Four Tools, One Target: Absorbing the Shock the 2019 Repo Market Couldn't Take

The Toolkit as a Set of Failure Points, Not a Single Machine

Most explanations of the Federal Reserve's short-term rate toolkit describe it as a system for pinning down one number, the federal funds rate. That is true but incomplete. A more useful way to think about interest on reserve balances (IORB), the overnight reverse repo facility (ON RRP), the standing repo facility (SRF), and the relationship between the effective federal funds rate (EFFR) and the Secured Overnight Financing Rate (SOFR) is as four separate patches, each sewn onto the system after a specific kind of failure was discovered. Understanding what each patch actually fixes, and what it does not fix, explains both how the floor system works day to day and why repo markets still cracked open in September 2019.

The Floor Itself: IORB as the Anchor

Banks hold reserve balances, which are deposits banks themselves keep at the Fed. Since 2008 the Fed has paid interest on those balances, a rate called IORB. The logic is simple: no bank should lend reserves to another bank overnight for less than it can earn risk-free by leaving them parked at the Fed. IORB therefore acts as a magnet pulling the federal funds rate, the rate banks charge each other for overnight loans of reserves, up toward it.

This only works cleanly if reserves are abundant. When reserves are scarce, banks compete for them and rates can spike above IORB, which is exactly what happened in September 2019 and is discussed below. When reserves are abundant, the opposite risk appears: rates can drift below IORB, because non-bank lenders in the market do not earn IORB and are willing to lend at less. That gap is where the second tool comes in.

The Backstop Below: ON RRP

Money market funds and other non-bank cash lenders cannot hold accounts at the Fed and cannot earn IORB directly. Left alone, they might accept very low rates just to park cash somewhere safe overnight, dragging the whole market down with them. The ON RRP facility gives these institutions a direct, risk-free place to lend cash to the Fed itself at a fixed rate, set below IORB. This creates a floor beneath the floor: no sensible cash lender should accept a rate lower than what the Fed itself offers overnight.

Together, IORB and ON RRP form a band. IORB pulls bank-to-bank lending rates up from above, ON RRP pulls broader market rates up from below. The effective federal funds rate and SOFR, which measures the rate on overnight loans backed by Treasury collateral in the much larger repo market, are both meant to settle inside this band under normal conditions.

The Gap That 2019 Exposed

September 2019 revealed what these two tools do not cover: a sudden shortage of reserves at the exact banks that need them most. In the years before, the Fed had been shrinking its balance sheet through quantitative tightening (QT), reducing the total stock of reserves in the banking system. Reserves were still large in aggregate, but they were unevenly distributed, concentrated at a handful of big banks that were not necessarily willing or able to redistribute them quickly through the market.

On September 16 and 17, 2019, two events collided. Corporate tax payments pulled cash out of bank accounts and into the Treasury's account at the Fed, and a large settlement of Treasury securities required dealers to fund a wave of new collateral. Both drained reserves from the system at the same moment. Banks that needed overnight cash to cover their positions found few willing lenders at normal rates. SOFR, which tracks repo market funding costs directly, spiked to nearly 5.25 percent intraday, roughly triple its recent level, and the effective federal funds rate briefly poked above the top of its target range. The mechanism that failed was not IORB or ON RRP. It was the absence of a reliable, pre-committed channel for banks holding Treasury or agency collateral to convert it into cash at a known rate, on demand, without waiting for the market to clear on its own.

The Fix: A Standing Facility for Exactly That Gap

The Fed's answer, formalized in July 2021, was the standing repo facility. The SRF lets a broad set of eligible banks and primary dealers post Treasury, agency debt, or agency mortgage-backed securities as collateral and borrow cash overnight at a fixed rate set above IORB, on a recurring daily basis. It is designed to be used routinely enough that borrowing from it carries no stigma, unlike the Fed's traditional discount window, which banks have historically avoided using for fear signaling weakness.

The SRF caps how high repo rates can rise, the same way ON RRP caps how low they can fall. If a dealer cannot find funding anywhere else at a reasonable rate, the SRF rate becomes the ceiling, because no rational borrower pays more when a known, always-available source sits right there. In principle, a September 2019-style spike should now hit that ceiling and stop, rather than spiraling as it did before the facility existed.

Why EFFR and SOFR Still Move Independently

EFFR reflects unsecured lending among banks in the federal funds market, a market that has shrunk over time. SOFR reflects secured lending backed by Treasury collateral in the much larger repo market, and it is more sensitive to swings in the supply of that collateral, to dealer balance sheet constraints around quarter-end and tax dates, and to Treasury issuance patterns. The two rates usually move together inside the IORB-to-ON RRP band, but SOFR is the more volatile of the pair precisely because it sits closer to the plumbing that broke in 2019.

What to Take Away

The Fed's toolkit is not one lever but four purpose-built valves: IORB anchors the floor from above, ON RRP reinforces it from below, the effective funds rate and SOFR are the readings that show whether the system is working, and the standing repo facility exists specifically to cap stress events that the other tools cannot reach. September 2019 happened because reserves were scarce and unevenly held, and no standing, stigma-free channel existed to move collateral into cash fast enough. The SRF was built to close exactly that gap. Whether it fully succeeds in a future, larger shock is something analysts will only be able to judge in hindsight, tied to whatever specific episode tests it next.