Ask the loan officers: why bank lending surveys track credit risk better than CDS

When I looked for the market signal that best explained movements in credit pricing, the answer was not the one most people reach for. It was a survey of bank loan officers.

The obvious candidate is CDS. Spreads are daily, liquid and quoted on thousands of names. If you want a market view of default risk, that seems the natural place to look. But when I lined up credit pricing against a range of macro and market indicators, the Federal Reserve’s Senior Loan Officer Opinion Survey, the SLOOS, tracked it more closely than CDS did.

Why the survey wins

Coverage. A trade credit or surety book is dominated by private, mid-market companies: buyers and contractors that have never issued a bond and will never have a CDS. CDS describes large, listed issuers. The SLOOS asks banks about lending standards for exactly the kind of firms that sit in an insurer’s book, including small businesses.

Mechanism. Mid-market defaults rarely surprise the lender. They follow a tightening of credit supply: banks raise standards, shorten terms and cut limits, and weaker companies lose the refinancing they need. The survey measures that tightening directly, often before it appears in insolvency statistics. It captures the cause rather than the symptom.

Noise. CDS spreads carry far more than default expectations: liquidity premia, hedging flows, risk-on and risk-off sentiment, and the technicals of a dealer-driven market. A spread can widen because a fund needs protection, not because a company got riskier. The survey is slower, but it is cleaner.

How I use it

I treat lending standards as a systematic factor: a forward-looking read on the credit cycle that feeds the economy a pricing model simulates. When net tightening rises, so does the probability of a stressed year, and price adequacy should be tested against that, not against last year’s benign loss ratio.

The caveats

The SLOOS is quarterly and US-focused, so for European books I pair it with the ECB Bank Lending Survey and the Bank of England’s Credit Conditions Survey. Surveys are opinions, not transactions, and they can break around structural shocks, as 2020 showed. And correlation is not a pricing model. It is an input that tells you which way the wind is blowing.

The broader lesson: the best signal for a risk is the one that measures its mechanism, not the one with the most decimal places.

Banu Munnangi is SVP, Credit Reinsurance AI at Marsh Re, where he leads data and AI for the Credit, Bond and Political Risk specialty. More writing · Expertise

Views are my own and do not represent my employer. Examples use illustrative figures, not client or employer data.