DRSFRMACBS: US Single-Family Mortgage Delinquency Rate Quarterly (1991–2026)

The US Mortgage Delinquency Rate tracks single-family residential mortgages at commercial banks at least 30 days past due, quarterly since 1991, from the Federal Reserve Board. One episode dominates the chart: the climb from ~2% to 11.48% (Q1 2010) through the housing collapse — a magnitude no other US consumer credit series has approached. Equally telling is the other extreme: the series low of 1.41% was set in Q4 2004, at the height of the bubble, and the 2026 reading of 1.89% sits near that historic floor.

Dataset: US Mortgage Delinquency Rate (1991–2026) · Updated 2026-01-01

Source: Board of Governors of the Federal Reserve System · FRED series DRSFRMACBS

Macro Takeaway

Dataset Overview

Dataset Variables

The CSV and Excel files contain the following columns.

A value of 1.89 means 1.89% of bank-held single-family mortgage balances were at least 30 days past due.

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 DRSFRMACBS:

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

Direct CSV Access — Eco3min Structured Dataset

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

print(f"Latest: {df['mortgage_delinquency_rate'].iloc[-1]:.2f}% of balances 30+ days past due")
print(df.tail())

Using the Dataset in R

library(readr)

url <- "https://eco3min.fr/dataset/mortgage-delinquency.csv"
df <- read_csv(url)

tail(df)
summary(df$mortgage_delinquency_rate)

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

Methodology

The Federal Reserve Board computes the rate from quarterly bank call reports: single-family (1–4 unit) residential mortgage balances 30+ days past due or in nonaccrual, over total such balances, seasonally adjusted.

The universe is mortgages held on bank balance sheets — predominantly portfolio loans (jumbos, ARMs banks retain, community-bank originations). Loans sold into agency securitizations, the majority of the market, are excluded.

This dataset is updated daily (Mon–Sat, 08:00 UTC) via automated pull from the FRED API; new observations appear with each source release.

Data Quality & Provider Notes

• Release latency. ~2 months after quarter end with the Board’s delinquency release.
• Revisions policy. Minor call-report corrections only.
• Universe. Bank-held single-family mortgages — a minority, higher-quality slice of the total market. MBA and CoreLogic cover the full market on different definitions.
• Forbearance treatment. During 2020, CARES Act forbearance kept many non-paying loans out of delinquency counts; the 2.83% blip of Q3 2020 understates pandemic payment disruption.
• Known gaps. None; continuous since Q1 1991.

Common Pitfalls When Using Mortgage Delinquency Data

1. Reading low delinquency as a housing all-clear. The series bottomed in Q4 2004. Delinquency reflects the price and refinancing regime as much as borrower quality — it is a lagging, regime-dependent indicator.
2. Quoting it as the national mortgage delinquency rate. Bank-held loans only; the MBA survey (full market, different past-due definitions) is the standard for market-wide statements, and its levels differ.
3. Ignoring the 2020 forbearance distortion. Millions of paused loans were not counted delinquent; comparing 2020 to 2009 on this series alone understates the pandemic shock by design.
4. Expecting card-like reactivity. Secured, payment-prioritized debt moves slowly; mortgage delinquency turns quarters after unsecured credit and the labor market.

Historical Regimes

1991–2003 — The quiet band. Delinquency held between 1.9% and 3% through a recession and recovery — the pre-bubble norm for bank-held mortgage books.

2004–2006 — The deceptive floor. The series low of 1.41% (Q4 2004) coincided with peak subprime origination: rising prices masked underwriting decay by letting distressed borrowers exit through sales and refis.

2007–2010 — The collapse. Delinquency rose eightfold to 11.48% (Q1 2010) — the credit core of the global financial crisis, unmatched in scale by any other consumer loan category.

2011–2019 — The decade-long workout. A slow grind down to ~2.5% by 2019 as foreclosure pipelines cleared and post-crisis vintages, underwritten to QM standards, replaced legacy loans.

2020–2021 — Forbearance, not delinquency. A modest blip to 2.83% (Q3 2020) despite record job losses — CARES Act forbearance absorbed the shock outside the delinquency definition.

2022–2026 — The lock-in floor. The rate fell to the 1.7–1.9% range and stayed there through the rate shock: fixed sub-4% payments, tight vintages, and price appreciation produced the strongest sustained mortgage performance in the series — by a mechanism opposite to 2004’s false floor.

Related Macroeconomic Datasets

Mortgage performance sits where household credit meets the housing market; both sides are below.

Macroeconomic Dataset Hub

This dataset is part of the Eco3min macro-financial data repository.

Sources

• Board of Governors of the Federal Reserve System — Charge-Off and Delinquency Rates on Loans at Commercial Banks
• Federal Reserve Bank of St. Louis — FRED series DRSFRMACBS

Dataset Reference