Macro Memo: Regime Change

When the Bank Term Funding Program (BTFP) backstopped interest rate risk from being realized in the banking system, we viewed that as an “all clear” signal to invest in growth. The last transmission mechanism

Six weeks later, that thesis had become the spine of the first ever paywalled piece on this newsletter, Artificial Intelligence: Global Equity Beneficiaries. CitriniResearch was born.

For this month’s macro memo, we’re returning to our roots and examining a setup three years in the making. One that has been brewing since the SIVB/FRC pseudo-crisis put GFC-era liquidity regulations back under the microscope.

But the implications of this setup extend beyond just the banks. They are part of a larger regime change that explains the motivation behind the Bessent “Twist”, Warsh’s balance sheet hawkishness, and what both mean for the bond market and money supply.

Our view: We see bank liquidity reform as the Trojan Horse for a new Treasury-Fed Accord, one in which the Fed shrinks and the banking system expands, both the Fed and the banks buy bills, the Treasury sells bills instead of bonds, and long duration paper gets scarcer.

Shrinking, Not Tightening

TLDR: We expect that monetary and fiscal authorities – Fed Chair Kevin Warsh and Treasury Secretary Scott Bessent – are aligned on a framework that solves for two critical problems, while still stimulating growth:

Reduce the Fed’s Balance Sheet

Improve US fiscal sustainability and reduce higher long-term rates

The plan looks like this:

First - Banks: Reform bank liquidity regulations, stimulating the private sector and expanding money supply via bank leverage.

Second - Twist: A new Fed-Treasury Accord 2.0 results in dual efforts focused on reducing open market UST duration (by increasing bills issuance). The Treasury alone is not doing a “Twist”, and investors are underestimating the role of the Fed in attempting it.

The result is a smaller balance sheet, lower coupons, and greater private sector growth. If we believe them, three trades follow…

Part I – Liquidity Reform: The Banking System Re-levers

Secretary Bessent in March fired a warning shot that the banking system was about to be unshackled:

“To unlock the vast promise of this transformation and secure America’s Golden Age, we confront the pressing necessity of unlocking hundreds of billions—potentially trillions—in new lending capacity to finance AI infrastructure, domestic supply chains, and the defense industrial base.

The problem, however, and the reason this roundtable is timely, is that the framework for supervising and regulating bank liquidity created in response to the 2008 financial crisis has excessively and unnecessarily

Which is all to say, liquidity regulation is an area of steady and serious focus for the Treasury Department.”

The two footnotes in Secretary Bessent’s speech that referenced liquidity policy proposals cite Jeremy Stein, a former Fed Governor from 2012-2014. The same Jeremy Stein now helps lead Warsh’s Federal Reserve “balance sheet task force”.

On the surface, liquidity regulation is presented as improving the safety and soundness of the banking system by preventing a repeat of the mechanics that led to SVB’s failure. In reality, however, we see it as the thread tying together Secretary Bessent and Chair Warsh’s goals – shrinking the Fed’s balance sheet without tightening.

Regime change is in the air.

Note: While we do our best to explain our view on banking liquidity regulation in plain English, if you need a refresher on how banks work, consider re-visiting our 2023 article “Evaluating Banks and Insurers”.

Safety & Soundness: A Trojan Horse

A bank can be insolvent for a long time and not die. Illiquidity is what kills a bank. Or, at least, illiquidity at a time when liquidity is in high demand among your depositors. You never want to run out of cash on the same afternoon all of your depositors want theirs back.

SVB was already technically underwater well before it collapsed. In fact, in October 2022, it was more insolvent than it was before the bank run started. But it wasn’t the number on paper that killed them.

The Fed’s post-mortem report on SVB indicates the bank had $40 billion of outflows on Thursday, March 9, 2023, and expected another $100 billion the following day.

This $140 billion comprised 85% of SVB’s deposits, its “runnable” funding. This is the type of crisis envisaged by the writers of the Federal Reserve Act in 1913 in creating the discount window; the Fed is supposed to be the ‘lender of last resort’ for the banking system, to prevent panics that crunch confidence (liquidity).

SVB had only $31 billion of borrowing capacity in collateral prepositioned at the Fed’s discount window/FHLBs, and it had not properly tested its pipes to draw on the discount window in years.

Despite this, SVB had a securities portfolio of $120 billion ($106.9 billion MTM) that was almost entirely Treasuries and MBS. This portfolio would’ve been eligible collateral to borrow against even if it was trading below par. It was contingent liquidity, but it could not be accessed as it was not properly positioned at the discount window.

In theory, SVB may have had sufficient contingent liquidity to get to the weekend if it was properly positioned. (It likely still would have failed, though). In any case, the deposits ran, and SVB didn’t have the money to give them back. The FDIC Grim Reaper closed the bank on March 10, 2023.

The failure of SVB demonstrated that a bank positioned to borrow from the Fed’s facilities is more liquid than one that isn’t.

