RRPONTSYD: Overnight Reverse Repo Facility Daily Take-Up Data Since 2003 (FRED)

RRPONTSYD is the FRED daily series for the Federal Reserve’s Overnight Reverse Repo Facility take-up — the dollar volume of cash parked overnight at the Fed by money market funds, government-sponsored enterprises, and other eligible counterparties. The series, published by the New York Fed Open Market Desk, peaked above $2.4 trillion in late 2022 and drained to near zero by early 2026. RRPONTSYD is the standard operational input for tracking excess system liquidity at the front end of the money market curve.

Dataset: Overnight Reverse Repo Facility (2003–2026) · Updated 2026-08-03

Source: FRED series RRPONTSYD · Federal Reserve Bank of St. Louis

Macro Takeaway

Dataset Overview

Dataset Variables

The CSV and Excel files contain the following columns.

Column names match the CSV headers exactly.

Download the Complete Dataset

The full dataset is available in CSV and Excel formats.

FRED Direct CSV Access

The underlying data is available from FRED under series code RRPONTSYD:

https://fred.stlouisfed.org/graph/fredgraph.csv?id=RRPONTSYD

Direct CSV Access — Eco3min Structured Dataset

https://eco3min.fr/dataset/reverse-repo-facility.csv

This URL returns the complete dataset in CSV format. It can be used directly in pandas, R, curl, or any data tool.

Using the Dataset in Python

import pandas as pd

url = "https://eco3min.fr/dataset/reverse-repo-facility.csv"
df = pd.read_csv(url, parse_dates=["date"])

print(df.head())
print(df["rrpontsyd"].describe())

Using the Dataset in R

library(readr)

url <- "https://eco3min.fr/dataset/reverse-repo-facility.csv"
df <- read_csv(url)

head(df)
summary(df$rrpontsyd)

Both examples load the dataset directly from the URL — no download or API key required.

Methodology

RRPONTSYD is reported by the Federal Reserve Bank of New York’s Open Market Desk and republished through FRED. Each observation is the total dollar amount accepted by the Fed in the overnight reverse repo operation conducted that afternoon, in which eligible counterparties (primary dealers, money market funds, government-sponsored enterprises, and select banks) deposit cash with the Fed against Treasury collateral and earn the offered rate, currently set at the lower bound of the federal funds target range.

The series is dated to the business day of the operation. There are no observations on weekends or US holidays. The ON RRP rate itself is published as a separate FRED series (RRPONTSYAWARD) and is not contained in this dataset, which focuses on take-up volume. Eco3min pulls the daily value from the FRED API on an automated schedule and republishes within hours.

Data Quality & Provider Notes

RRPONTSYD is a daily operational series published by the New York Fed and considered authoritative. Eco3min mirrors the FRED feed with an automated daily pull, so each business day’s take-up is available within hours of the operation. The notes below cover release cadence, revisions, and structural breakpoints relevant to data engineering.

• Release latency. The ON RRP operation closes at 1:15 PM ET. The take-up is reported by the New York Fed in the daily Statement of Money Market Operations the same afternoon. FRED publishes within hours, typically by 4:00 PM ET. Eco3min refreshes daily.
• Revisions policy. RRPONTSYD is rarely revised. Operational data of this type is reported once and not subject to later restatement. Format-level changes to the facility (offered rate, counterparty cap, collateral eligibility) do not retroactively revise the series.
• Alternative sources. The same data is available directly from the New York Fed (newyorkfed.org/markets/desk-operations/reverse-repo), and through Bloomberg, Refinitiv/LSEG, and Haver Analytics. FRED is the standard free source.
• Known gaps. No observations on weekends and US federal holidays (per the business-day calendar). The series exists from 2003 onward, but the facility’s permanent standing format dates from September 2013; pre-2013 values reflect ad-hoc reverse repo testing operations and are not directly comparable to the post-2013 standing facility on a behavioral basis. The April 2021 counterparty cap increase (from $30B to $80B per counterparty) is a key structural break that enabled the 2021–2022 explosion in take-up.

Practical recommendation: when analyzing pre- and post-2021 RRPONTSYD dynamics, treat the counterparty cap regime as a structural break and avoid comparing distribution statistics directly across the two periods.

Common Pitfalls When Using RRPONTSYD

RRPONTSYD looks straightforward but several interpretation errors recur in macro and money market analysis.

