DRTSCILM: Net Percentage of US Banks Tightening C&I Loan Standards (SLOOS) Since 1990

DRTSCILM is the Federal Reserve’s quarterly measure of the net percentage of domestic banks tightening standards on commercial and industrial loans to large and middle-market firms. Reported through the Senior Loan Officer Opinion Survey (SLOOS) since 1990 Q2, DRTSCILM captures bank intentions before they show up in hard credit data — making it one of the most-watched survey-based leading indicators in US macro analysis. Positive values mean more banks are tightening than easing; negative values mean the reverse. Distributed via FRED.

Dataset: US Bank Lending Standards (1990–2026) · Updated —

Source: FRED series DRTSCILM · Federal Reserve — Senior Loan Officer Opinion Survey (SLOOS)

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 DRTSCILM dataset spans more than 35 years of quarterly SLOOS observations across every US credit cycle since 1990.

FRED Direct CSV Access

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

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

Direct CSV Access — Eco3min Structured Dataset

https://eco3min.fr/dataset/us-bank-lending-standards.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/us-bank-lending-standards.csv"
df = pd.read_csv(url, parse_dates=["date"])

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

Using the Dataset in R

library(readr)

url <- "https://eco3min.fr/dataset/us-bank-lending-standards.csv"
df <- read_csv(url)

head(df)
summary(df$net_pct_tightening)

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

Methodology

DRTSCILM is sourced from the Federal Reserve’s Senior Loan Officer Opinion Survey (SLOOS), a quarterly survey of approximately 80 large domestic banks and 24 US branches and agencies of foreign banks. Each quarter, the Fed asks senior loan officers whether their bank has tightened or eased standards on various loan categories. DRTSCILM specifically captures the response for C&I loans to large and middle-market firms (defined as firms with annual sales of $50 million or more).

The published value is a diffusion index: the percentage of banks reporting tightening minus the percentage reporting easing. A reading of +30% means 30 percentage points more banks tightened than eased — not that 30% of loans are subject to tightened standards.

The survey is conducted in the first or second month of each quarter and released approximately one month after the quarter-end (January, April, July, and October). Responses cover the prior three months and are weighted equally across respondents — no asset-weighting. This dataset is updated quarterly (Q+1 month) via automated pull from the FRED API.

Data Quality & Provider Notes

DRTSCILM is well-suited for cyclical analysis but has structural limitations that matter for time-series work. Eco3min refreshes the FRED mirror weekly; new observations appear approximately one month after each quarter-end.

• Release latency. The Federal Reserve releases SLOOS quarterly, typically in the first half of the second month following quarter-end (early February, May, August, November). Responses reflect bank policy over the prior three months, so the structural lag at first observation is roughly four months between the start of the surveyed period and publication.
• Revisions policy. DRTSCILM is generally not revised after publication. The respondent panel changes marginally over time as banks merge, exit, or are added — small enough not to require backward adjustment, but worth noting for long-horizon analysis. The Fed occasionally adds new survey questions; DRTSCILM itself has been continuous in its current form since 1990 Q2.
• Alternative sources. The Federal Reserve publishes the full SLOOS release directly at federalreserve.gov/data/sloos.htm, including small-firm C&I standards (DRTSCIS), commercial real estate, residential mortgage, and consumer loan standards. Haver Analytics and Bloomberg mirror the FRED series. For pre-1990 historical context, the Fed maintains an earlier discontinued series that is not directly comparable due to methodology changes.
• Known gaps. None within the 1990 Q2–present window. The survey has been conducted every quarter without interruption, including during the pandemic.

For analytical work, pair DRTSCILM with DRTSCIS (small-firm C&I standards) and with the corresponding loan-demand series (DRSDCILM) to separate supply-side tightening from demand-side credit weakness.

Common Pitfalls When Using DRTSCILM

DRTSCILM is widely cited as a recession indicator, but recurring interpretation errors distort the signal.

