The Coordination

Warsh promises a balance-sheet and a rate outcome under oath. Bessent supplies the Treasury-side instruments the outcome depends on. Druckenmiller endorses the arrangement. The three principals, on the fourteen-month record The Axis documented, have specified a coordinated program that points in a single direction. But the program cannot close at the scale promised without moving the reserve-demand floor, and the reserve-demand floor is controlled from a seat the three principals do not hold and do not discuss. Four instruments are authorized on the axis's public record. A fifth is required. The seat that holds the fifth is outside the axis, and the archive's record on whether that fifth instrument has ever worked is five decades of failure. These are testable claims.

The Commitment and the Seat

The contradiction is structural. Warsh's doctrine, stated under oath on April 21, 2026, narrows the Federal Reserve's independence to a precise operational area:

Bessent's companion formulation, issued on July 21, 2025 the same day he delivered a keynote inside the Federal Reserve building, compresses the architecture into a single metaphor: "the Fed's conduct of monetary policy is a jewel box that should be walled off to preserve its independence." The narrow peak and the jewel box are the same figure. Independence is operationally tight and scoped. Coordination with the rest of government is everything that lies outside the scope. The architecture is published.

The commitment attached to that architecture runs the other way. Warsh on Fox Business with Larry Kudlow, July 8, 2025:

At the confirmation hearing nine months later, the commitment becomes testimony: "Working with the Treasury Secretary, we are going to have to find out the way in which we can take the balance sheet and make it smaller." The scale is specific — a couple trillion. The partner is named. The promise — a "big rate cut," "materially lower interest rates" — is measurable on the weekly Federal Reserve balance-sheet release. The instruments the chair holds under his own doctrine are administered-rate decisions, runoff caps, and active open-market operations. These reach the short end of the curve by fiat and the asset side of the balance sheet by decree. None of them reaches the liability side, where reserves sit, and none of them determines who absorbs the duration the Fed sheds. The commitment authorizes an outcome. The doctrine narrows the instruments. The gap between them is where the other principals enter — and where the fifth instrument waits.

The Runoff Ceiling

The first instrument is passive runoff. When a Treasury matures, the Federal Reserve declines to reinvest. The Treasury pays the Fed from the General Account; Fed Treasury holdings fall by the matured amount; the Treasury auctions new debt into the private market to rebuild the account; primary dealers pay for the new issuance with reserves. The balance-sheet reduction lands on Fed Treasury holdings and on bank reserves, in equal amounts. Duration transfers from the Fed to the private market. The T-account is clean and the political cost is minimal.

The ceiling of this instrument is determined by the maturity structure the Fed holds, not by the chair's preferences. At the current $50 billion per month cap on Treasury runoff, passive Treasury capacity is approximately $600 billion per year. MBS prepayments at the current rate environment add roughly $200 billion per year. The combined ceiling is $1.6 trillion over twenty-four months — eighty percent of the stated target.

A second constraint binds before the asset-side ceiling is reached. Every asset reduction lands on the liability side primarily as falling reserves, because currency growth is slow and the Treasury General Account and reverse repo facility can absorb only cyclical swings. The Federal Reserve has approached a similar reserve level once before, in September 2019, and money-market operations broke. That constraint will arrive before month twenty-four. The reduction the chair's own tools can deliver is therefore bounded not by what runoff and sales can produce but by what the liability side will tolerate — and the tolerance is set by a framework the chair's tools do not reach.

The residual four hundred billion, plus any political demand for acceleration, also cannot be delivered by runoff alone. Warsh's hearing testimony names what it requires. The balance sheet has become an "ordinary recurring force" whose normalization puts the Fed "in the business of politics." Mortgage-backed securities in particular are "fiscal policy in disguise." Bessent in *Gain-of-Function* centers the same critique on Thomas Hoenig's 2010 dissent against the formal asset-purchase program. Druckenmiller, on CNBC in May 2024, converted the diagnosis into a single operational phrase:

"Great enabler" is diagnostic authorization, not endorsement. The phrase appears as developed institutional claim in Bessent's *Gain-of-Function* fourteen months later — "The Fed's actions fostered a culture... that encouraged reliance on the central bank to bail out poor fiscal policies" — and as hearing testimony in Warsh's April 2026 appearance. Three voices, one vocabulary, one operational consequence: the residual gap that runoff cannot close will be closed by active sales of agency MBS. The Fed SOMA desk sells roughly $25 billion per month into primary dealers, who distribute to REITs, insurers, and foreign buyers. Unlike runoff, this is a transaction; the market has to clear it at a price; MBS yields widen relative to comparable-duration Treasuries. Staff studies of the 2017–2019 balance-sheet normalization suggest $100 billion of MBS supply widens yields 10 to 15 basis points. At $300 billion per year, the mortgage-rate drag is on the order of 30 to 45 basis points. The active-sales instrument closes the supply-side gap and raises a new problem: the "materially lower interest rates" promise requires something to pull the long end down even as QT and MBS-spread widening push it up.