1. Confusing take-up volume with the offered rate. RRPONTSYD measures the dollar amount placed at the facility. The rate the Fed offers — a separate parameter, typically aligned with the lower bound of the federal funds target — is published as RRPONTSYAWARD. The two move on different drivers: take-up reacts to the spread between the offered rate and alternatives (T-bills, private repo), while the offered rate is set administratively by the FOMC.
2. Reading the ON RRP drain as a tightening signal. A counterintuitive feature of the facility: when RRPONTSYD declines, the cash typically returns to T-bill purchases, private repo lending, or bank deposits. The dollars do not exit the financial system — they reshuffle. Treating the drain as analogous to QT or rate hikes mistakes a portfolio rotation for an aggregate liquidity contraction.
3. Comparing pre-2021 and post-2021 levels directly. The April 2021 counterparty cap expansion changed the facility’s operational ceiling by an order of magnitude. Pre-2021 RRPONTSYD typically traded between $0 and $200 billion. Post-2021 saw routine readings above $1 trillion. Comparing distributional statistics across the structural break — without segmenting the periods — produces misleading conclusions about facility utilization.
4. Conflating ON RRP with the Standing Repo Facility. The Standing Repo Facility (SRF, FRED series RPONTSYD) is the symmetric tool on the opposite side: counterparties pledge collateral and borrow from the Fed when private repo rates spike above the offered rate. ON RRP absorbs excess cash; SRF provides emergency cash. They activate under opposite conditions and should not be aggregated or compared interchangeably.

Historical Regimes

2003–2013 — Operational testing. RRPONTSYD existed as an ad-hoc Open Market Desk operation. Take-up was sporadic and typically below $50 billion. The facility was not yet a standing tool, and the data reflects experimental operations rather than systematic monetary policy.

September 2013 – 2020 — Standing facility, low usage. The Fed introduced the facility in its permanent standing format in September 2013, capped initially at $30 billion per counterparty. Take-up oscillated between near zero and several hundred billion at quarter-ends, reflecting normal money market funds’ management of excess cash. The facility played a modest role in the floor system.

2020–April 2021 — COVID expansion baseline. Massive QE during COVID flooded the system with reserves, but RRPONTSYD remained quiet through 2020 because the offered rate (5 bps) sat below alternative private repo rates. T-bill scarcity and rate compression at the front end of the curve set the conditions for the subsequent explosion.

April 2021 – September 2022 — Cap expansion and explosion. The April 2021 counterparty cap increase to $80 billion per counterparty removed the binding constraint. As money market funds searched for yield above near-zero T-bill rates, RRPONTSYD climbed from under $100 billion to a peak of approximately $2.37 trillion in September 2022. The expansion absorbed the excess liquidity created by 2020–2022 QE and the deposit growth at money market funds.

September 2022 – Early 2024 — Drain phase one. The 2023 debt-ceiling resolution unleashed a flood of T-bill issuance to refill the Treasury General Account. Money market funds rotated from RRPONTSYD into T-bills, draining the facility from $2.37T to roughly $400B by mid-2024. This drain mechanically offset most of the Fed’s quantitative tightening, keeping bank reserves stable.

Mid-2024 – 2026 — Approach to zero. Take-up declined to near zero by early 2026. With the buffer effectively exhausted, further QT — or any liquidity shock — would land directly on bank reserves rather than being absorbed by the facility. The dynamic is detailed in the net liquidity index dataset.

Related Macroeconomic Datasets

RRPONTSYD is one of the three core liquidity components alongside the Fed balance sheet and the Treasury General Account. The series below cover the broader liquidity framework, the money aggregates, and the policy rate that calibrates the facility’s offered rate. How the ON RRP, alongside the TGA, separates reported balance sheet from real system liquidity is set out in Eco3min’s reading of real versus reported Fed liquidity.

Macroeconomic Dataset Hub

This dataset is part of the Eco3min macro-financial data repository. Explore all available datasets including inflation, interest rates, equity returns, credit spreads, and currency indicators.

Explore the Eco3min Dataset Hub

Sources

• Federal Reserve Bank of St. Louis — FRED series RRPONTSYD
• Federal Reserve Bank of New York — Open Market Desk, Statement of Money Market Operations

Dataset Reference

Eco3min Research. (2026). Overnight Reverse Repo Facility — RRPONTSYD (2003–2026).<br> https://eco3min.fr/en/overnight-reverse-repo-facility-dataset/