The Fed’s Overton Window Is Shifting

From supply-driven reserves to demand-driven liquidity.

What do the Bank of England (BoE), the Reserve Bank of Australia (RBA), the European Central Bank (ECB), the Riksbank, the Reserve Bank of New Zealand, and the Bank of Canada all have in common today?

Over the past few years, these central banks have begun shifting toward demand-driven operating systems. Some, such as the RBA, ECB, and BoE, have conducted formal reviews of their frameworks, while others have moved more incrementally in the same direction. These operating systems rely more heavily on repo and lending operations as the primary elastic suppliers of liquidity and place greater weight on interbank lending to redistribute reserves across the banking system.

We have covered these developments both in this newsletter and on the podcast, including conversations with Isabel Schnabel about the ECB’s demand-driven ‘soft floor’ system, Laurie Bristow on the RBA’s demand-driven full allotment system, Bill Nelson on the BoE’s demand-driven repo system, and Per Åsberg Sommar on the Riksbank’s demand-driven, state-contingent operating system.

What unites these central banks is a growing recognition that interbank markets should play a larger role in reallocating liquidity, while central bank lending facilities should serve as elastic backstops that allow reserves to expand on demand rather than being pre-positioned in advance. In short, the global trend is away from balance sheet management and toward price-based, demand-driven liquidity provision.

The notable exception to this emerging consensus is the Federal Reserve. Its ample reserve system discourages interbank lending and relies on Reserve Management Purchases (RMPs) to grow reserves over time. While this approach may appear demand-driven too, it is fundamentally different. The Fed manages reserve quantities in advance through asset purchases, whereas its peers supply reserves elastically through routine repo and lending operations that respond to demand in real time.

But that may be changing. A number of recent official speeches suggest the Fed is poised to join the demand-driven central bank club. These remarks point to a potential break from the Fed’s long-standing commitment to its supply-driven ample reserve framework and toward a leaner, more demand-driven approach. The Overton window on the Fed’s operating system appears to be shifting. This is a huge change.

In the remainder of this newsletter, I will review the speeches that indicate there is a growing openness toward a more demand-driven operating system and then discuss why this change is warranted.

A Growing Openness at the Fed

The speeches below reveal a marked shift over the past year in policymakers’ openness toward a more demand-driven operating system. There are likely other forces contributing to this change, but my focus here is on the visible change in how Fed officials are thinking about the Fed’s operating framework.

The first signs of a shift appeared in late summer 2025. In an August 25 speech, Dallas Fed President Lorie Logan reaffirmed her support for ample reserves but suggested, in the spirit of a demand-driven framework, that if the Fed wanted to minimize its financial footprint, it should do so by supplying reserves through "ceiling tools... and temporary operations" rather than maintaining large outright holdings. A month later, Fed Vice Chair for Supervision Michelle Bowman went further, expressing a preference for reserves “closer to scarce than ample” and arguing that lower reserve levels would improve market functioning and reduce the Fed’s footprint. Together, these remarks signaled that the Fed’s supply-driven ample-reserves framework was no longer beyond question.

The shift became more consequential on March 3, 2026. Governor Bowman directly challenged the post-2008 liquidity framework, arguing it may be encouraging liquidity hoarding and weakening the role of the Fed’s Discount Window (DW). At the same time, Treasury Secretary Scott Bessent and FDIC Chairman Travis Hill proposed allowing DW capacity to count toward liquidity requirements. While this idea had been raised before, the fact that three key policymakers advanced it simultaneously marked a turning point. Their speeches helped shift the focus from how the Fed supplies reserves to why banks demand so many of them. Regulatory policy and monetary policy implementation were now intersecting in public discourse in a way that pointed toward a more demand-driven system.

A more explicit shift emerged in late March and early April, as FOMC participants began outlining how to actually shrink the Fed’s balance sheet. In a March 26, 2026 speech, Governor Stephen Miran argued that a smaller balance sheet is both desirable and feasible, and that reducing banks’ demand for reserves is central to achieving it. He also released a coauthored paper, A User’s Guide to Reducing the Federal Reserve’s Balance Sheet, which provides a menu of policy options. Taken together, these reforms—many of which work by lowering reserve demand—could allow the Fed to shrink its balance sheet by roughly $1 to $2 trillion.

Then on April 2, President Lorie Logan delivered a speech and a coauthored paper with Sam Schulhofer-Wohl that also provided a menu of policy options for reducing the Fed’s balance sheet. They explicitly distinguished between returning to scarce reserves versus “shifting the demand curve inward” by reducing banks’ need for reserves. Their preferred approach is the latter option since it keeps the ample-reserves framework but makes it leaner and more demand-driven, one defined by spreads rather than the pre-positioning of large quantities of reserves.

Taken together, these developments suggest the Fed’s once sacrosanct, supply-driven ample-reserves system is likely to be replaced by a leaner, demand-driven operating system. It likely helps that the incoming Fed chair, Kevin Warsh, has also expressed support for rethinking the Fed’s balance sheet and operating framework. And, perhaps, the Macro Musings podcast’s relentless coverage of these issues and this newsletter has played some small role too.