You might be thinking: “Wow, this victory lap is oddly informative”. But we aren’t taking a trip down memory lane for the sake of tallying past wins. This is more important now than it was in the immediate aftermath of SVB.

Bank regulations move slowly. More than three years ago, SVB exposed a mismatch between economic liquidity and regulatory liquidity. The focus shifting to liquidity regulation matters now.

On March 3 of this year, Treasury Secretary Bessent gave a speech with the explicit ask that liquidity rules recognize borrowing capacity at the Fed’s discount window and the FHLBs against prepositioned collateral. The same day, Michelle Bowman, the Fed’s Vice Chair for Supervision, also gave a speech on the bank liquidity framework, echoing Bessent’s explicit call for recognition of borrowing capacity at the discount window.

So, what does reforming liquidity regulation mean for the banks?

Less cash required. More loans. More securities. In other words – more leverage on tangible liquidity.

Put simply, contingent liquidity may start counting as liquidity. That may sound tautological, but it would be a transformative development for the banking system.

Banks have to carry high quality liquid assets (HQLA) to comply with the key current liquidity regulation, the Liquidity Coverage Ratio (LCR) rule. The LCR rule identifies how much money theoretically could flow out in a 30-day panic and forces each bank to hold enough HQLA to cover that potential outflow.

A few weeks after Secretary Bessent and Vice Chair Bowman’s remarks in early March, then-Fed Governor Stephen Miran wrote ‘A User’s Guide to Reducing the Federal Reserve’s Balance Sheet’, described as a menu of options rather than any policy endorsement. On that menu, Miran proposed to recognize prepositioned discount window capacity in the numerator of the Liquidity Coverage Ratio, capped at 20% of the bank’s HQLA.

Say a bank holds $114 of cash and government bonds against a theoretical $100 of deposits that could leave in a panic: a 114% LCR. If it prepositions its collateral at the discount window, it can count up to $23 more

(20% of $114) towards the ratio (now a 137% LCR).

But banks do not need higher ratios – they need the same ratio with less collateral. To stay at 114%, the bank can now hold just $95 of cash and bonds, count 20% of that ($19) and still show the same $114 of liquidity against the $100 of deposits.

What does this mean when extrapolated to the entire banking system, or rather the entire economy?

In plain English, Miran outlined $500 billion to $1 trillion of cash that the banking system could redeploy from potential liquidity regulation reform.

I understand that the past few years have made large numbers lose a bit of their impact, so let’s recharacterize – that’s 1.5-3% of nominal GDP that could become freed up to be lent out (with a multiplicative effect).

There’s already some evidence that banks are frontrunning what’s coming. JP Morgan specifically is running its liquidity book to the framework Bessent and Bowman described in March, rather than to the rule that is still on the books.

JP Morgan has reduced its cash asset ratio to 6% – the lowest level of the post-GFC era. Only a small portion of that ($52 billion) sits in reserves at the Fed. But if JPM, the largest bank in the country – with nearly $5 trillion of assets (19.5% of the banking system!) – can operate with just 6% of its assets in cash, why is the rest of the banking system operating with ~13%?

In Jamie Dimon’s shareholder letter in April 2026, he included a regulatory proposal:

“The liquidity component of loans and securities should be equal to what the Fed discount window would lend against those securities. We should eliminate duplicative or unnecessary liquidity buffers. These actions would create an enormous amount of lendable liquidity and also allow banks to use their capital far more flexibly in a crisis. They would also reduce the need for the Fed to step in every time there is a kerfuffle in the market. Credit for the Fed discount window alone would increase JPMorganChase’s lendable liquidity by almost $500 billion.”

Consider that Jamie Dimon has spent the last two years explicitly warning of higher rates… while JPM rotated >$250 billion of overnight-earning cash reserves at the Fed largely into buying Treasury bonds.

If you thought rates were going higher, why would you materially lengthen your asset duration?

We don’t know what JPM is sniffing out from regulators or the administration. But we do know what JPM has committed to:

The SRI funds the exact sectors that have been used to justify the reform, demonstrating the channel by which the administration can get the banking system instead of the Fed to finance the economy the way the US did post-WW2.

More interestingly, the sizing rhymes with the reform math. Dimon’s letter puts discount-window recognition at “almost $500 billion” of new lending liquidity for JPM. This initiative raises JPM’s decade-long financing commitment by up to $500 billion over a previous baseline of $1 trillion. They’re not identical, but one of the

The Trade: Bank Stock Implications & Basket Idea

Liquidity reform transmits to bank earnings through three channels:

Funding Relief – or paying down borrowings whose costs exceed what they earn on cash – is the most immediate, albeit small and balance sheet reductive.

Asset Mix Shift – or redirecting liquidity towards higher yielding assets – is the most earnings accretive, but will take a few years and is discretionary to risk appetite.

Growth via incremental balance sheet capacity (higher leverage) is likely the most multiple accretive.

