No wonder the Federal Reserve is 'panicking'? Just like six years ago, Wall Street's largest banks are no longer depositing money.

JPMorgan has withdrawn roughly $350 billion from its Federal Reserve account since 2023, cutting its Fed deposits from $409 billion at end-2023 to about $63 billion in Q3 while increasing U.S. Treasury holdings from $231 billion to $450 billion. The bank’s large reallocation is cited as a key driver of a sharp decline in aggregate bank reserves (U.S. bank Fed deposits down from $1.9 trillion to ~$1.6 trillion) and recurring repo-market stress, prompting the Fed to launch Reserve Management Purchases (around $40 billion/month in T-bills) and signal a dovish easing cycle; the article notes Fed interest-on-reserves payouts of $186.5 billion in 2024 and highlights political scrutiny (Sen. Rand Paul) after JPMorgan earned an estimated $15 billion in Fed interest in 2024 versus $58.5 billion in total profit.

Analysis

Market structure: JPMorgan’s $~350bn shift from Fed balances into Treasuries (Fed reserves down < $3.0tn) reallocates a very large marginal buyer into duration markets while simultaneously draining lendable reserves in the repo plumbing. Winners: long-duration Treasuries, primary dealers/asset managers holding bills (RMP supports bills at $40bn/month). Losers: smaller banks/wholesale funding borrowers, repo users, and banks reliant on excess reserve interest income (regional banks, parts of BAC).

Risk assessment: Tail risk remains a replay of 2019—single-bank liquidity moves triggering repo dislocations and emergency Fed operations; probability ~10–20% over next 6 months if reserves slip further or deposit dynamics accelerate. Immediate (days) risk: repo volatility and front-end bill yield moves; short-term (weeks–months): bank earnings volatility and regulatory scrutiny (Senate attention could lead to IOER/interest-on-reserves tweaks). Hidden dependency: concentrated balance-sheet management by a few GSIBs creates nonlinear system risk via collateral rehypothecation chains.

Trade implications: Expect downward pressure on yields as big banks lock into Treasuries and Fed’s RMP bids for bills cap bill yields — favor medium-duration long Treasuries (2–10yr) as a core hedge for 1–3 months and short bank equity beta (regional banks, BAC) on funding-margin compression. Use relative-value: long JPM vs short BAC/KRE to express flight to scale and deposit stickiness, and buy protection (puts) on the KBW bank index for tail hedging if repo rates spike.

Contrarian angles: Consensus views JPM as the villain and will short JPM; that may be overdone — JPM’s move is strategic risk management locking yields, not clear earnings impairment short-term. Markets may underprice the stabilizing effect of a large Treasury buyer plus Fed RMP ($40bn/month) — if reserves stabilize above ~3.2–3.5tn the “scarce reserves” narrative could reverse quickly and produce a snap tightness unwind. Watch thresholds: reserves <3.0tn (risk-on liquidity shock) vs >3.5tn (risk-off unwind).

Key Decisions for Investors

• Establish a 2–3% portfolio position long iShares 7-10yr Treasury ETF (IEF) for 1–3 months as a tactical yield-fall hedge; target 25–50bp fall in 7–10yr yields, take profits if yield falls ≥40bp or close if yields rise >30bp from entry.
• Implement a relative trade: long JPM (2% weight) and short BAC (2% weight) or short KRE (regional bank ETF) for 3 months to capture scale-loss vs margin pressure; set stop-loss at 12% adverse move on either leg and trim when JPM outperforms by 300bps.
• Buy a protective put spread on the KBW Bank Index (3-month) to cap tail risk from repo dislocation (buy 5–10 delta put, sell 2.5–3 delta lower strike) sized to cover 1–2% portfolio drawdown exposure; unwind if Fed reserves recover above 3.5tn.
• Allocate 3–5% to ultra-short Treasury/bill ETFs (BIL or SHV) for liquidity runway and roll into longer Treasuries if the Fed’s RMP keeps bill yields depressed; rebalance if 3-month bill yield moves >50bp.
• Monitor three catalysts over next 30–90 days before scaling: Fed weekly H.4.1 reserves (thresholds: <3.0tn = add protection; >3.5tn = reduce hedges), monthly RMP purchase cadence ($40bn/month) and any legislative proposals on IOER affecting bank economics.