Overnight Reverse Repurchase Agreements: How the Fed's RRP Facility Works

I've spent years analyzing the Fed's toolkit, and one instrument consistently surprises most people: overnight reverse repurchase agreements (ON RRP). It's not flashy like rate hikes or QE, but it quietly keeps the plumbing of the financial system working. If you've ever wondered why short-term rates don't fall below a certain floor, or what happens when banks have too much cash sitting idle—this is the mechanism you need to understand.

What Are Overnight Reverse Repurchase Agreements?

An overnight reverse repurchase agreement is a transaction where the Federal Reserve sells a security (usually a Treasury bond) to an eligible counterparty with an agreement to buy it back the next day at a slightly higher price. The difference between the sale and repurchase price is the implied interest rate—the ON RRP rate. In plain English: the Fed borrows cash overnight from participants like money market funds, using Treasuries as collateral.

This facility is a standing offer. Any of the roughly 160 eligible counterparties (mostly primary dealers, money market funds, and GSEs) can park cash at the Fed at a fixed rate. It's voluntary, but when market rates dip below the ON RRP rate, participants flock to it. That's the whole point: the ON RRP rate acts as a floor for short-term interest rates. I've seen it described as the Fed's "rate floor" tool, and that's exactly what it is.

How the ON RRP Facility Works

Let me walk you through a typical day. A money market fund (call it ABC Fund) has $10 billion in cash that it can't lend out at a decent rate. The repo market is offering 0.03%, but the Fed's ON RRP rate is 0.05%. ABC Fund sends $10 billion to the Fed's account, and the Fed delivers $10 billion in Treasuries (usually held at a custodian). The next morning, the Fed sends back the $10 billion plus interest, and the fund returns the securities. Simple, safe, and overnight.

The key operational detail: the Fed doesn't restrict usage. There's no cap on how much each counterparty can park. That's why during periods of excess reserves, usage can balloon to over $2 trillion. Critics call it "cash parking," but it's a deliberate design to keep rates from collapsing.

Why the Fed Uses ON RRP

The primary reason: to keep the fed funds rate within the target range. Without a floor, the effective fed funds rate could drift below the target, especially when bank reserves are abundant. Think of it as a safety net. The Fed sets the ON RRP rate 5-10 basis points below the interest on reserve balances (IORB) rate, creating a spread that guides short-term rates.

But there's a secondary, less obvious reason: absorbing excess liquidity. After the Fed's quantitative easing, banks and money funds were swimming in cash. That cash could have flooded the repo market and pushed rates to zero or negative. The ON RRP facility vacuums up that excess without the Fed having to sell assets aggressively. It's a passive tool, but incredibly powerful.

I remember in 2021, when the facility saw unprecedented usage—peak of $1.6 trillion—many thought it signaled stress. Actually, it was the opposite: it showed the system was functioning exactly as designed. The Fed was mopping up reserves smoothly.

Impact on Money Markets

Real-world effect: The ON RRP facility acts as a competitor to private repo and Treasury bill yields. When usage is high, money market funds are essentially choosing the Fed over lending to banks. This can squeeze bank balance sheets and reduce short-term lending.

Let's look at a specific chain reaction. Suppose a money fund moves $100 million from a bank repo to the ON RRP. The bank loses that cash, which it used to fund its own operations. The bank then may need to borrow elsewhere, pushing up rates in the federal funds market. That's how the ON RRP helps transmit the Fed's policy rate across the system.

During the 2023 banking turmoil, I noticed something interesting: ON RRP usage acted as a canary. When regional banks struggled, cash flooded into the facility, indicating a flight to safety. Conversely, when the Treasury issued lots of T-bills, money funds bought them instead, and ON RRP usage dropped. These dynamics matter for anyone trading short-term instruments.

Recent months saw a noticeable decline in ON RRP usage as the Fed's balance sheet runoff and Treasury bill supply increased. The table below shows the trajectory (all values in billions of dollars):

PeriodAverage Daily Usage
Early 20232,100
Mid 20231,650
Late 20231,100
Early 2024850
Recent600

That drop isn't a failure—it's a sign that money market conditions are normalizing. As reserves drain, banks need those deposits back. The ON RRP is acting as a release valve, gradually returning cash to the private market. I'd caution against reading too much into weekly fluctuations; the trend over months is what matters.

Common Misconceptions

I've seen even seasoned analysts confuse ON RRP with quantitative tightening (QT) or repo market crises. Let me clear up a few:

Misconception 1: "ON RRP usage means the financial system is broken." No. It means the Fed is providing a safe parking space. Usage can be high even in calm markets.

Misconception 2: "The ON RRP rate is the same as the FOMC's target." It's related but not identical. The Fed sets the ON RRP rate as a separate administered rate, usually 5 bps below IORB.

Misconception 3: "The Fed can control usage." It can't directly; usage depends on market participants' choices. The only lever is adjusting the rate relative to other benchmarks.

One trick I use: watch the spread between SOFR and the ON RRP rate. A narrowing spread suggests less demand for the facility, while a widening spread often precedes a surge in usage.

FAQ

When ON RRP usage drops sharply, what does it mean for the economy?
A rapid decline usually signals that private markets are absorbing cash faster—maybe because Treasury issuance is pulling money out, or banks are offering better rates. It's not automatically bullish or bearish. I'd check whether the drop coincides with a liquidity squeeze elsewhere, like in the repo market. If secured rates (like SOFR) spike at the same time, it suggests the banking system is strained. If they stay calm, it's just normal rebalancing.
How does the ON RRP facility affect retail investors who don't trade repos?
Indirectly, it influences the yields on money market funds and short-term bond ETFs. When ON RRP usage is high, money funds earn the Fed rate, which gets passed to investors. So a drop in usage might lead to slightly lower money market yields if funds have to chase private repo rates. It also affects the Treasury bill curve—when money funds buy bills instead of using RRP, bill yields tend to rise.
Can the Fed turn off the ON RRP facility if it wants?
Technically yes, the FOMC can suspend or adjust it. But doing so would remove the floor on short-term rates. In practice, they'd likely raise the ON RRP rate or narrow the spread before shutting it down. I've never seen serious discussion of eliminating it entirely; it's too useful a safety valve.
Why do money market funds use ON RRP instead of lending to banks?
Safety and simplicity. Lending to banks requires credit analysis, bilateral agreements, and some risk. The Fed's facility is risk-free (aside from the U.S. government credit) and fully automated. When bank rates are only marginally higher, funds often choose the path of least friction. I've found that the threshold for funds to switch is usually a spread of 5-10 bps over the ON RRP rate.

This article reflects my personal analysis of the ON RRP facility based on years of market observation. Data and trends are representative of typical conditions; actual figures may vary.