Below the paywall, we detail the banks most likely to benefit from this and those who wont and could be used as a hedge. We also detail Part II of the plan, the Treasury’s Twist and the role the Fed will (and is already) playing in it.

The four large-cap banks that stand out as the biggest winners to us are Bank of America, US Bancorp, Truist Financial, and Capital One Financial.

Bank of America (BAC US) is the clearest example. BAC bought a ton of mortgage bonds in 2020-21 at very low rates. A year later, when rates rose, those bonds lost value on paper. Because the liquidity rule marks bonds at fair market value, the cushion shrank with them. This explains why its securities-backed borrowing jumped in early 2023 and has remained elevated since: the quickest way to rebuild the cushion was to borrow

US Bancorp (USB US), the sixth largest independent bank in the US, is the same story in a smaller package. Relative to capital, its paper losses on bonds are the largest of the big regional banks. Its balance sheet is propped up by expensive borrowed money it can now let run off. USB used to be valued above its peers because its businesses earn more; with room to grow, the premium should come back.

Truist (TFC US) is US Bancorp with some of the pain already taken. In 2024, it sold a chunk of its low-rate bonds at a loss, and it still carries paper losses on the rest. Truist’s balance sheet has actually shrunk over the past four years. It has pole position in the country’s fastest-growing deposit markets and more capital than it needs; add freed-up cash, and growth should turn.

Capital One (COF US) is different. While it is similarly a large bank with securities losses, its book is mostly credit cards, which earn the highest risk-adjusted margins in banking — roughly 8% NIM against ~2.5% charge-offs — so each redeployed dollar generates more net spread than at any peer, though it also consumes more capital.

Our long basket consists of 30% BAC, 25% USB, 30% COF, 15% TFC.

Other longs that fit our criteria but which we are not high conviction on are CFG, MS, SYF, PNC and MTB.

While a short leg may be unnecessary if this regime change manages to lower rates durably, it’s valuable to take a look at who won’t benefit and can provide a hedge against higher rates.

As SVB made vividly clear, contingent liquidity recognition only works for banks whose depositors, raters, and examiners trust the contingency.

Flagstar Bank (FLG US) (formerly New York Community Bank) is two years removed from a near-death deposit scare. Its supervisors, and more importantly its confidence-sensitive deposit base, require self- insurance regardless of what any rule says.

The risk to our basket of bank longs is that rates rise precipitously and the market starts to price in lower capital levels from widening securities losses. That same risk could break the construct of reflexive confidence for banks whose depositors have panicked in recent years, and who are already light on liquidity. Flagstar (FLG US) had its near-death deposit scare in January 2024; Western Alliance (WAL US) and Banc of California (BANC US, formerly Pacwest/PACW) each saw depositors panic and withdraw in the aftermath of the SVB contagion.

Today, while all three banks are exempt from the LCR given their size, it is possible to construct an estimate for LCR via each bank’s regulatory call report. We build HQLA from each bank’s call report cash and securities lines and apply the LCR rule’s standardized runoff weights to their deposit mix. All three banks under today’s rules sit well below 100%, and even pro forma for the liquidity reform posited above, remain nicely below

100%. This means that they are not sufficiently liquid to cover the outflows the LCR rule’s formula estimates.

Accordingly, if our long basket were to not perform due to precipitously higher rates… these banks would be significantly worse off, possibly existentially.

Our short basket consists of 60% FLG, 25% WAL, 15% BANC, sized dollar neutral against the longs. These banks won’t benefit from reform because they are, and post-reform will remain, light on liquidity.

Other shorts that fit our criteria but that we’re lower conviction on are CFR, OZK, and COLB.

The main risk to this trade is that reform could be diluted – for instance, a cap below Miran’s 20%, or a fee on borrowing capacity.

Notably, banks don’t have to spend the freed up cash. Lending is the biggest lever, but underwriting and deploying is not instantaneous. We’ve focused our picks on the names that have the fastest levers (repaying debt and growth).

You can find our Liquidity Reform basket on the Citrindex portal here.

Part II – The Twist: Accord 2.0 and the Long End

Chair Warsh is well known for his publicly voiced opposition to QE 2/3 in 2010-2014; his objection was the Fed becoming a price-maker in the Treasury market and therefore transitioning into an enabler of irresponsible fiscal policy.

It appears that the goal now is to enact a regime shift where normal-course liquidity propagation occurs in the private banking sector, rather than from the Federal Reserve.

The Fed wants the banking system’s balance sheet to expand rather than its own. In fact, it appears they want to reduce the Fed’s balance sheet as they expand the banks’ balance sheet. Famed money market strategist Zoltan Pozsar has written on this recently, labelling it ‘Private QE’.

In Miran’s ‘A User’s Guide to Reducing the Federal Reserve’s Balance Sheet’ menu, all these policy options are framed in the context of reducing bank reserves.

Recall that bank reserves are just deposits that banks hold at the Fed. But bank reserves alone aren’t the critical driver of economic activity. Rather, it’s the spendable money supply.

