than a tenth of a percent of the total. The Fed’s QE programs ran into the trillions.

But the significance of this small operation lies in the pattern it establishes. Treasury launched buybacks in 2024 as a technical liquidity program, expanded their frequency in 2025, and has now doubled their size within days of long yields hitting a 19-year high. While each step was small, the new policy appears to be that when long-term yields rise to politically uncomfortable levels, the fiscal authority will sell more near-money and retire duration to bring yields back down.

What To Watch

Treasury said it will announce future buyback sizes at the November 4 refunding. Watch, too, whether Warsh uses Jackson Hole to draw a line between the Fed’s balance sheet and Treasury’s operations. If he says nothing, investors may reasonably conclude that the Fed accepts a fiscal hand on long-term rates, and inflation expectations will be set with that in mind. Also watch the bill share of the debt, which Treasury reports each quarter to its advisory committee. If it keeps climbing past 22 percent, the government is holding down long rates by financing itself with ever shorter maturities. The same number signals the rollover risk taken to manage the new strategy. A buyback program paid for with borrowed money leaves the debt just as large but moves its due date closer.