The Rate That Has to Fall

The second instrument is the administered-rate decision. The Fed votes to lower the interest rate on reserve balances — 4.40 percent on the February 25 release — and the overnight reverse repo rate, 4.25 percent. No securities move; no T-accounts change on the day of the decision. Banks earn less on reserves; money market funds earn less on RRP; short-term market rates drop mechanically, because no institution lends below what the Fed pays, and competition drives the rate to the administered floor. The short end of the curve moves by fiat. The political authorization arrived on the day of the hearing:

The authorization is not confined to a single cut. Trump on the same day named the Greenspan-era Fed posture as the template the chair is expected to restore — rate reductions delivered on the expansion of supply capacity rather than on the Phillips Curve. The rate cut is not a technical accommodation. It is the operational signal that the new chairmanship has taken over from Powell's posture. The short end moves as soon as the administered-rate decision is made. The long end is the problem.

Balance-sheet reduction is a quantity operation. It removes Fed demand for duration from the primary market and forces private buyers to absorb more duration per year than they otherwise would. Standard term-structure theory assigns this pressure to the term premium: long rates, all else equal, rise. The natural operational outcome of simultaneous rate cut and quantitative tightening is a bull steepener — short rates fall, long rates rise or fall less. Warsh's promise attaches to the entire curve. The promise contradicts the natural outcome unless something else is pulling long rates down.

Two pulling forces are available. The first is disinflation from the AI productivity shock, the first of Warsh's four pillars: if expected inflation falls faster than the term premium rises, nominal long yields drop. The second is Treasury issuance composition, named directly by Warsh:

"Let Secretary Bessent handle the fiscal accounts" is an operational assignment. Composition of new issuance is chosen quarterly through the Treasury Borrowing Advisory Committee and announced in the Quarterly Refunding Statements; it is not a Federal Reserve decision. If the Treasury Secretary shortens the weighted-average maturity of new issuance — more Bills, fewer long bonds — the private market does not have to absorb long duration at the pace Fed runoff would otherwise require. Bessent's publicly stated preference is shorter issuance; the pattern continues the approach that dominated Treasury debt management in 2023 and 2024 and absorbed most of the RRP drawdown into money-market-fund demand for Bills. This is the structural endorsement Druckenmiller named on the day of Warsh's nomination:

The word "accord" echoes Warsh's July 2025 CNBC proposal for a new Treasury-Fed accord modeled on the 1951 agreement. The three voices converge on the same operational structure. The issuance-composition lever is the second principal's instrument, authorized by the first principal's testimony and endorsed by the doctrinal ancestor. Bessent extended the authorization one day after Warsh's hearing. In Senate Appropriations testimony on April 22, 2026, he treated Fed and Treasury swap-line authorities as operationally interchangeable:

The doctrinal separation the jewel-box architecture would require is absent. The two authorities are treated as a single dollar-funding instrument routed through whichever seat is operationally convenient. Bessent's own framing names the operational purpose: maintaining order, preventing disorderly sales. Swap lines are a stabilizing margin, not a programmable demand engine. They do not, on their own, produce a specific quantity of foreign-CB absorption for the duration the Fed sheds. What they do is hold open a backstop that keeps pension, insurance, and mutual-fund buyers from pricing disorder into long-end yields when the Fed's asset reduction reaches its active-sales phase. The contribution is real but contingent — a defensive margin against the worst scenarios on the long end, not a replacement for the issuance-composition lever. The second principal's seat carries two operational pieces the first principal's seat cannot carry: issuance composition, which absorbs duration directly, and swap-line authority, which prevents the scenarios in which duration absorption breaks down. Both are authorized on the record. One is programmable. The other is contingent.

Four instruments are now active across two seats: rate action and runoff under the chair, issuance composition and swap-line authority under the Treasury Secretary, with active MBS sales available to the chair as a fifth operation authorized by the axis's diagnostic vocabulary. But the program still has to pass through the liability side of the Federal Reserve's balance sheet.