When a bank makes a new loan, it expands the money supply via incremental leverage on its cash. Other things equal, lower liquidity requirements allow M2 money supply to expand without balance sheet expansion at the Fed.

To recap:

1. Liquidity reform may be coming via reduced LCR requirements.

1. Incremental liquidity is “released” into the private banking system, meaning banks redeploying

their cash into loans/securities.

1. The net effect is likely an expansion in money supply through private sector lending and

incremental bank leverage, even potentially as the Fed shrinks its balance sheet.

Now, the argument against our third point is that the entire purpose is to lower the reserve floor so the Fed can shrink further without breaking repo. Just because the banks can go increase lending doesn’t mean they will.

But ask what Chair Warsh does with a balance sheet that is $500 billion smaller and a banking system that hasn’t grown at all. That is just tightening with extra steps. Tightening that’s delivered into a midterm year, with an AI capex cycle to finance, by an administration whose Treasury Secretary has been publicly demanding “hundreds of billions, potentially trillions, in new lending capacity” since March. We don’t think Jamie Dimon raised a $1 trillion financing commitment by $500 billion because he expects to be running a smaller bank.

Everything above concerns the liability side of the Fed’s balance sheet: bank reserves. The asset side, though, belongs to the Treasury.

The Treasury’s Dilemma

A Fed that shrinks is a Fed that stops buying government debt right as debt to GDP crosses levels last seen in WW2. So, while Warsh gets the balance sheet he has wanted for fifteen years… he also hands the country’s Chief Bond Salesman a funding gap.

A key argument of now-Fed Chair Warsh’s for the last fifteen-plus years has been that a large Fed balance sheet is an enabler of excessive government borrowing. For example, April 2025 at the IMF:

“QE – with some fits and starts in the 2010s — has become a near permanent feature of central bank power and policy.

Fiscal policymakers — that is, elected members of Congress — found it considerably easier appropriating money knowing that the government’s financing costs would be subsidized by the central bank.”

When COVID hit in March 2020, the fiscal and monetary authorities responded as if it were the start of a world war. This is chronicled well in Nobel laureate (and Fed inflation task force co-lead) Tom Sargent’s 2022 co- authored paper, ‘Three World Wars: Fiscal-Monetary Consequences’.

Deficits above 10% of GDP have only occurred three times in the Federal Reserve era: WW1, WW2, and COVID. Spending in the post-GFC period compounded the COVID borrowing, and debt-to-GDP is today >100% for the first time since the WW2 period.

Elevated deficit spending continued well past the pandemic emergency period, extending through both the Biden and Trump administrations to-date. Very little of this was financed via taxes, compared to the previous episodes of drastically elevated borrowing.

The most important distinction between WW1/WW2 and COVID, however, is that the Federal Reserve bought a majority of the Treasury issuance during COVID, effectively monetizing the Treasury’s spending.

In monetizing a significant portion of the sovereign borrowing in response to the pandemic, the Fed created the precondition for fiscal primacy. Once fiscal capacity was demonstrated at war-like scale, the political incentive to use it propagated.

That is how we got the dynamic that we dubbed U.S. Fiscal Primacy in 2023.

It is clearly a central concept that Chair Warsh has been focused on for a long time. In aiming to reduce the Fed’s balance sheet, he is attempting to return the punchbowl of liquidity back to the private markets, rather than the Fed, which at the end of the day is a financial agent of the Treasury (even if they are legally independent).

This is the transition he aims to initiate via liquidity reform for the banking system.

Chair Warsh stated in July 2025 on CNBC:

“We need a new Treasury-Fed accord, like we did in 1951 after another period where we built up our nation’s debt and we were stuck with a central bank that was working at cross purposes with the Treasury. That’s the state of things now.

So if we have a new accord, then the... Fed chair and the Treasury secretary can describe to markets plainly and with deliberation, ‘This is our objective for the size of the Fed’s balance sheet.’”

If the target size of the Fed’s balance sheet is smaller (i.e. the Fed is not an active incremental buyer of Treasuries), then the target size of the banking system’s balance sheet must be larger (i.e. the banks are the incremental buyer of Treasuries.)

Banks own a smaller share of the Treasury market today than almost any point in the Fed era. They used to be the predominant financing mechanism for the economy and the government, rather than alternative asset managers and the Fed.

At the same time, bank ownership of outstanding Treasury debt is coming off its all-time lows, while Fed ownership of outstanding Treasury debt is coming off its all-time highs.

The Treasury wants the banks to lend more. Like a lot more. Similar to the post WW2 era, we are in the early stages of a generational boom in capital expenditures, and banks awash in liquidity will be able to lean into that.

When we look at how loaned up the US banking system balance sheet is vs the past, it clearly has room to go higher.

But baiting banks back to the Treasury market also requires a twist.

