NextFin News - The Federal Reserve's bond portfolio manager signaled that future Treasury bill purchases will be guided entirely by market conditions rather than a preset schedule, as the central bank judges that banking-system reserves remain comfortably ample. Roberto Perli, manager of the System Open Market Account at the New York Fed, said at the 2026 U.S. Treasury Market Conference on Tuesday that the Open Market Trading Desk's reserve management purchases can be adjusted up or down — or paused altogether — depending on money market conditions.
The message lands as the Desk has already done exactly that. It suspended reserve management purchases for the operating period through September 14 — the first full pause since the program launched in December 2025 — and extended the hiatus through the period ending October 14. Bank reserves stood at $3.04 trillion as of September 9, up from $2.85 trillion at the end of last year and above the year-to-date average of $3.01 trillion. That is the kind of buffer that lets the Fed treat balance sheet management as calibration rather than crisis response.
The Pause Is a Signal, Not a Policy Shift
The most important thing about the pause is what it is not. It is not a return to quantitative tightening, and it is not a reversal of the ample-reserves framework the Federal Open Market Committee adopted when it stopped shrinking the balance sheet on December 10, 2025. Reserve management purchases were never meant to ease financial conditions or push down long-term yields the way the post-2008 asset-buying programs did. Their purpose is narrower and more technical: keep reserve balances inside the range the FOMC considers ample, so the federal funds rate stays under firm control.
That distinction matters because it explains why the Desk can stop buying one month and restart the next without sending a monetary policy signal. Perli has framed the program in those terms from the start, and the FOMC itself made the option explicit at its June meeting, stating that temporary pauses could occur if money market conditions warrant. In July remarks at a New York Fed-hosted conference, the SOMA manager put the flexibility in plain terms:
"The Desk can adjust amounts up or down for any given month, depending on money market conditions," Roberto Perli, manager of the System Open Market Account, said in July.
The pace history shows the flexibility in action. Purchases began at $40 billion a month in December 2025, were cut to $25 billion in April, reduced again to $10 billion in May, and then halted. For the period through mid-October, the Desk still plans roughly $15.6 billion in reinvestment purchases — the recycling of principal payments from agency mortgage-backed securities into Treasury bills — even as the incremental reserve management line sits at zero. The balance sheet is therefore not shrinking; one component of its growth has simply been put on hold.
The total asset side of the Fed's balance sheet stood at $6.74 trillion as of September 9, according to the Board of Governors' weekly H.4.1 release, little changed from the roughly $6.7 trillion level reported in the July Monetary Policy Report. Reserves, the largest liability, have held between $3.0 trillion and $3.1 trillion since early this year. Usage of the overnight reverse repo facility has remained near zero on most days, meaning the marginal supplier of short-term funding has shifted from the Fed's facility to private money markets. Against that backdrop, the pause reads as the Desk saying the plumbing is working as designed — not that the policy stance has changed.
It was also a mild surprise. Market participants had broadly expected monthly purchases to remain around $10 billion, and the cut to zero was the first full suspension since the program began. The surprise is not a policy one; it is a signal that the Desk's read of funding conditions is looser than the street's.
Why the Fed Believes Reserves Are Still Ample
The Desk's confidence rests on a stress test the system just passed. In April, tax-season flows into the Treasury General Account drained liquidity from the banking system at precisely the pace the FOMC had feared when it launched the program. The TGA peaked at $1.04 trillion, and reserves fell by roughly $300 billion in the weeks around April 15. Repo rates ticked above the standing repo facility rate on the tax date but reversed the next day, returning to levels near the interest rate on reserve balances. Prior reserve management purchases ensured that the April trough in reserves stayed above the levels seen late last year, when the FOMC determined reserves had declined from abundant to ample.
In other words, the framework was tested under the exact conditions it was built for — a large, rapid, predictable reserve drain — and it held. Perli noted in July that there was "no evidence of a material change in banks' demand for reserves" despite the volatility. That is the empirical foundation for the pause: if reserves absorb a $300 billion seasonal swing without money market dysfunction, the marginal benefit of adding more through bill purchases is low.
The definition of "ample" itself is operational, not ideological. As Perli explained in May, ample refers to the range of reserves that makes the federal funds rate only modestly sensitive to changes in reserves. It is the flat part of the reserves demand curve, where an extra dollar of reserves does little and a missing dollar does little either. The Desk's job is to keep the system on that flat portion — which is why the tool is a dial, adjusted month by month, rather than a commitment.
The Second-Order Read: Bills, Money Market Funds, and the Supply That Must Be Absorbed
Here is where the pause carries a market implication that goes beyond the plumbing. Reserve management purchases have been conducted entirely in Treasury bills. When the Fed buys bills, it removes short-duration government supply from the private sector; when it stops, that supply must be absorbed somewhere else. The natural buyers are money market funds — government MMFs, which can hold only government or agency debt and repo, and prime MMFs, which can also hold commercial paper and certificates of deposit.
If the Fed steps back from the bill market while Treasury issuance continues, the private sector must take more bills onto its balance sheet. That can push bill yields slightly higher relative to repo rates and the interest on reserve balances, and it can shift money market fund portfolios away from overnight instruments toward slightly longer paper. It is a subtle reallocation, not a dislocation — but it is the kind of transmission channel that matters at the margin when funding conditions are already loose. Think of the Fed's bill buying as a price-insensitive bid at the short end of the curve; when that bid disappears, the market-clearing yield has to do a little more work.
