What is the allowance for credit losses on a bank's balance sheet?

The allowance for credit losses is a reserve on a bank's balance sheet that reduces the total loan balance to account for expected losses. Banks set this money aside because some borrowers will inevitably fail to repay, and the allowance reflects the bank's best estimate of those future losses across its entire loan portfolio.

The allowance for credit losses (ACL) shows up on the balance sheet as a negative number subtracted from gross loans. If a bank holds $10 billion in total loans and carries a $150 million allowance, its balance sheet reports net loans of $9.85 billion. That net figure is what flows into total assets.

Older filings and industry discussions sometimes use the term allowance for loan and lease losses (ALLL). The terminology shifted when banks adopted new accounting rules under CECL (Current Expected Credit Losses), but the underlying concept is the same: money reserved against loans that may not be fully repaid.

How the Allowance Grows and Shrinks

Three forces move the allowance balance:

Provisions add to it. Each quarter, the bank records a provision for credit losses on its income statement, and the corresponding amount flows onto the balance sheet as an increase to the allowance.

Charge-offs reduce it. When a bank determines that a specific loan is uncollectible, it writes that loan off against the allowance. The gross loan balance drops and the allowance drops by the same amount, leaving net loans unchanged.

Recoveries add back to it. If the bank later collects on a loan it previously charged off, that recovery increases the allowance.

A quick example: a bank starts a quarter with a $200 million allowance. It records a $15 million provision, charges off $20 million in loans, and recovers $3 million on previously written-off loans. The ending allowance is $198 million ($200 + $15 - $20 + $3). Over time, the allowance balance reflects the cumulative net result of these three flows stretching all the way back to the bank's formation.

What CECL Changed

Before CECL took effect, banks reserved for credit losses under an incurred loss model. They only set aside reserves when losses were probable and reasonably estimable based on current conditions. In practice, allowances often stayed relatively low during good economic periods and then spiked during downturns, lagging the actual deterioration in credit quality.

CECL requires banks to estimate lifetime expected losses on every loan from the day it originates. The allowance must cover losses the bank expects to occur over the remaining life of each loan, not just losses already signaled by delinquency or default. Because this approach is forward-looking, the allowance now incorporates management's economic forecasts. When the outlook darkens, the allowance can increase even if borrowers are still making payments on time.

The practical effect is more volatility in the provision line on the income statement. A recession forecast can trigger a large provision expense well before actual losses show up, and an improving forecast can lead to provision releases that boost reported earnings.

Evaluating the Allowance

Analysts use two primary ratios to assess whether a bank's allowance is adequate:

The loan loss reserve ratio (allowance divided by total loans) measures the overall reserve level as a percentage of the portfolio. For most U.S. commercial banks, this ratio has historically ranged from about 1.0% to 1.5% during normal economic periods. A bank with a heavy concentration in commercial real estate or consumer lending will typically carry a higher reserve ratio than one focused on lower-risk residential mortgages.

The reserve coverage ratio (allowance divided by nonperforming loans) shows how many times the allowance covers identified problem loans. A coverage ratio above 100% means the bank has reserved more than the total of its current nonperforming loans, providing a cushion against additional deterioration. Coverage ratios between 150% and 200% are common among well-reserved banks in stable credit environments.

Neither ratio tells the full story on its own. A low reserve ratio might look concerning, but if the bank's loan mix is heavily weighted toward low-loss categories like single-family mortgages, a lower ratio could be perfectly appropriate. Context always matters.

Why Two Banks Can Report Different Allowances

The allowance is a management estimate, and judgment runs through every part of it. Two banks with nearly identical loan portfolios can report materially different allowance levels because they use different economic forecast scenarios, different loss rate models, and different qualitative adjustment factors.

CECL gives banks latitude in choosing their modeling approach. Some use discounted cash flow models, others rely on loss-rate methods, and larger banks often combine multiple models for different loan segments. The choice of forecast period, scenario weighting, and reversion assumptions can all shift the calculated allowance by meaningful amounts. Regulators and auditors review these assumptions, but there is no single correct answer for what the allowance should be.

This subjectivity is exactly why analysts don't take any one bank's reported allowance at face value. Peer comparisons, trend analysis, and examining how the allowance relates to actual charge-off experience all provide additional perspective.

What a Changing Allowance Tells You

A rising allowance relative to loans generally signals that the bank expects credit conditions to worsen. This can reflect genuine deterioration in the portfolio, a shift in the economic forecast, or increased uncertainty about specific loan segments. Pay attention to whether the increase is driven by higher provisions (the bank is actively building reserves) or by lower charge-offs (losses are coming in below expectations, so the allowance simply accumulates).

A declining allowance can mean credit quality is genuinely improving, with charge-offs and nonperforming loans both trending down. But it can also mean the bank is releasing reserves to pad earnings during a soft quarter. The way to distinguish between the two: look at charge-off trends and nonperforming loan levels alongside the allowance. If all three are improving together, the reserve release is likely legitimate. If the allowance is declining while nonperforming loans are flat or rising, that gap deserves closer scrutiny.

Differences Across Bank Types

Large banks with sophisticated modeling teams and diverse loan portfolios tend to maintain more granular, model-driven allowances. Their ACL models often incorporate multiple economic scenarios with probability weightings, and they update forecasts frequently.

Community banks and smaller regionals typically use simpler approaches, often based on historical loss rates for each loan category with qualitative overlays. Their allowances tend to be more stable quarter to quarter because the methodology is less sensitive to forecast changes. Regulators hold all banks to the same standard of maintaining an adequate allowance, regardless of modeling sophistication.

Loan mix is the biggest driver of differences in the reserve ratio across bank types. A community bank with 70% of its portfolio in commercial real estate will carry a different reserve level than a large bank with a mix of credit cards, auto loans, and commercial lending. Credit cards and other unsecured consumer loans carry much higher expected loss rates than secured commercial loans, so banks with large consumer portfolios will naturally report higher reserve ratios.