Treasury Fed Accord 2.0: Beginning With a Twist

Everything above builds to one question: once the Fed stops absorbing the Treasury’s duration, who does?

Our answer is the Treasury will simply sell less long duration paper. The Fed is already shortening what it holds, and the two together are – in practice – an Accord 2.0.

But first, why does Secretary Bessent need the long end lower?

At the Manhattan Institute in June 2024, seven months prior to his confirmation as Treasury Secretary, Scott Bessent outlined a version of Shinzo Abe’s “3 Arrows” for the United States, targeting 3% real GDP growth, a 3% deficit-to-GDP, and a 3 million bbl/d increase in oil production.

Since then, Secretary Bessent has been ridiculed for the 3% deficit goal by the media, the financial community, and the political sphere.

The administration initially pursued a ‘move fast and break things’ approach to getting the numerator down – galvanizing the DOGE effort in pursuit of austerity.

This backfired on Liberation Day. The 10y Treasury yield went from 3.86% on April 7 to 4.58% on April 11, a 72bp move in just four days. It’s one of the most violent moves in modern bond market history and served as the catalyst for Trump to cool off on his tariff approach.

In the weeks following, Secretary Bessent began to message that the administration was now focused on stimulating (nominal) economic growth above the rate of growth in the deficit. Secretary Bessent has repeated the ‘grow our way out’ plan numerous times since.

In other words, the new approach to reducing the deficit’s share of GDP is via the denominator, i.e. ‘growing into the debt’.

Outgrowing the debt was the primary mechanism of how the United States materially reduced debt-to-GDP post WW2, the last time the national debt was this high. Secretary Bessent is attempting to engineer a similar feat as in the postwar-1940s.

Note however that in the post WW2 period while real growth was strong, the bulk of the nominal GDP growth came from inflation. We inflated away the debt.

Also note that the US government post WW2 often ran balanced budgets, not deficits, and thus held the nominal level of the debt flattish for decades.

Secretary Bessent, having taught financial history at Yale, is familiar with this playbook: restrain growth in the debt as much as possible, and grow the economy as fast as you can.

Last Thursday, Secretary Bessent went on CNBC and gave explicit color on his plans for fiscal consolidation. We believe the signal is underappreciated.

“I think a very, very good chance we have \[seen the peak deficit during this administration.\]”

As Secretary Bessent noted on CNBC, the recent uptick in the deficit is entirely a function of tariff refunds. Looking ahead, this trend should reverse as the administration re-implements tariffs under Section 301. The yellow dotted line shown above indicates that the primary deficit ex customs hasn’t really increased.

The notional deficit (i.e. actual spending) has been declining since last summer, driven by both spending cuts as well as a revenue tailwind from tariffs.

Secretary Bessent said in the same CNBC interview:

“The administration, we are announcing \[...\] an increased focus on fiscal consolidation. And it’s coming from President Trump… combined with the vice president’s Fraud Task Force, where I think we could save several hundred billion dollars… \[T\]he CBO projections… are always wrong…

There’s nothing magic about the $40 trillion \[government debt\] number… we can grow our way out of that.

When I look, we have 6.0-6.5 percent nominal growth”

Still, even fairly aggressive assumptions around growth and government finances get you to maybe a 5% deficit with 6% nominal growth – in this case the current 106% debt-to-GDP ratio only declines by ~0.7% per

A 5% deficit is still very high by historic standards; in the Federal Reserve era, the government has only run a deficit that high in 20 years out of 113. Virtually all of those periods were during wartime or financial crisis.

We believe the deleveraging outlined above is insufficient in the medium/long term to restrain the secular selloff in long bonds.

Secretary Bessent consistently highlights that his goal is for nominal economic growth to run above the effective coupon on the national debt. In other words, Bessent’s stated strategy is to deliver negative real returns to bondholders.

That is the medium-term view, and it is bearish for duration. It is also, however, exactly why the near term matters for Secretary Bessent: a plan that needs nominal growth to outrun the coupon needs a lower coupon, and the cheapest way to get there is to sell fewer bonds while the market is short them.

Who buys the bills?

Secretary Bessent keeps repeating that the Treasury will have a greater role in financial regulation… and he needs to find buyers for our sovereign paper.

“Importantly, being “Regular and Predictable” does not mean that Treasury’s issuance policy cannot or should not evolve as investor demand changes. If our borrowing outlook changes, so will the amount we issue. And if structural demand for certain products or tenors evolves over time, we will be responsive and adjust how we allocate issuance accordingly.

For example, we are closely monitoring growth in money market funds and the stablecoin market, which are both large investors in Treasury bills. Money market funds are now valued at about $7.5 trillion, having grown by nearly $1 trillion in the last year alone. The stablecoin market, meanwhile, is valued around $300 billion and could grow tenfold by the end of the decade thanks to the innovation made possible by the GENIUS Act. As money market funds and stablecoins grow, so too will the demand for Treasury bills.