1. Confusing DRTSCILM with DRTSCIS. DRTSCILM covers C&I loans to large and middle-market firms; DRTSCIS covers small firms. The two series often diverge meaningfully — small-firm standards tend to tighten earlier and ease later than large-firm standards. Users tracking only one and attributing economy-wide significance to it miss the segmentation that has historically been informative around turning points.
2. Misreading the diffusion index as a percentage of loans. A +30% DRTSCILM reading means 30 percentage points more banks tightened than eased — not that 30% of loans face tighter standards. If 40% of banks tightened and 10% eased while 50% reported no change, DRTSCILM is +30%. The diffusion structure means the indicator is a directional signal about bank-policy momentum, not a quantitative measure of credit supply contraction.
3. Treating positive and negative readings symmetrically. The asymmetry of credit cycles makes negative DRTSCILM less informative than positive readings. Banks easing standards in late expansions often reflects competitive dynamics — including the search for yield, share-of-wallet pressure, and regulatory tolerance — rather than improving economic fundamentals. The deeply negative DRTSCILM of 2005–2006, alongside historically tight HY OAS spreads, illustrated this pattern.
4. Ignoring the publication lag. SLOOS releases roughly one month after quarter-end and reflects bank policy over the prior three months. By the time a tightening reading is published, the policy changes have already been in effect for two to five months. For real-time analysis of credit conditions during fast-moving stress episodes, the more contemporaneous market-priced measures — HY OAS, IG OAS, NFCI — provide signal that DRTSCILM mechanically cannot.

Historical Regimes

1990 Q2–1992 — Inaugural recession reading. The modern SLOOS launched in 1990 Q2, immediately into the 1990–1991 recession. DRTSCILM peaked at +59.7% in Q4 1990, signaling near-universal tightening across the surveyed panel. Although the recession officially ended in March 1991, lending standards remained tight through 1992 — an early demonstration of the asymmetric persistence of credit tightening relative to economic recovery.

1993–1999 — Late-cycle easing. DRTSCILM moved into and stayed in negative territory for most of the period, with banks easing standards as the expansion matured and competition for corporate lending intensified. The Asian crisis and LTCM episode of 1997–1998 produced a brief tightening response, but the rebound to easing was fast.

2000–2003 — Dot-com bust and aftermath. DRTSCILM spiked to +59.7% in Q4 2001, matching the 1990 peak. The tightening cycle lasted approximately eight quarters — longer than the recession itself — and was accompanied by the largest sustained widening of HY OAS spreads since the launch of the spread series in 1996.

2004–2007 — Pre-GFC easing. DRTSCILM reached -20% in 2005–2006, among the loosest readings in the series. Banks competed aggressively for corporate lending as covenant-lite structures spread from leveraged loans to broader C&I. Parallel deeply negative readings in the NFCI and historically tight HY spreads reinforced the picture of broad-based risk-premium compression.

2007–2009 — Global Financial Crisis. DRTSCILM reached its all-time high of +83.6% in Q4 2008 as banks froze new lending nearly universally. The reading exceeded +50% for five consecutive quarters — the longest sustained extreme-tightening regime in the dataset. The episode also produced the clearest historical example of DRTSCILM leading the hard data: the peak occurred two quarters before the trough of US GDP, and the easing back below zero in 2010 preceded the recovery of bank loan growth by roughly six months.

2010–2019 — Selective easing with episodic tightening. Banks gradually loosened standards (DRTSCILM averaged approximately -10% from 2011 to 2014) before mild tightening episodes in 2015–2016 (energy and EM stress spilling into C&I underwriting) and late 2019 (yield curve inversion). Neither episode reached the +30% historical recession threshold, and both were partially reversed before the pandemic shock.

2020–present — Pandemic, tightening, soft landing debate. DRTSCILM jumped to +71.2% in Q2 2020 — the second-highest reading on record — then eased through 2021 as fiscal support and Fed accommodation insulated corporate balance sheets. The 2022–2023 hiking cycle pushed DRTSCILM back above +50% by Q3 2023, a level historically followed by credit-contraction-driven recessions within four quarters. Readings eased somewhat through 2024–2025, but the divergence between elevated DRTSCILM tightening and contained HY OAS spreads has been one of the more distinctive features of the current cycle. Whether this reflects post-pandemic structural shifts in the credit channel or a delayed but still mechanically operative transmission remains debated.

Related Macroeconomic Datasets

DRTSCILM captures the survey-based supply side of credit. To form a complete picture, pair it with the market-priced credit risk measures (HY and IG spreads), the composite financial-conditions indices, and the stock-of-credit context that determines how tightening propagates to the real economy.

Macroeconomic Dataset Hub

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

Explore the Eco3min Dataset Hub

Sources

• Federal Reserve Board — Senior Loan Officer Opinion Survey (SLOOS)
• Federal Reserve Bank of St. Louis — FRED series DRTSCILM

Dataset Reference

Eco3min Research. (2026). US Bank Lending Standards — DRTSCILM (1990–2026).<br> https://eco3min.fr/en/us-bank-lending-standards-dataset/