To be clear, a shift to a more demand-driven operating system will take time and it will require the Fed to transform its ceiling facilities into business-as-usual sources of liquidity. But the path is now being set and the destination is becoming increasingly clear.

Why This Change is Warranted

So it appears the Fed may be moving toward a more demand-driven operating system. But is this shift warranted? While such a transition may make sense for other central banks, it is worth asking whether it also makes sense for the Federal Reserve.

In a recent Macro Musings podcast with Bill Nelson, we revisited the challenges of the Fed’s supply-driven ample-reserves system and the case for moving toward a more demand-driven approach. From that conversation, four key reasons emerge for why such a transition may be desirable: (1) a revival of interbank lending, (2) reduced liquidity fragility, (3) stronger bank lending to the real economy, and (4) greater central bank independence. Let’s consider each of these in turn.

(1) Revival of Interbank Lending

A central motivation behind the shift toward demand-driven systems abroad is the desire to revive interbank lending. In these frameworks, central banks want private money markets to play a larger role in allocating liquidity across banks, thereby enhancing price discovery and market discipline. By contrast, the Fed’s supply-driven ample-reserves system largely satiates banks’ demand for liquidity, leaving little need for an active interbank market.

This has important consequences. With abundant reserves readily available, banks have less incentive to borrow from or lend to one another. As a result, interbank activity atrophies over time. At the same time, limited use of private funding markets reduces banks’ use on the Fed’s ceiling facilities. This further reinforces banks’ preference to self-insure through large reserve holdings rather than engage in market-based liquidity management.

(2) Reduced Liquidity Fragility

Raghuram Rajan, in a series of papers with Viral V. Acharya, Rahul S. Chauhan, and Sascha Steffen, argues that the supply-driven ample-reserves system may itself be partly a byproduct of distortions created by QE that increase liquidity fragility.

They argue that as reserves flood into the banking system during QE, banks become more comfortable funding themselves with demandable deposits and other runnable liabilities while reducing more stable funding sources. The reserves created by QE are effectively treated as a backstop for these liabilities, leading banks to expand claims on liquidity even as reserves increase. As a result, the increase in reserves is offset by a corresponding increase in liquidity claims on banks, leaving net liquidity little changed.

Evidence from Rajan and his coauthors shows that these runnable liabilities rise during QE and do not unwind proportionately during QT. In other words, as reserves decline, these runnable liabilities remain causing net aggregate liquidity to shrink faster than headline reserve balances suggest. The result is a system that appears liquid on the surface but is more fragile in practice. This fragility helps explain why QT fails to fully offset QE and instead contributes to a ratchet effect on the size of the Fed’s balance sheet.

In this sense, the supply-driven ample-reserves system both reflects and reinforces a growing dependence on central bank liquidity. While it helps accommodate the higher demand for reserves created by QE-induced balance sheet distortions, it also entrenches that dependence by sustaining an environment of continuously ample reserves. A shift toward a leaner, demand-driven framework could help reduce this fragility by encouraging more active liquidity management and reducing reliance on persistently high reserve levels.

(3) Stronger Bank Lending to the Real Economy

Another concern with the ample reserve system, as shown by William Diamond, Zhengyang Jiang and Yiming Ma, is that large injections of reserves can crowd out lending by raising the marginal cost of expanding bank balance sheets. Thomas Hogan reports similar findings. These results are complemented by Raghuram Rajan et al.’s work on liquidity dependence, discussed above, which shows that QE-induced reserves are largely financed with runnable deposits and accompanied by a shift away from term funding toward short-term liabilities.

This shortening of bank liabilities makes banks more liquidity-sensitive and less willing to engage in maturity transformation, the very activity that underpins traditional bank lending. In effect, the asset-side intent of QE—to encourage longer-term lending—is partially offset by a liability-side response that pushes banks in the opposite direction. The result is a banking system that is less inclined to supply medium and longer-term loans.

(4) Greater Central Bank Independence

One of the key advantages of an ample-reserves system is that it gives the Fed two independent monetary policy levers: its policy rate and its balance sheet. That is, the Fed can adjust its target policy rate without having to change the size of its balance sheet. Alternatively, if the banking system is under stress and needs an injection of liquidity, the Fed can expand its balance sheet without altering its policy rate. Prior to 2008, this separation was not possible; the policy rate was tightly linked to the size of the balance sheet.

This feature of the ample-reserves system, however, is a double-edged sword. While it provides the Fed with greater flexibility, it also creates a temptation for Congress and other political actors to view the balance sheet as a policy tool for broader fiscal or credit allocation goals. If the size of the balance sheet does not affect the policy rate, why not use it to support student debt cancellation or finance other politically salient initiatives?

Moreover, the supply-driven ample-reserves system can create the misleading impression that the Fed is subsidizing banks and is subject to large operating losses, especially after encountering the zero lower bound. These perceptions invite political criticism and further undermine the Fed’s independence.

Conclusion

Taken together, these considerations suggest that the case for a more demand-driven operating system is not merely about aligning with global trends, but about addressing real shortcomings in the Fed’s current supply-driven ample reserve operating system. A leaner, demand-driven system would not eliminate all risks or tradeoffs, but it would move the Fed toward a more resilient, market-oriented, and politically sustainable operating regime.