Repo Spike and the Fed's Corridor
At a quarter-end, the federal funds rate trades at 5.10% against a target range of 2.25–2.50%. SOFR jumps from 2.30% to 5.25%. The Fed's balance sheet has shrunk by $1.2 trillion; reserves are $1.5 trillion, or 7% of GDP, down from 18% two years earlier. IORB is 2.40%; ON RRP is 2.30%. Treasury settlement and month-end dealer balance-sheet constraints coincide. The desk has a standing repo facility (SRF) with a 2.50% offer rate but stigma prevents use. Banks hold $3.1 trillion in high-quality liquid assets. Dealer repos are scarce because quarter-end window dressing reduces intermediation. The FOMC wants to maintain the current target range without resuming aggressive asset purchases, and the next meeting is in three weeks.
1. Diagnose the spike. Is it a reserve shortage, a collateral problem, or a balance-sheet constraint? What evidence supports each?
Read the worked answer
Three diagnoses are consistent with the data; the art is weighting them.
- Reserve shortage: reserves are 7% of GDP, down from 18%. In an ample-reserves framework that is still large in absolute terms, but quarter-end balance-sheet windows can make the effective supply scarce for a few days. The fed funds rate above IORB and SOFR at 5.25% while IORB is 2.40% shows that banks are paying more than the administered rate for reserves.
- Collateral problem: if dealers cannot pledge the right collateral or haircuts widen, repo rates spike even with abundant reserves. Treasury settlement absorbs collateral.
- Balance-sheet constraint: SLR and dealer intermediation limits can make balance sheet scarce at quarter-end. The coincidence of month-end and settlement points to this.
The spike is a temporary money-market plumbing problem, not proof that the entire reserve stock is too low. But repeated spikes indicate the framework needs a standing facility that converts collateral into reserves at a controlled rate.
2. Propose a sequence of operations and explain the effect on the corridor.
Read the worked answer
Sequence:
- Conduct open market repo operations against Treasury and agency collateral at a rate inside the target range. This adds reserves temporarily and directly caps the repo rate.
- Activate or lower the SRF offer rate so it is a ceiling near the top of the target range, and broaden eligible collateral. The SRF is designed to reduce stigma by being available to all eligible counterparties.
- If pressure persists, slow or pause balance-sheet runoff and adjust IORB/ON RRP spreads. Lowering the ON RRP rate can push money out of the facility, but that is not the first tool when the problem is quarter-end collateral.
- Communicate that the desk will use temporary operations, not a new QE program.
Effect on the corridor: a well-functioning ceiling keeps repo and fed funds inside the range. If the SRF is above the target range, it does not cap the rate; if it is below, it can become the effective floor for eligible collateral. The desk should also publish usage data to reduce stigma.
3. What are the long-run implications for the Fed's operating framework: scarce reserves or ample reserves?
Read the worked answer
The choice is between a scarce-reserves corridor and an ample-reserves floor. In a scarce-reserves system, the desk manages the reserve supply daily and rates are more volatile; the corridor is defined by the discount rate and the deposit rate. In an ample-reserves system, the central bank supplies enough reserves that the opportunity cost of holding them is near the administered rate; the floor is IORB and the ceiling is a standing lending facility.
The recent spike shows that an ample-reserves framework still needs a standing repo ceiling and attention to quarter-end balance-sheet costs. The Fed should not return to scarce reserves just to shrink the balance sheet, because that raises payment-system risk and makes the fed funds target harder to control. The better path is a large balance sheet with a well-designed ceiling, broad collateral, and a clear reaction function. The framework should be judged by control of the policy rate, not by the size of the balance sheet alone.