In addition to growing demand for Treasury bills from money market funds and stablecoin providers, we are witnessing increased demand from banks as they shake off the excessive oversight that held them back. Since the start of this year, bank portfolios have expanded their Treasury holdings. Additional reforms… could further accelerate this process… and we will adjust our long-term issuance plans accordingly.”

The banking system is set to take up the mantle as the incremental buyer of Treasury issuance. A key distinction must be made: they are not captive buyers, and the yield (price) is not fixed. In the aftermath of the SVB crisis where deposit duration was proved to be extremely short, the banks are incrementally sensitive to the length of their asset duration. Kevin Warsh’s distaste for forward guidance, continued inflationary pressures and fiscal uncertainty all combine to form a preference for the short end.

The banks will be duration-sensitive buyers of Treasury issuance. They will predominantly buy T-bills and short- term bonds.

A change is necessitated in the Treasury’s issuance (supply) to duration-match incoming demand. Bill issuance needs to increase; long bond issuance can decrease.

But banks are not the only incremental buyer the Secretary is chartering (no pun intended…)

Stablecoins

The GENIUS Act’s implementation in January 2027 is a major catalyst for incremental demand for Treasury bills. Harnessing stablecoins’ implicit T-bill reserve demand is a compelling candidate if one is searching for a reason the administration has been so keen to pump crypto markets (beyond Trump’s own PnL).

As crypto market cap goes higher, so does stablecoin issuance. That means more demand for T-bills, which is monetizable funding for the Treasury.

After MMF reform led to increased demand for bills, the Treasury increased supply to ensure collateral did not become scarce – if current growth forecasts for USD-pegged stablecoins are accurate, it’s entirely plausible we could see the same reaction of increased bill issuance to accommodate.

Just this past week, the Treasury put out its Notice of Proposed Rulemaking (NPR) for stablecoin issuance.

The announced theme of this year’s Jackson Hole is “Financial Innovation”. That could refer to several different topics – perhaps a symposium on why innovation in finance is so highly correlated with eventual financial crises directly linked to said innovations – but we feel it’s most likely related to stablecoins.

Stablecoins provide a direct and growing source of global demand for the short end. And the GENIUS Act has paved the way for new, bank-issued stablecoins that could result in even more incremental demand if those stablecoin accounts were to come at the expense of demand deposits.

We’ve spoken about eSLR relief before – the eSLR recalibration became effective on April 1, 2026. Without going into detail again, the leverage ratio stops penalizing a low-margin, Treasury/reserve/repo heavy book and effectively lowers the cost of holding and intermediating treasuries.

By cheapening banks’ treasury warehousing at the subsidiary level (where a bank-issued stablecoin’s reserve pool would sit), it enables more effective incentives for banks to issue stablecoins that must be backed by short-dated treasuries. Stablecoins are barred from paying yield, but can compete on other rewards like points and cashback (imagine a checking account with credit-card style rewards). Demand deposits that would otherwise mostly become loans and have a small “safe” remainder put into bills could transform into stablecoins that require the issuer to hold about a dollar in bills.

TLDR: Secretary Bessent confirms that 1) liquidity reform and leverage relief will drive banks to buy more Treasury bills 2) stablecoins are another major incremental source of Treasury bill demand and 3) the Treasury is explicitly willing to adjust its issuance of bills vs. bonds to meet that demand. We believe that Jackson Hole will be constructive for risk assets as it becomes clearer what the intentions behind the new Fed-Treasury Accord are.

Clues from the QRA

In just the last few weeks, we’ve already gotten a clue that this change in Treasury issuance mix is impending.

The Treasury holds a Quarterly Refunding Announcement (QRA) where it announces the details of its planned bill and bond auctions for the following 3 months. The 3Q26 QRA was on August 5, 2026, and there was a key

In the 2Q26 QRA on May 6:

“Looking ahead, Treasury continues to evaluate potential future increases to nominal coupon and FRN auction sizes, with a focus on trends in structural demand and potential costs and risks of various issuance profiles.”

In the 3Q26 QRA on August 5:

“Looking ahead, Treasury continues to evaluate potential future changes to nominal coupon and FRN auction sizes, with a focus on trends in structural demand and potential costs and risks of various issuance profiles.”

If we read the tea leaves, the Treasury is considering reducing its mix of issuance in bonds (i.e., coupon auction size changes don’t have to be increases / could be decreases) while increasing its mix of issuance in bills. More bills, less bonds.

What a ‘twist’ is, and what it isn’t

The two dimensions of debt are how much and for how long.

A “twist” alters the second (duration) without changing the first (quantity of issuance). The Treasury can twist by tilting its issuance resulting in the same total borrowing with a different maturity mix. The Fed, on the other hand, twists by tilting what it holds, buying bonds or selling bills (or vice versa).

We have seen the Treasury and/or Fed twist twice before.

