← Back to home

Bank Lending Standards: The Recession Indicator Nobody Watches

Every quarter since 1990, the Federal Reserve has surveyed senior loan officers at dozens of banks with one question: "Are you tightening or easing lending standards?" The answers predict recessions with remarkable accuracy—an average lead time of 4 months. Yet almost nobody watches this data. We test 35 years of the Senior Loan Officer Opinion Survey (SLOOS) and find it outperforms the yield curve on timing. But there's a catch: 2 major false signal episodes since 1990, including a massive tightening in 2022-2023 that produced no recession.

What is SLOOS?

The Senior Loan Officer Opinion Survey on Bank Lending Practices (SLOOS) is a quarterly survey conducted by the Federal Reserve. They ask roughly 80 large domestic banks and 24 branches of foreign banks about:

The key questions:

  1. Are you tightening or easing standards for approving loans?
  2. Is demand for loans increasing or decreasing?
  3. Why are you making these changes? (economy weakening, capital requirements, competition, etc.)

The Fed reports results as net percentages: (% tightening) - (% easing). A positive number means more banks are tightening than easing. A number above 20% typically signals significant credit contraction.

Why This Matters

Banks are forward-looking. When they tighten standards, they're saying: "We think economic conditions are deteriorating, so we're getting more selective about who we lend to." This happens before the economy actually weakens, because banks have access to real-time data on their borrowers' finances and payment histories.

Once standards tighten, credit growth slows. Businesses can't invest. Consumers can't spend. The economy contracts. It's a self-fulfilling prophecy, but one that's highly predictable.

Test 1: Does SLOOS Predict Recessions?

Let's look at the full history: net percentage of banks tightening C&I lending standards from 1990 to 2025, with NBER recession periods shaded.

Bank Lending Standards vs. Recessions (1990-2025)

Net percentage of banks tightening C&I lending standards (blue line). Gray shaded areas are NBER recessions. Notice how the blue line spikes before each gray bar. Average lead time: 4 months.

Recession Lead Times

Recession Peak Tightening Tightening % Lead Time

Result: SLOOS has predicted all 4 recessions since 1990 with an average lead time of 4 months. Much shorter than the yield curve (12-18 months), but shorter leads are actually better for actionable signals.

Test 2: SLOOS vs. Yield Curve

We've previously tested the yield curve as a recession predictor. It works, but with long and variable lags (6-24 months). How does SLOOS compare?

Comparing Two Leading Indicators

Red: SLOOS tightening %. Blue: 10Y-2Y yield curve spread (negative = inverted = recession signal). Yield curve inversions typically lead recessions by 12-18 months. SLOOS spikes appear 4 months before recessions.

Indicator Comparison

Indicator Avg Lead Time False Signals (1990-2025) True Signals Signal Accuracy
SLOOS Tightening >20% 4 months 2 episodes (1998, 2022-23) 4 recessions predicted 67% (4 of 6 signals)
Yield Curve Inversion 12-18 months 2 episodes (1998, 2022-25) 2 recessions predicted (2001, 2008) 50% (2 of 4 signals)

The Key Difference: SLOOS leads by only 4 months (more actionable timing) and predicted all 4 recessions since 1990, giving 67% signal accuracy (4 true, 2 false). The yield curve leads by 12-18 months (harder to time) and only predicted 2 of 4 recessions, with 50% accuracy (2 true, 2 false). SLOOS missed COVID-2020 briefly but caught it during, while yield curve failed completely on the short COVID recession and hasn't been followed by recession in 2022-2025 despite inversion.

Test 3: The False Signal Problem

Here's the uncomfortable reality: since 1990, there have been multiple distinct tightening episodes where SLOOS crossed above 20%. Of these episodes, only the ones that actually led recessions were useful signals. The rest were false alarms where you'd have gone defensive for nothing.

False Signals: When Tightening Didn't Lead to Recession

Red shaded areas show tightening episodes above 20% that did NOT predict a recession (no recession within 6 months before or 12 months after the peak). Two major false signals visible: late 1990s and 2022-2023.

Notable False Signal Episodes

Why These False Signals?