The Floor the Seats Do Not Reach

The Federal Reserve has reached a similar reserve level once before. In September 2019, with reserves at roughly $1.4 trillion on a balance sheet of $3.8 trillion, the Tealbook prepared for the September 17–18 FOMC meeting reported the staff's assessment of where reserves stood relative to the demand curve:

Three days later, on September 17, the overnight repo rate spiked outside its target range. The federal funds rate followed. The Federal Reserve intervened with emergency repo operations. The subsequent internal review revised the projected reserve floor upward, repeatedly, over the following eighteen months. The chair's operational summary in December 2019 named what the staff analysis had missed:

The floor was not set by Fed asset holdings. It was set by banks' demand for reserves as a liquidity buffer. Bank demand for reserves is generated, structurally, by the Liquidity Coverage Ratio — the regulation finalized in 2014 requiring large banks to hold high-quality liquid assets sufficient to cover thirty days of modeled stressed outflows. The floor is a regulatory artifact. It is not within the chair's open-market authority to move. It is not within the Treasury Secretary's debt-management authority to move. The four instruments the axis has authorized do not reach it.

The arithmetic is unambiguous. Under current LCR calibration, the ample-reserves floor is estimated at roughly $2.2 trillion. With reserves starting at $2.97 trillion, the available reduction from the liability-side constraint alone is approximately $0.8 trillion before money-market stress returns. Passive runoff and active MBS sales would deliver that reduction between months twelve and fifteen of the program. At that point the balance sheet has contracted by roughly $1.4 trillion — short of the stated target by six to eight hundred billion dollars. The remaining gap cannot be closed by continuing to sell assets. The only path forward is to move the floor, which requires modifying the Liquidity Coverage Ratio. That modification sits in a seat the axis does not occupy.

The seat is the Vice Chair for Supervision, created by Dodd-Frank in 2010 and held since June 2025 by Michelle Bowman. Bowman is not an axis principal. Her reform program predates Warsh's nomination and is doctrinally distinct from the restorationist critique. But the technical apparatus her seat controls includes exactly the framework the arithmetic requires to be modified. On March 3, 2026, at the Committee on Capital Markets Regulation Roundtable, she named the linkage:

LCR hoarding drives reserve demand. Reserve demand forces a larger Fed balance sheet. Recalibrate the LCR and the balance sheet can be smaller. Bowman's compatibility claim closes her own argument:

The fifth instrument the arithmetic requires is authorized by its own speaker, operationally specified in its own venue, on its own calendar. The axis's four principals do not discuss it. The technical seat does. The two programs — the axis's monetary-doctrine program and Bowman's regulatory-reform program — interlock at the reserve-demand floor, which is the single piece of plumbing both require. Without the fifth instrument, the program stops at month fifteen. With it, the floor moves. Whether it delivers the behavioral change its speaker presumes is the remaining question.

The Distributional Theory

The strongest version of the axis's case is not the arithmetic of runoff, the mechanics of rate cuts, or the coordination of Treasury issuance. It is the distributional theory. Warsh, on CNBC in July 2025, making the case to a sympathetic anchor:

The frame is the most analytically attractive piece of the restorationist case. It does not claim that quantitative tightening lowers rates through conventional transmission. It claims something different and structurally more ambitious: that shrinking the Federal Reserve's financial-market footprint redeploys liquidity from Wall Street to Main Street. The asset-market froth comes down — IPO pricing, narrow credit spreads, renewed SPAC activity — while real-economy credit expands. The sympathetic reader's instinct is to agree. The critique of financial-market accommodation is real. The distributional record of the post-2008 QE era is real. The question of whether monetary expansion reached the borrower it was supposed to reach has been one of the Federal Reserve's most uncomfortable internal debates for a decade.

The archive has tested the channel Warsh names. Across four decades of easing cycles, Robert Kaplan at the Dallas Fed, Elizabeth Duke on the Board of Governors, Charles Evans at the Chicago Fed, and Loretta Mester at the Cleveland Fed surfaced the same institutional finding: the expansion of Federal Reserve liquidity did not reach the small business borrower through the channel policy documents described. Large corporations with capital-markets access received rate-cut transmission. Households with mortgage refinancing access received rate-cut transmission. Small businesses dependent on bank lending received tightened credit standards from the same banks the Fed was supplying with reserves. The 2011 Tealbook documented the result:

A staff economist made the institutional point direct in a September 2011 FOMC deliberation:

The mechanism Warsh invokes operates in the opposite direction from what his prescription requires. Adding liquidity did not transmit to the real economy through the channel the post-crisis framework described. There is no documentary basis in the Fed's own record for the claim that removing liquidity will transmit in reverse through the same channel. The accommodation-withdrawal half of the story may well be correct — reducing Federal Reserve footprint in financial markets may bring down asset prices, narrow the froth Warsh names, and undo some of the distributional effects the archive has documented. The redeployment half is the claim the archive does not support. The pipe that did not carry liquidity to Main Street in the expansion direction has no record of carrying it in the contraction direction.

