# Credit conditions beyond Brazil's policy rate

Policy Brief · Brazil · Monetary Policy

Brazil's Selic rate is central to monetary conditions, but it is not the interest rate on every loan. Borrowing costs depend on funding, credit risk, operating expenses, taxes, and the product's contractual and regulatory structure. Earmarked and non-earmarked credit can consequently respond differently to the same policy environment.

The Banco Central do Brasil's [May 2026 Financial Stability Report, section 2.1](https://www.bcb.gov.br/content/publications/financialstabilityreport/202605/fsrFullRep.pdf) provides a concrete decomposition. For 2025 it reports an adjusted average interest cost of credit of 29.30 percent for non-earmarked lending and 9.59 percent for earmarked lending. The difference is 19.71 percentage points. These are annual portfolio measures under the report's methodology, not the latest offered rate for an individual borrower.

## A portfolio cost differs from the price of a new loan

The interest cost of credit reflects the relevant outstanding credit portfolio. New-loan rates describe contracts entering the flow during a period. Their movements can differ because existing loans reprice gradually and because the mix of new borrowing changes. A company refinancing a particular obligation needs the actual offered terms rather than the portfolio average.

Composition can change an aggregate without every contract changing. More borrowing through an expensive product can raise the average even if product-specific rates are stable. A shift toward secured or regulated lending can lower it. Comparisons should separate within-product repricing from changes in the weights.

Maturity and borrower type matter as well. A household credit-card balance and a corporate investment loan carry different risks and cost structures. Combining them can describe total intermediation but cannot establish conditions facing a particular business segment. The relevant series should match the borrowing purpose and population.

## The spread is not identical to bank profit

The BCB decomposition separates funding cost from delinquency, administrative expense, taxes and the credit-guarantee fund, and a financial margin. The difference between a lending rate and funding cost therefore contains several components. Describing that entire gap as profit would omit expected credit losses and other costs.

The components are estimated under a stated methodology. They should not be treated as an itemized invoice on each borrower or a causal decomposition of every policy change. An increase in the delinquency component can reflect worsening repayment conditions, while a funding-cost increase can reflect repricing. These mechanisms have different implications for what a rate cut can accomplish.

The financial margin is also not automatically the return on equity. Capital, risk, provisions, and the accounting perimeter matter for that comparison. A portfolio-level spread should not be transferred directly into a profitability claim without reconciling those concepts.

## Directed credit changes the route of transmission

Earmarked credit operates under specific financing and regulatory arrangements. The report notes its greater reliance on regulated or parafiscal sources and lower sensitivity to market conditions. That can buffer some borrowers from immediate policy-rate changes, while non-earmarked lending responds through a different funding structure.

Buffering does not make the financing cost disappear. Its allocation depends on the source of funds, any regulation, eligibility, and the institution bearing risk. The economic appraisal should identify those arrangements rather than assume that a lower borrower rate establishes a lower social cost.

Eligibility also matters. A firm unable to access a directed program may face market-based credit conditions even when a supported category expands. Aggregate lending growth can therefore coexist with tight conditions for a particular borrower group. The distribution of access is as relevant as the average price.

A comparison between the two segments also needs borrower and product controls. Their rate difference cannot all be assigned to the policy framework if collateral, maturity, purpose, and expected losses differ. A causal evaluation would seek comparable borrowers or another defensible design.

## Household and corporate balance sheets mediate the response

A policy-rate change can alter debt service, deposit income, working-capital cost, and the discount rate for investment. The net effect depends on whether a borrower has fixed or variable rates, when debt matures, and how much liquidity it holds. The same borrower can benefit on one side of the balance sheet while facing higher cost on another.

Suppose a hypothetical firm has 20 million reais of floating-rate debt and 5 million of deposits. If its debt rate rises one percentage point and deposit earnings rise half a point for a full year, additional interest expense is 200,000 reais and additional income 25,000. The direct net cost is 175,000 reais before taxes and changes in principal. Applying the policy-rate change only to net debt would give a different, unjustified result under these assumptions.

The commercial effect also depends on sales, margins, and cash collection. A firm with growing earnings may absorb the increase, while one with falling revenue may reduce investment. A credit assessment should test both operating and financing scenarios rather than label all deterioration a direct consequence of monetary policy.

Collateral requirements provide an additional adjustment margin. A Brazilian lender can leave the advertised rate unchanged while requiring more security, a shorter maturity, or a larger borrower contribution. Credit conditions can therefore tighten without an obvious movement in the quoted price. Borrower surveys and contract terms are useful complements to rate data when determining whether access has changed for otherwise comparable firms. The financing assessment should retain these nonprice terms rather than treating every unchanged rate as unchanged credit availability.

## A diagnostic approach to lending conditions

| Evidence | Economic role |
|---|---|
| Selic and market funding | Marginal monetary and liquidity conditions |
| Product-specific new-loan rates | Current pricing for a defined borrowing flow |
| Outstanding portfolio cost | Cost embedded in the existing loan stock |
| Spread decomposition | Funding, losses, operating and fiscal components |
| Applications, approvals, and drawdowns | Demand and access beyond the quoted rate |

Brazil's credit system requires these layers to be read together. The BCB's segmented evidence demonstrates that one policy rate cannot reproduce all financing conditions. A useful assessment identifies the product, source of funding, borrower risk, and repricing schedule before judging how a monetary decision reaches household spending or business investment.


---

American Economic Senate