沃什要缩表？美银：恐怕只是「雷声大雨点小」

Original Author: Xu Chao

Original Source: Wall Street CN

Market concerns about new Fed Chair Kevin Warsh's aggressive balance sheet reduction may have significantly overestimated what he can actually achieve.

Warsh was recently confirmed by the Senate as Fed Chair. Due to his long-standing criticism of the Fed's massive balance sheet, the market widely expects him to rapidly push for large-scale quantitative tightening. However, in a research report published on May 18, Bank of America rate strategists Mark Cabana and Katie Craig stated bluntly: In terms of both scale and composition, the room Warsh can actually maneuver is extremely limited, and the direct impact on the market is expected to be near zero.

Bank of America's core judgment is: In terms of balance sheet size, the Fed completed quantitative tightening normalization in Q4 2025. To further reduce the size, it would need to shrink one of three core liabilities—currency, the Treasury General Account (TGA), or reserves—and the only one Warsh can meaningfully operate on is reserves. The path is limited and the pace is slow.

Regarding asset composition, the MBS reinvestment plan is already underway and fully priced into the market. Any impact from compressing the weighted average maturity (WAM) of Treasuries is offset by market mechanisms, trending towards zero. Neither constitutes a tightening of financial conditions nor triggers rate cut signals.

The report also presents a scenario not yet widely considered by the mainstream market: If the Standing Repo Facility (SRP) rate for banks is set equal to the Interest on Reserve Balances (IOR) rate, coupled with reducing information disclosure requirements to lower the "stigma" effect, it could more effectively reduce bank reserve demand, thereby creating genuinely operational space for balance sheet reduction. Bank of America believes the actual impact of this approach could exceed the market's traditional expectations for Warsh's tightening path.

Scale Dilemma: Of the Three Big Liabilities, Only Reserves Can Be Acted Upon

After completing QT normalization, the Fed's balance sheet stands at approximately $6.78 trillion, driven by the liability side. The three core liabilities are: Reserves (approx. $3.12 trillion, 46% of total), Currency (approx. $2.46 trillion, 36%), and TGA (approx. $807 billion, 12%).

Currency is considered an "exogenous" liability by central banks, beyond the reach of policy tools. BofA notes that theoretically, it could be compressed by eliminating high-denomination banknotes, but "that's not happening in the US."

Regarding the TGA, the Treasury has already indicated it has no intention of reducing it. BofA expects the TGA to rise to $900 billion by the end of Q2 2025 and further to $950 billion by the end of Q3 2025.

BofA believes that marginal adjustments to the TGA via repo investments are possible, but the impact would be minimal; adjustments through the Treasury Tax and Loan (TT&L) accounts are "highly unlikely."

Reserves are the most realistic option for Warsh to shrink the balance sheet, but each path has its constraints.

A "bank-unfriendly" approach—such as setting reserve caps or tiered interest rates—would compress bank liquidity, weaken market-making and lending appetite, and subsequently drag on the economy. BofA believes Warsh is unlikely to choose this path.

A "bank-friendly" path involves relaxing regulations and allowing banks to pre-position collateral at the discount window to expand High-Quality Liquid Assets (HQLA), thereby reducing reserve demand.

BofA estimates this path could ultimately yield a reserve reduction of about $200 billion to $500 billion, but the process would be slow. And since it wouldn't tighten financial conditions, it wouldn't constitute a reason to cut rates.

Composition Adjustment: WAM Compression Impact Nullified by Offset Mechanisms

At the asset composition level, Warsh's operational space is also constrained by mechanisms.

The Fed currently holds about $1.98 trillion in MBS. It is gradually reducing this by allowing maturing and prepaid MBS to roll off and reinvesting the proceeds into Treasury bills, at a pace of $10-20 billion per month.

BofA believes the possibility of selling MBS is very low (unless directly repurchased by Fannie Mae or Freddie Mac, which is seen as a low-probability event). The current approach is already fully priced in by the market and does not constitute a new source of disruption.

Shortening the Weighted Average Maturity of Treasury holdings is another focus area.

Warsh could steer reinvestment of maturing Treasury coupon proceeds away from the current proportional allocation across all maturities, concentrating instead on short-term instruments (e.g., 2 to 3-year Treasuries) to accelerate WAM compression. However, the Fed executes its reinvestments using an "add-on" approach in auctions—meaning it participates in auctions to add directly to the supply of short-term Treasuries, rather than replacing longer-term ones.

This raises a key question: Will the Treasury offset the Fed's shorter WAM by adjusting its own debt issuance structure? BofA's answer is no. If this assessment holds, and the Fed's WAM compression has zero net impact on the overall Treasury market and financial conditions, Warsh would have no logical basis to push for rate cuts on this premise.

Ample Reserves: Warsh Lacks Both Will and Ability to Change This

In BofA's view, the most critical question regarding Warsh is: Will he favor an "ample" or "scarce" reserve regime? BofA's answer is ample, with a high degree of certainty.

The advantage of an ample reserve system lies in its ease of operation, ensuring sufficient funding for the banking system, curbing money market volatility, and supporting relatively loose financial conditions, with the only cost being a slightly larger balance sheet. In contrast, while a scarce regime could further shrink the balance sheet size, it would introduce real risks like increased money market volatility and liquidity pressure.

BofA provides two supporting points. First, President Trump places far greater importance on loose financial conditions than on the size of the Fed's balance sheet, and Warsh is expected to be receptive to this policy preference.

Second, the Fed formally adopted the ample reserve regime in 2019, which is fully supported by the current leadership, with some officials holding quite strong views. BofA cites Fed Governor Waller from a speech in February 2026: "You don't want banks to be looking for money under sofa cushions every night... That is extremely inefficient and extremely stupid." BofA's report specifically highlighted the last word.

BofA believes that Warsh is not only subjectively inclined towards the ample system but will also be objectively constrained by the consensus within the Fed.

SRP=IOR Mechanism Could Be the True Game-Changer

BofA's report proposes a mechanism reform beyond the traditional framework, drawing on previous comments by Dallas Fed President Lorie Logan: Setting the Standing Repo Facility (SRP) rate for banks equal to the Interest on Reserve Balances (IOR) rate.

Specifically, banks could pledge Treasuries or agency debt to the Fed around the clock in exchange for cash, at a price equal to their deposit rate. This would function similarly to a discount window but be open 24/7 without discrete operational time windows.

Because the rate is competitive and carries no premium, banks would be more willing to use it, consequently reducing their need to hold large precautionary reserve buffers. This would create operational space for the Fed to shrink its balance sheet. The Bank of England currently employs a similar mechanism.

To further enhance the effect, BofA suggests reforming the information disclosure system, specifically by eliminating the current weekly report that discloses the distribution of reserves across regions. This report is currently used by market participants to track institutions potentially facing liquidity pressure. Removing it would effectively reduce the "stigma" associated with using the SRP tool, making banks more willing to use it when needed, rather than hoarding excess reserves.

The report also notes that the Bank SRP and Dealer SRP should be distinct: The Dealer SRP rate should be set about 5 to 10 basis points higher than the Bank SRP rate to ensure banks have an incentive to lend funds in the repo market, while preserving pricing space for free-market transactions.

BofA concludes that the "Bank SRP=IOR" mechanism, combined with reforms to the reporting system, could substantially lower bank reserve demand, thereby providing Warsh with a truly viable path to compress the balance sheet. This approach has not yet entered mainstream market discussion, but BofA expects it will eventually garner widespread attention—its impact potentially far exceeding the market's current conventional estimates of Warsh's ability to shrink the balance sheet.