  1. 1998: Fed acted quickly with rate cuts to stabilize markets before credit crunch spread to real economy
  2. 2022-2023: Strong labor market, excess consumer savings from COVID stimulus, and low corporate refinancing needs created resilience despite tight credit

Test 4: Does Loan Growth Actually Decline?

When banks say they're tightening standards, does credit actually contract? Let's check by comparing SLOOS to year-over-year growth in commercial & industrial (C&I) loans outstanding.

"Loans outstanding" is the total dollar amount of C&I loans on bank balance sheets. When this number grows year-over-year, banks are extending more credit to businesses. When it shrinks, credit is contracting. We expect a strong correlation: tighter standards → fewer loan approvals → slower loan growth → possible credit crunch.

Lending Standards vs. Actual Loan Growth

Blue: SLOOS tightening % (left axis). Orange: C&I loan growth year-over-year (right axis). When standards tighten, loan growth slows with a lag of 1-2 quarters. But not all tightening produces negative loan growth.

The Lag Structure

Quarter 0 (Survey): Banks report tightening standards
Quarter 1-2: Loan growth begins to slow as new applications get rejected
Quarter 2-3: If severe, loan growth turns negative
Quarter 3-4: Recession begins (if credit contraction is deep enough)

Key Insight: The severity matters. Tightening from 10% to 30% might slow loan growth from 8% to 4% (no recession). Tightening from 10% to 70% might turn loan growth negative (recession likely).

Test 5: What About Consumer Credit?

We've focused on C&I (business) lending, but what about consumer lending standards? Do they predict recessions independently?

Consumer vs. Business Lending Standards

Blue: C&I lending standards (1990-present). Orange: Consumer lending standards (starts mid-1990s - FRED data not available earlier). Both tend to move together, but consumer standards lag slightly and are less volatile.

Consumer vs. Business Standards

C&I standards lead: Banks tighten business lending first because business loans are riskier and more sensitive to economic conditions.

Consumer standards follow: 1-2 quarters later, banks tighten consumer credit as unemployment rises and household finances weaken.

Exception - 2022: Consumer standards tightened MORE than C&I standards, possibly due to rising credit card delinquencies and auto loan defaults. This unusual pattern may explain why recession didn't materialize—businesses stayed healthy even as consumers struggled.

Test 6: The 2022-2023 Mystery

The most important test: what happened in 2022-2023? SLOOS showed the most extreme tightening since the 2008 financial crisis—over 60% net tightening in Q2 2023. Yet no recession came. What went wrong?

2022-2023: Record Tightening, No Recession

SLOOS tightening exceeded 60% in Q2 2023 following Silicon Valley Bank collapse and Fed rate hikes to 5.25%. Historical rule: anything above 40% predicts recession. This time: no recession through end of 2025.

Why 2022-2023 Was Different

  1. Fiscal support still flowing: Unlike past cycles, government transfer payments to households remained elevated through 2023. Consumers had cushion.
  2. Strong labor market: Unemployment stayed below 4%. Hard to have recession when everyone's employed.
  3. Sector-specific crisis: Banking stress was concentrated in regional banks exposed to commercial real estate and tech deposits. Money center banks stayed healthy.
  4. Pandemic excess savings: Households entered 2022 with record savings. Even as credit tightened, they could draw down cash.
  5. Corporate borrowing front-loaded: Many companies refinanced at low rates in 2020-2021. They didn't need new loans in 2022-2023 even as standards tightened.

The Lesson: SLOOS measures supply of credit (banks willing to lend). But recessions require both supply contraction AND demand for credit. If businesses/consumers don't need to borrow (because they're flush with cash or already locked in low rates), tight standards don't bite.

Key Takeaways

For Your Toolkit

SLOOS is worth monitoring as a confirmation signal:

  1. If yield curve inverts AND SLOOS tightens >30% → high recession probability
  2. If yield curve inverts but SLOOS stays loose → wait (like 2022-2025)
  3. If SLOOS tightens but yield curve is positive → probably false signal (like 2016)

Where to find it: Federal Reserve Board publishes SLOOS quarterly. FRED database has series DRTSCILM (C&I standards) and DRTSCLCC (consumer standards). Reports come out ~30 days after quarter end.

Data & Methodology

Data Sources:

Calculations: