Loan Loss Forecasting
Executive Summary
Key Takeaways
- ✓ Loan loss forecasting extends segment-level credit loss provisioning into a forward-looking exercise, projecting how expected loss rates evolve as macroeconomic conditions and portfolio composition change across the forecast period.
- ✓ Loss-rate assumptions should be held current, not static — a loss rate calibrated to historical experience at the start of a forecast can become materially unrepresentative several periods forward as conditions genuinely change.
- ✓ The forecast should link loss-rate assumptions to a defined macroeconomic scenario (base, stressed, or otherwise), so that a scenario change flows through to the loan loss forecast consistently rather than requiring a separate, manual adjustment.
- ✓ Portfolio composition changes — growth concentrated in a higher-risk segment, for example — should themselves shift the forecast's blended loss rate, even before any change in the underlying segment-level loss-rate assumptions.
- ✓ The forward-looking loss forecast should reconcile to the allowance for credit losses roll-forward, so the balance sheet reserve projected across the forecast period is internally consistent with the income statement provision charges driving it.
Objective¶
This guide covers how to extend Credit Loss Provisions into a forward-looking, scenario-based forecast, within the Banking Financial Modelling pillar.
From a Single Period to a Forecast¶
Credit loss provisioning, as described in Credit Loss Provisions, applies segment-level loss-rate assumptions to segmented loan balances for a single period. Loan loss forecasting extends this across the full forecast horizon, projecting how those loss rates themselves should evolve as conditions change, rather than holding the initial period's calibration static throughout.
Segment Loss Rate (Period t) = Segment Loss Rate (Period t-1) × Scenario-Linked Adjustment Factor
Segment Provision Charge (Period t) = Segment Loan Balance (Period t) × Segment Loss Rate (Period t)
Linking Loss Rates to Economic Scenarios¶
The forecast's segment-level loss-rate assumptions should be explicitly linked to the macroeconomic scenario in use, so that switching from a base to a stressed scenario flows through consistently rather than requiring loss rates to be manually re-entered for each case — see Banking Scenario Analysis for how scenarios should be structured across the model generally, and Stress Testing Models for how a specific adverse scenario translates into a stressed loss forecast.
Portfolio Composition Effects¶
Because loss rates are held at the segment level, a shift in where forecast growth is concentrated changes the blended loss rate across the total book even without any change to the individual segment-level assumptions. Growth concentrated in a higher-risk segment raises the blended rate; growth concentrated in a lower-risk segment lowers it. A single blended loss-rate assumption applied to total loans cannot capture this composition effect at all.
Reconciling to the Allowance Roll-Forward¶
The forecast provision charges for each period should feed directly into the allowance for credit losses roll-forward, so the projected balance sheet reserve remains internally consistent with the income statement charges driving it across the entire forecast, not only in the current period.
Common Construction Pitfalls¶
- Holding loss-rate assumptions static across the entire forecast period rather than updating them as scenario conditions change.
- Failing to link segment-level loss rates explicitly to the scenario in use, requiring manual re-entry when scenarios are switched.
- Modelling loss rates only at a blended, total-book level, missing the composition effect of growth concentrated in a specific segment.
- Projecting provision charges without reconciling them to the allowance roll-forward across the full forecast horizon.
Continue Reading¶
Prerequisites¶
- Credit Loss Provisions — the parent guide
- Loan Portfolio Modelling
Related Technical Guides¶
Related Glossary¶
How OXXON tests thisRun a free structural check with FMAE
Frequently Asked Questions
What is loan loss forecasting?
The forward-looking extension of segment-level credit loss provisioning, projecting how expected loss rates evolve across the forecast period as macroeconomic conditions and portfolio composition change, rather than holding provisioning assumptions static from the base period onward.
Why can't loss-rate assumptions simply be held static across the forecast?
Because a loss rate calibrated to historical experience at the start of a forecast can become materially unrepresentative several periods forward as economic conditions or the portfolio's risk composition genuinely change — a static assumption silently understates or overstates the true forward-looking credit cost.
How should the forecast connect to defined economic scenarios?
By explicitly linking segment-level loss-rate assumptions to the scenario in use (base, stressed, or otherwise), so that switching scenarios flows through consistently to the loan loss forecast, rather than requiring loss rates to be manually re-entered for each scenario.
How does portfolio composition change affect the forecast even without a loss-rate change?
If forecast growth is concentrated in a higher-risk segment, the blended loss rate across the total book rises even if none of the individual segment-level loss rates themselves changed — a model built at the segment level captures this automatically, while a single blended assumption would miss it entirely.
How does the loan loss forecast connect to the allowance for credit losses?
The forecast provision charges each period should feed directly into the allowance roll-forward, so the projected balance sheet reserve stays internally consistent with the income statement charges driving it across the entire forecast period, not just the current period.
How does this guide relate to Stress Testing Models?
This guide covers the forward-looking loss-rate projection mechanics generally; Stress Testing Models covers how a specific adverse scenario should be translated into a stressed variant of this same loss forecast.
Related Articles
Credit Loss Provisions
Credit loss provisioning is the income statement charge that builds up the allowance for credit losses held against a bank's loan portfolio. Provisions should be derived from portfolio-segment loss-rate assumptions applied to segmented loan balances — not a single blended provisioning rate applied to the total book — since default risk varies substantially by product type and risk grade. This guide covers how to structure that segment-level provisioning build and how it connects to the allowance roll-forward on the balance sheet.
Loan Portfolio Modelling
Loan portfolio modelling is the asset-side counterpart to deposit modelling: the loan book should be segmented by product type, risk grade, or business line, each carrying its own origination, repayment, yield, and expected loss assumptions. This guide covers how to structure that segmentation, how to roll forward segment-level balances period over period, and how the segmented output feeds both the interest income build and credit loss provisioning.
Stress Testing Models
A bank stress test should vary the same volume, rate, and credit-loss drivers already present in the base model under a defined adverse macroeconomic scenario, rather than being built as a separate, structurally disconnected stress workbook that cannot be reconciled back to the base case. This guide covers how to structure a stress test as a set of parameter overlays on the existing model, how to translate a macroeconomic scenario into the specific driver changes it implies, and how the resulting capital and liquidity impact should be presented against the base case.
Banking Scenario Analysis
Scenario analysis in a bank model means building multiple forward-looking cases — a base case and one or more alternative cases — as parameter variations of the same underlying model structure, not as separate, disconnected workbooks. This guide covers how to structure a bank's scenario framework generally, how scenarios should be selected and switched cleanly, and how stress testing and loan loss forecasting fit as specific, more prescriptive applications of this same underlying discipline.
Allowance for Credit Losses
The allowance for credit losses is a contra-asset account on a bank's balance sheet, representing the reserve held against expected credit losses on the loan portfolio. It is built up through periodic provision charges against the income statement and drawn down as specific loans are written off, following the same roll-forward discipline a corporate model applies to a bad debt reserve, but at a scale and centrality that makes it one of the most closely scrutinized figures on a bank's balance sheet.
Non-Performing Loan Ratio
The non-performing loan (NPL) ratio measures non-performing loans — those in significant default or unlikely to be repaid in full without recourse to collateral — as a percentage of a bank's total loan book. It is the core asset-quality indicator, and should be read alongside the provision coverage ratio, since a rising NPL ratio without a corresponding increase in provisioning coverage signals building, unrecognized credit risk.