Is the Fed really going to shrink its balance sheet? Citi: Watch closely whether the reserve requirement will be lowered.

```

The Federal Reserve’s balance sheet policy is at a new crossroads. As Kevin Warsh is about to preside over his first FOMC meeting as the new chairman, the market is increasing speculation about the Fed’s quantitative tightening (QT) path.

Citi Research believes that Warsh is unlikely to make any major announcements on balance sheet policy at this meeting, but the medium-term direction is already clear—if QT is to resume, a key prerequisite must first be met: reducing the banking system’s demand for reserves.

According to ZF Trading Desk, in a June 12 report, Citi Research analysts Alejandra Vazquez Plata and Jason Williams point out that the order of operations is critical. Before restarting any QT, the Fed and regulators need first to lower banks’ demand for reserves through regulatory reforms, then watch for confirmation signals from the market before advancing further balance sheet reduction. Citi expects these first two steps to inject relatively ample liquidity into the banking system over the next few years, supporting the short- and medium-term yield spreads.

For investors, the core trading logic is that reducing reserve demand will lead to a temporary rise in systemic liquidity, benefiting short-term yield spreads. Citi’s scenario analysis shows domestic banks’ reserve levels could fall by $200–600 billion, but this space is limited by the behavior of foreign banks and regulatory constraints.

Warsh will not make “big moves” at his first FOMC meeting

Citi believes it is not the right time for Warsh to launch a full-scale adjustment of balance sheet policy at his first FOMC meeting.

In April this year, Warsh testified before the Senate Banking Committee, saying he would "work with the Treasury Secretary" to shrink the Fed’s balance sheet, but he gave no specific numerical target, and emphasized that any change would be “fully communicated” to the market in advance.

Citi also cited former FOMC member Stephen Miran’s view that QT can serve as a justification for rate cuts—Miran has argued that trimming the balance sheet tightens monetary policy, so as long as the federal funds rate is not near its zero lower bound, the policy rate should be lowered to offset. This framework may also shape Warsh’s policy thinking.

A Necessary Precondition for QT: Reserve Demand Must Fall First

Citi stresses that the sequence of operations is the core variable in this round of balance sheet adjustments. The path includes three steps:

First, lower banks’ demand for reserves via regulatory reform. Citi thinks the most likely move is to modify the way the liquidity coverage ratio (LCR) is calculated—regulators are now considering including the Fed’s liquidity tools (such as the discount window and standing repo facilities) in the recognition of high-quality liquid assets (HQLA). Once implemented, banks could convert some reserves into short-term Treasuries, pushing swap spreads wider at the front end. Citi expects relevant proposals could be released as early as this summer.

Additionally, including standing repo facilities (SRPs) in central clearing is also an option. Roberto Perli, SOMA head at the New York Fed, has said that central clearing for SRPs will lower balance sheet costs for repo market intermediaries, and help “further improve counterparty participation and the SRPs’ effectiveness in containing rate pressures and maintaining market stability.”

Second, wait for and confirm that reserve demand has substantively declined. Citi notes observable signals include: repo rates dropping below the interest rate on reserve balances (IORB), persistent increases in overnight reverse repo (ON RRP) usage, and lower advances from the Federal Home Loan Bank. Perli previously said that if reserve demand drops, “the federal funds rate may soften, repo rates may fall significantly, ON RRP usage may rise, or FHLB advances may fall.”

Third, only after sufficient evidence, proceed to further QT. Citi expects that if RMP (Reserve Management Purchases) are still ongoing, these would be stopped before restarting balance sheet normalization. Citi also points out that the degree of QT is unlikely to push the balance sheet below its projected level for Q4 2025—meaning the Fed would not shrink the balance sheet so much that swap spreads would drop further below the Q4 2025 level.

How Much Can Reserve Demand Drop? Space Is Limited

Citi built three scenarios using individual bank data to estimate the potential for cutting domestic banks’ reserves.

By Q1 2026, G-SIBs’ (Globally Systemically Important Banks) average reserve/asset ratio is 11.6%, but individual variance is wide—the lowest at just 1.8%, the highest at 26.5%. Non-G-SIB banks (assets >$100bn) are at 9.2%, and banks under $100bn in assets at 6.0%.

In the most aggressive scenario, if all banks’ reserve/asset ratios are set at 4.9% (the 25th percentile for large non-G-SIBs), total domestic bank reserves would fall by about $700 billion from Q1 levels. If the floor is each bank’s lowest ratio of the past three years, the drop is about $426 billion; if based on the three-year average, the drop is about $116 billion. Overall, Citi thinks the feasible compression range for domestic bank reserves is $200–600 billion.

Citi also points out several limiting factors: foreign banks hold substantial reserves, but their behavior is unlikely to change due to domestic regulation, setting a floor on demand; banks with less than $100 billion in assets hold about $300 billion in reserves, with little further scope to compress; G-SIBs have highly uneven reserve distribution, for which the main drivers are unclear.

Can RRP Return to the Spotlight? Much Larger Demand Reduction Is Needed

Citi also evaluated whether the coming debt ceiling cycle in 2027 could see a sustained rise in ON RRP usage if reserve demand falls.

The logic is: once Treasury exhausts extraordinary measures, the TGA balance will fall, boosting reserves and injecting liquidity into the system. Citi estimates extraordinary measures could begin in summer 2027, with the “X date” in Q4 2027.

However, Citi’s "net effective supply model" (private holdings of bills minus Fed liquidity) shows that if reserve demand falls by only $500 billion, there would be no RRP usage in 2027. The model suggests reserve demand needs to drop about $800 billion for RRP to return meaningfully in 2027. Another Citi model, which only uses Fed liquidity as an explanatory variable, shows a $500 billion drop in demand could bring $300 billion in RRP usage—but Citi thinks this model likely overestimates RRP usage, and they will continue using the net effective supply model as their main framework.

Recent Repo Market: Temporary Easy Liquidity, Volatility Ahead

On recent market dynamics, Citi notes that repo rates this week remain soft, as expected. Although SOFR fixings are relatively low, the August SOFR/FF spread became more negative this week. Citi thinks this reflects expectations for a higher SOFR fixing in July–August.

Key drivers include larger Treasury auction sizes and a rising TGA balance (approaching $1 trillion by late July)—both could push SOFR fixings higher in late July. In the short term, Monday’s $80 billion coupon settlement and quarterly corporate tax payments will push up the TGA balance temporarily and reduce reserves, with SOFR potentially approaching IORB, but Citi expects this volatility to be a temporary disturbance.

Additionally, the Fed is maintaining the RMP (Reserve Management Purchases) pace at $10 billion per month. Citi’s base expectation is to continue at that pace for the year, but if repo rates rise sharply (approaching the top of the fed funds target range), purchases could accelerate; if repo rates remain soft, purchases could be paused temporarily.

~~~~~~~~~~~~~~~~~~~~~~~~

The above content comes from ZF Trading Desk.

For a more detailed analysis, including real-time coverage and frontline research, please join [ZF Trading Desk Annual Membership]

Risk Warning & DisclaimerThe market carries risks, and investments should be made with caution. This article does not constitute individual investment advice and does not take into account the specific investment objectives, financial situations, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article fit their specific circumstances. Investment based on this information is at your own risk. ```