In 1961, Kennedy inherited a recession that demanded lower long term yields to stimulate growth. The gold standard, however, meant foreign short rates sat higher than the US’s, so holding short yields down would have sent gold draining abroad. The Treasury tilted its issuance towards bills and the Fed abandoned the bills- only doctrine it had adopted in 1953. It bought bonds while selling bills – a ‘Fed Twist’. Long rates came down, short rates held. The curve flattened.

In 2011, the Fed ran the same playbook without any help from the Treasury. By selling short paper and buying long rather than another round of QE, the curve flattened again. These were two unique setups with the same result.

Today, we’re likely witnessing the same thing.

Bessent’s CNBC appearance last Thursday had an interesting tidbit:

“Well, again… people have bad information. I have asymmetric information, so I think that the market should think, well, why would we have joined the Japanese in the intervention at this time?

Do we know something the market doesn’t know… in terms of being willing to do \[...\] what I would call a Treasury twist here in terms of the bond market? What do I know that the market doesn’t know? So I think the market’s probably gotten a little ahead of itself.”

Remember, this is the man who left a note reminding him to buy “Buy Japanese Yen (JPY) $5-$10 bil”.

Just like that piece of paper, we believe this is much more about signaling to the market that the government is ready to get started with a whole host of programs – and the Treasury twist is only a supporting character in that play.

Bessent pretty much said as much:

“Well, again, we have a big toolkit, so we will see. And part of it is signaling here, and to show that we believe that the yields don’t reflect the underlying fundamentals.”

Maybe he has more information that the market doesn’t have? Maybe something that he didn’t say out loud on CNBC?

We think his intentions are not only to signal whether or not there will be a Treasury Twist, but to signal the new regime and the incoming Accord 2.0. We don’t think it’s a bluff, but we also recognize that he understands the value of determining whether he can scare money short bonds into doing his job for him.

On September 9, we will find out.

If the Treasury uses the Treasury General Account, as has been rumored, the first enlarged operation will function as a cash-funded retirement of long-duration debt – removing long bonds from the market without issuing replacement securities at the time of settlement. In other words, this won’t be a classic Treasury buyback that retires off-the-run, low-liquidity issues simply to issue new bonds that can trade with a deeper book. But whether the buyback ends up actually resulting in a net buying effect on the long end won’t be determined immediately.

The second half of the transaction will determine the actual impact – if the TGA is rebuilt by issuing bills then it’s a full-fledged Treasury twist. If it is rebuilt with long coupons, it is pretty much duration neutral. If they simply leave the TGA depleted, the impact on duration will persist but with a reduced cash cushion.

The Fed’s half of the twist is already underway

This may not be just a ‘Treasury twist’. As in 1961, a ‘Fed twist’ is waiting in the wings. In fact, it has already started.

Chair Warsh in his July 2026 testimony to Congress said in his opening statement referencing the task force:

“The Fed balance sheet, both its size and duration, are worthy of a very worthwhile review. I’m inclined to think that there are better regimes we can go to, but we’re not going to do it without due consultation with the markets and with members.”

Later, in the Q&A:

“As we sit here today the average duration — that is the average outstanding stock of assets that are held by the Fed in U.S. Treasuries — is about six and a half years. The Treasuries that \[are\] held in the market are about four and a half years. I understand in periods of crisis like the 2020 pandemic and the ’08 crisis, central banks by design needing to step into markets to create a fair price. But in more benign times when our overall holdings are larger than the market, that’s something where monetary policy is, in the language of former Chairman Volcker, on the edge of its authority.”

Warsh wants the duration and the size of the treasury’s portfolio to decrease.

This is quite similar to the view of Jeremy Stein, the primary citation in Bessent’s speech and now the head of the Fed’s new task force on the balance sheet. Not only that, former governor Stein is also the architect of the liquidity regulations described in Part I.

Chiefly, Stein argues – in parallel to Chair Warsh’s Q&A response above from a month ago – that duration absorption should be a temporary crisis tool, not a standing feature.

The Fed has been discussing this matter for a couple years now; Cleveland Fed President Hammack and Governor Waller have been particularly vocal on it. In fact, minutes suggest it was the central topic at the October 2025 Fed meeting:

“Overall, most participants favored a long-run composition of the SOMA portfolio that matched the composition of Treasury securities outstanding, indicating that a proportional allocation would provide enough flexibility and may be simpler to communicate.

Participants also agreed that a larger share of Treasury bills than the current portfolio allocation would be desirable in the long run.”

The meeting’s Implementation Note subtly directed the NY Fed desk:

“Beginning on December 1, reinvest all principal payments from the Federal Reserve’s holdings of agency securities into Treasury bills.”

After late 2025’s funding stress, the Fed initiated Reserve Management Purchases (RMPs). RMPs buy T-Bills on the open market each month in order to rebuild reserves, keeping repo tight. While this is characterized as liquidity and plumbing operations, at the end of the day it’s still reserves buying new bills. In other words, monetization.