The diagnosis survives. The prescription does not. The asset-market accommodation is documented, the distributional effects of post-2008 QE are documented, the institutional frustration with monetary transmission is documented. What the archive refuses is the mechanical claim that shrinking the Fed's balance sheet redeploys financial-market liquidity into real-economy credit. The channel runs one way or it does not. The Federal Reserve's own forty-year record says it does not.

The Seat That Does Not Hold

The doctrine names the seats it needs. The arithmetic requires a seat it does not name. That is the shape of what the archive has shown.

The doctrine's four instruments — rate action, runoff, active MBS sales, Treasury-side coordination through issuance composition and swap-line fungibility — reach roughly $1.4 trillion of balance-sheet reduction over twenty-four months. That is seventy percent of the stated target, delivered through the chair's seat and the Treasury Secretary's seat operating together, on political authorization from the president and doctrinal endorsement from Druckenmiller. It is a substantial program. It is not the program Warsh committed to under oath.

The remaining gap, and the rate-cut promise attached to the full commitment, require the ample-reserves floor to move. The floor moves only if the fifth instrument delivers. The fifth instrument is the regulatory reform program the Vice Chair for Supervision has been rolling out since December 2025. The enhanced supplementary leverage ratio was finalized then, restoring the SLR's role as a backstop so that banks can hold Treasuries without leverage-capital penalty — which, in the axis's arithmetic, expands the domestic bank absorption channel for the duration the Fed sheds. The three banking agencies jointly rescinded the public LCR FAQs on February 10, 2026. The Basel III reproposal was previewed at the Cato Institute on March 12. Small-business risk weights were reduced at CBA LIVE on March 31. Each of these is a rule change that, assembled alongside the LCR recalibration Bowman previewed on March 3, moves the reserves floor downward and expands bank capacity to hold Treasuries and extend loans. The doctrinal coordination the axis describes and the technical coordination the arithmetic requires intersect at these specific rule changes on this specific calendar.

The distributional theory, too, rests on this seat. If the redistribution from financial markets to the real economy that Warsh promises is to occur, it does not occur through the Fed's asset composition — the archive has shown the channel does not run. It occurs, if at all, through the bank-lending channel itself. Bowman's risk-weight reductions on small-business and community lending, her mortgage-capital recalibration, her LCR recalibration — these are the mechanisms by which bank capacity to extend real-economy credit can expand. Warsh's seat cannot produce that expansion. Bessent's seat cannot produce it. The seat that can produce it is not in the axis.

Whether the regulatory seat delivers the behavioral change its reforms presume is the remaining question. Bowman herself, speaking at the same Committee on Capital Markets Regulation roundtable venue in April 2024, stated the institutional position on that question:

The 2024 analysis was the synthesis of five decades of institutional experience — the 1984 FDICIA-era reforms, the 2003 primary-credit redesign, the Term Auction Facility of 2007–2010, the 2019 repo event, the 2023 Bank Term Funding Program — across which administrative reforms to the discount window failed to eliminate stigma each time they were attempted. The Window tested the 2026 Bowman prescription against that record and found the archive supports the 2024 position: stigma operates on the demand side, in the rational calculations of bank treasurers who know that detection carries reputational costs no regulator can offset, and administrative reforms to the supply side cannot reform that dynamic. The current reform cycle proceeds against the archive's five-decade record of exactly this attempt failing.

Two endpoints sit at the twenty-four-month mark. If the fifth instrument delivers behavioral change alongside rule changes, the Federal Reserve's balance sheet reaches approximately $5 trillion on reserves of roughly $1.5 trillion — in the range of eighty percent of the stated target, with the remainder rolling into year three on continued runoff against a now-lower floor. If the fifth instrument delivers rule changes without behavioral change, the program reaches approximately $5.8 trillion on reserves of roughly $1.9 trillion — around forty percent of the stated target, paused visibly around month fifteen. Both outcomes are available on the operational sequence the four principals have authorized. Which obtains is determined by the seat the axis's architecture does not name, and by a behavioral dynamic the archive has documented unchanged across five decades.

The Coordination