Fed Sustains $26.5B Monthly T-Bill Buying While Debt Machine Runs Hot
The New York Fed's Open Market Desk is running $10B/month in Reserve Management Purchases on top of ~$16.5B/month in MBS-to-T-bill reinvestments. The FOMC reaffirmed the standing directive on June 17 with no sunset language. TBAC projects ~$540B in total SOMA T-bill demand for 2026.

The FOMC's standing T-bill purchase directive has no end date, no sunset clause, and a $540B annual tab.
Key takeaways
- The New York Fed's Open Market Desk is running $10B/month in Reserve Management Purchases plus approximately $16.5B/month in MBS-to-T-bill reinvestments, totaling roughly $26.5B/month in new SOMA T-bill demand.
- The FOMC's June 17, 2026 implementation directive reaffirmed the T-bill purchase program with no sunset language; the Treasury Borrowing Advisory Committee projected total SOMA T-bill demand of approximately $540 billion for full-year 2026.
- Monthly purchase amounts are announced on a rolling basis with no multi-month commitment locked in, which means the program persists indefinitely absent a formal FOMC decision to end it.
The New York Fed's Open Market Desk confirmed $10 billion in Reserve Management Purchases (RMPs) for the period ending approximately July 13, on top of roughly $16.5 billion in separate agency-to-T-bill reinvestment purchases for the same window, per the NY Fed's operational details page. The June 17, 2026 FOMC Implementation Note reaffirmed the standing directive to increase SOMA holdings through T-bill purchases with no modification and no end date.
The combined flow is roughly $26.5 billion per month. Annualized, the Treasury Borrowing Advisory Committee's Q1 2026 charge projected approximately $360 billion in RMPs plus $180 billion in MBS reinvestments, totaling around $540 billion in SOMA T-bill demand for calendar year 2026.
How the Program Got Here
The FOMC formally ended quantitative tightening on November 30, 2025, and directed the New York Fed to begin RMPs starting December 12, 2025, at $40 billion per month. The rationale was reserve maintenance: keep bank reserves "ample" as seasonal flows and Treasury General Account dynamics drained liquidity. NY Fed President Roberto Perli confirmed reserves sat around $3 trillion when the program was running at full pace, per his March 26, 2026 remarks.
The $40B pace was always intended to step down after the April tax-season reserve drain passed. It has. The $10B/month figure represents the reduced post-April pace. What has not stepped down is the directive itself.
Each monthly amount is announced on or around the ninth business day of the month. There is no announced ceiling, no terminal date, and no mechanism that automatically winds the program down.
The NY Fed's RMP FAQ draws a clean line between this program and QE: RMPs target reserve levels, not long-duration yields. The TBAC's own analysis noted these purchases "do not significantly affect longer-maturity yields." That distinction is real and worth taking seriously. But the more important question is whether the program can stop.
The Structural Dependency the Math Reveals
The Fed and its defenders have a coherent technical argument: RMPs offset structural growth in Fed liabilities (currency in circulation, TGA balances, regulatory reserve demand from banks) and seasonal swings. They are, in this framing, reactive management of a balance sheet that grows because the economy grows.
The problem with that framing is the denominator. The U.S. Treasury is rolling over a debt load that has reached approximately $39 trillion per the Treasury Fiscal Data as of early July 2026, with Treasury bills alone accounting for roughly $6.7 trillion that must roll within twelve months. The short end of that market needs buyers who are not price-sensitive. The Fed, as a captive purchaser of T-bills, is absorbing roughly $540 billion in 2026 paper that would otherwise need to clear on price.
That absorption is the plumbing, not a side operation.
The thesis: the Fed cannot exit this program without either materially shrinking the federal debt load or triggering the reserve scarcity and money-market dislocation the program was built to prevent. The sovereign debt spiral dynamic at the long end of the curve is mirrored here at the short end by a different mechanism but the same dependency. The falsifiable version: if the Fed announces a complete cessation of both RMPs and agency reinvestments and short-term funding markets clear without repo-rate spikes for six or more consecutive months, the program is genuinely discretionary. That test has never been run.
Bitcoin's supply is fixed at 21 million. It cannot be enrolled in a reserve management program by decree, re-hypothecated into a SOMA portfolio, or expanded to offset a TGA drawdown. Every month the Fed announces another RMP tranche is another data point for that comparison. The corporate treasury trade is partly a bet on exactly this dynamic continuing.
What to Watch
Monthly RMP amounts are announced rolling, with no forward schedule locked in. The next announcement comes on or around the ninth business day of July.
Any deviation from $10B, up or down, will signal whether the Desk sees reserve conditions tightening again or whether further reduction is on the table. A move back toward zero would be the first real test of the structural-dependency thesis. Watch the repo market alongside each announcement.
Sources
- FOMC Implementation Note, June 17, 2026
- NY Fed Treasury Securities Operational Details
- NY Fed Operating Policy Statement, December 10, 2025
- NY Fed FAQs: Reserve Management and Reinvestment Purchases
- NY Fed Teller Window: Implementation of Reserve Management Purchases, March 31, 2026
- TBAC Q1 2026 Charge: Bill Purchases and the Consolidated Balance Sheet
- TBAC Report to Secretary, February 3, 2026
- NY Fed President Roberto Perli, March 26, 2026
- Treasury Fiscal Data: Debt to the Penny
Frequently Asked Questions
The Fed says no, and the technical distinction holds: RMPs target reserve levels and concentrate on short-duration T-bills, while QE targeted longer-duration securities to suppress yields across the curve. The TBAC confirmed RMPs do not significantly move longer-maturity yields. The harder question is whether the label matters when the balance sheet is still expanding and the Treasury requires the buying regardless.
No formal end date exists. Monthly amounts are set on a rolling basis, announced on or around the ninth business day of each month. The June 17 FOMC directive carries no sunset language. Absent a deliberate FOMC decision to wind the program down entirely, some level of T-bill purchasing continues by default.
Bills are short-dated instruments that roll over quickly, giving the Desk flexibility to adjust the pace without the yield-curve-control optics that longer-duration purchases would carry. The tradeoff is that the Treasury's short-end financing needs, which are substantial given the current maturity profile of federal debt, receive real-time central bank support whether or not that framing is politically comfortable.