The scale involved is not trivial. In its presentation to the Treasury Borrowing Advisory Committee earlier this year, the Treasury estimated that reserve management purchases to maintain ample reserves would total roughly $360 billion in 2026, with agency MBS paydowns adding roughly $180 billion more — about $540 billion of Treasury bill demand flowing into the SOMA portfolio over the year. A pause through mid-October removes a chunk of that demand right as the Treasury ramps up net bill issuance. Perli flagged exactly this dynamic in July: "This month and next, money markets will have to absorb a large amount of net bill issuance; as such, money market conditions may tighten, and the reserves demand curve could shift back up."
That is the tension embedded in the pause. The Desk is pausing because conditions are loose today, even as it has told the market to expect tightening tomorrow. The 10-year Treasury yield was around 4.93 percent on September 22, with the 2-year near 4.73 percent — levels that reflect a market pricing in neither a liquidity crisis nor a return to aggressive easing.
The Backstop That Makes the Pause Safe
There is a reason the Desk can afford to be this flexible. In his September 22 remarks, Perli underscored the Standing Repo Facility's role in monetary policy implementation and smooth market functioning. The SRF lets eligible counterparties borrow reserves from the Fed against Treasury collateral at a fixed rate, effectively putting a ceiling on how high short-term funding rates can spike. Alongside the discount window modernization discussed at the same conference by Vice Chair Philip Jefferson, it forms a backstop that lets the Fed run reserves lower than would have been safe a decade ago.
The backstop changes the risk calculus. In September 2019, a reserve drain collided with corporate tax payments and Treasury settlement to send overnight repo rates above 10 percent, forcing the Fed to intervene with emergency purchases. Today, the SRF is designed to prevent that kind of spike from becoming a crisis in the first place. That is why the Desk can pause purchases, wait for data, and react only if the effective funds rate or repo rates show sustained pressure — rather than pre-funding every conceivable drain.
The Counter-Thesis: Could the Fed Be Pausing Too Soon?
The strongest argument against the Desk's read is that "ample" can turn into "scarce" faster than the monthly adjustment cycle can catch. Several Wall Street strategists already see the pause extending well beyond October. Citigroup strategists argue that bank reserves have been pushed back to a "lightly abundant regime" and expect the Fed to remain on pause for the rest of the year. TD Securities projects purchases could stay at zero until mid-November, then restart at only $5 billion to $10 billion a month. If those calls are right, the pause is not a temporary breather but evidence that the system has more liquidity than the Desk assumed.
The risk runs the other way, too. The pause rests on the assumption that reserve drains remain predictable — tax dates, TGA rebuilds, year-end window dressing. A surprise on any of those fronts, or a larger-than-expected Treasury refunding announcement, could drain reserves faster than the Desk's monthly calendar allows it to respond. In that scenario, the Fed would be reacting to tightness after it appears, rather than pre-positioning against it — the very problem the December decision to launch reserve management purchases was meant to solve.
The Desk's answer to both critiques is built into the framework. Perli and his colleagues wrote in March that if continued bill purchases ever led to strains in the bill market itself, the Desk has the flexibility to purchase short-term Treasury coupon securities instead, to preserve bill market functioning. And on the downside side, the standing repo facility gives banks and primary dealers a backstop source of liquidity before reserves fall to scarce levels. The toolkit is deeper than the monthly purchase announcement suggests.
Still, the falsifying signal is concrete and observable. If reserve balances fall back toward $2.9 trillion — the trough the Fed navigated late last year — or if the effective federal funds rate trades persistently above the interest rate on reserve balances, the "reserves remain ample" call would be wrong and the pause would look premature. Weekly H.4.1 data, the effective funds rate, and overnight repo rates are the dials the market should watch, not the headline purchase schedule.
What Comes Next: Three Horizons
In the short term, the pause is likely to hold through mid-October, with money markets stable and bill yields firming modestly as issuance ramps up. The reinvestment line — roughly $15.6 billion through mid-October — keeps the SOMA portfolio from contracting, so there is no tightening impulse from that side.
Over the medium term, the fourth quarter brings the forces Perli flagged: heavier net bill issuance and the usual year-end balance sheet window dressing that pulls reserves out of the system. Most strategists expect purchases to resume then, though they disagree on the size. Bank of America's rates team, citing "persistently loose funding conditions," sees the schedule staying at zero through the year before returning to roughly $10 billion a month. Barclays expects $10 billion in October and $20 billion in November. Wells Fargo sees the pause lasting at least through mid-October. The spread of views itself tells you how little the market can pin down a data-dependent Desk.
Over the long term, the structural conclusion is the clearest part of the story. Reserve management purchases are now a permanent technical instrument of monetary policy implementation, sitting alongside the standing repo facility and the overnight reverse repo tool. They are not a crisis program and not a stimulus program. They are the mechanism by which the Fed keeps an ample-reserves regime ample — and the fact that the Desk can pause them without drama is evidence the regime is working.
Three scenarios frame the path ahead. The base case is a pause through mid-October followed by a resumption near $10 billion a month as fourth-quarter issuance hits. The upside case for liquidity is a larger-than-expected refunding or a sharp year-end reserve drain that forces purchases back toward $20 billion. The downside case is funding conditions that stay loose enough to keep the Desk at zero through year-end.
The Fed has turned balance sheet management into a dial rather than a switch. The market's job is to watch the dial — reserves, the funds rate, repo rates — and not mistake the silence of a paused purchase schedule for a change in direction.
Explore more exclusive insights at nextfin.ai.