The Dec 1 note also details the reinvestment of maturing agency MBS into bills, which shifts duration even more towards the front end even as the overall book runs off. Active (RMPs) and passive (reinvestment) flows both work towards a proportional SOMA that matches the bill-heavy composition of the treasury market and brings the duration of the Fed’s book more in line with their goals.

Whatever you call it, the Fed is taking actions that result in the accumulation of the exact paper the Treasury will be selling lots more of.

Our ‘Fed twist’, already in action.

A forthcoming Treasury twist that lowers long bond incremental supply and a matching Fed twist that has already begun (and may accelerate) under Chair Warsh.

Taken all together, we can see that Bessent has found his answer to “Who buys the debt?”.

A Tactical Trade on the Long-End

A few reasons stand out as to why Secretary Bessent has made his loudest stand now.

Secretary Bessent communicated on CNBC on Thursday he’d have more detail out publicly on the fiscal consolidation plans in the next week or so. He’s preparing to move – or for a move in – bond prices. Higher in price, lower in yield. That timing seems to align with Jackson Hole and G20 over the coming week.

But the most obvious reason is the recent breakout in the yield of the long bond.

With the recent breakout in yields of long bonds, net short positioning in the long bond is extreme. The COT report from the CFTC shows speculative net positioning in the 30yr Treasury bond, which is at a multiple- standard deviation extreme of the last many years.

The long bond (30yr) is primed for a squeeze, but it’s going to take some doing.

Secretary Bessent knows the appetite for the long bond is less and he needs to make supply more scarce, rotating the supply to T-bills instead. His message – between the yen intervention, the QRA hint, the Treasury buyback signaling, his TV appearance – is that it’s go-time. Game on.

We believe the long bond will stage a rebound, and that calls out to November are underpriced. We’ll initiate a 5y30y flattener to isolate the long end, where the Treasury’s supply cut and the positioning squeeze bite hardest, and expect the trade to materialize in the next 3 months through the next QRA on November 4, by which time we think the ‘Treasury twist’ will become apparent to the market.

As noted previously, we are skeptical that the Bessent/Warsh plane lands smoothly in the medium-to-long term; even with Secretary Bessent’s reductions in the deficit, the outlined sovereign deleveraging is not enough to draw a parallel with the post-WW2 period. Furthermore, it is very possible that a lower long bond yield stimulates the economy (i.e. housing) and promotes credit creation, which could be inflationary.

The flattener trade is a tactical idea, and timing will be crucial. Our global macro positioning remains long gold and short US dollars which we believe will continue to perform.

It is evident that regime change is here for the Fed, the Treasury, their interaction with each other and with the banking system. Whether it works or not remains to be seen, but the implications if we’re right about the plans being executed upon between the Fed and Treasury are broad and far-reaching.

Acknowledgement: Our thanks to Ronick Sanon for his involvement in this article.

Part III – The Plumbing Risk: Rethinking Ample Reserves and the STIR Hedge

Prior to the GFC, the system operated in a ‘scarce reserves’ regime where the banks operated with almost no cash balance day to day. Post the GFC, the system has been purposefully operated in an ‘ample reserves’ regime where banks are flush with liquidity.

Per Warsh’s Congressional testimony in mid-July, the Fed’s balance sheet task force is set to ‘review the ample reserves regime’ as it assesses potential avenues to reduce the Fed’s balance sheet size and duration.

Additionally, we think central clearing adds skew toward wider repo markets, given collateral will need to be posted that wasn’t required prior. The margin of collateral posted will be set by a procyclical risk model that demands more collateral when volatility rises. It is not apparent that this is priced into repo spreads, given central clearing is still in the comment period pending implementation for repo by end-June 2027.

We are also mindful that the September 2027 debt ceiling deadline could present funding challenges. (Note, the initial effect of the debt ceiling actually draws down the TGA which is liquidity positive for banks – the real challenge is after the debt ceiling is lifted and the TGA draws those funds back. Further macro spillover concerns can lead to financial volatility.) The last shutdown – the longest in history – occurred in a Republican- dominated Congress, and betting markets currently price a split Congress in 2027.

Meanwhile, given some small stresses in funding markets in 2H25, the Fed began RMPs in December, basically buying Treasury bills in the open market every month to soothe imbalances in the short-term funding markets. The Fed’s current T-bill buying has kept repo / STIR markets very tight.

There is convexity out the STIR curve to some widening from a potential reserve regime shift. It has not priced this in because it is so early the outcome likely hasn’t even been decided yet; Warsh only convened his

We would buy 1y2y SOFR-FF swaps to isolate incremental funding market stress getting priced in between Aug ‘27 – Aug ’29.

These are both overnight instruments so implicitly we take no duration risk, just SOFR-FF curve risk, and should get paid some term premia to hold the position. If we are wrong, the loss is small, and if any funding stress from any of the above gets priced, we could win big. We isolate summer ’27 to summer ’29 because liquidity reform is shaping up to be a 1Q27 proposal after the balance sheet task force reports to the Fed in December.