Credit Risk: Definition, Measurement & Management

Updated 12 Mar 2026Content was accurate at the time of publication

What is credit risk?

Credit risk is the possibility that a borrower or counterparty will fail to honour contractual repayments, creating a loss for the lender or investor. It is distinct from market risk (price moves) and liquidity risk (inability to meet cash needs): credit risk is about creditworthiness and default. You assume credit risk whenever you lend money, provide trade credit, extend a guarantee, or take exposure to bonds and derivatives.

Credit risk affects banks, non-bank lenders, corporates offering supplier credit, and investors who hold corporate or sovereign debt. It often overlaps with counterparty risk (the chance a contracting party won't settle), settlement risk (failure during the settlement process), and concentration risk (excess exposure to a single borrower, sector or geography).

Types of credit risk

Understanding the types helps you choose the right controls.

Borrower / issuer risk: Risk that an individual, SME or corporate issuer won't repay principal and interest on a loan or bond.
Counterparty risk: Common in derivatives and trading — the counterparty may default before settlement.
Settlement risk: Failure in the settlement process (e.g., payment systems, cross-border settlement).
Sovereign risk: Default or restructuring by a government or an inability to transact due to capital controls.
Concentration risk: Over-reliance on a single borrower, industry or region; raises portfolio volatility.
Industry / portfolio risk: Cyclical sectors (construction, commodities) where correlated defaults can spike losses.

Each type requires different control strategies. Asset-backed solutions such as asset finance or invoice finance can help reduce exposure for small businesses by improving liquidity and asset security.

How credit risk is measured

Credit risk measurement combines borrower-level estimates and portfolio analytics.

Key metrics: PD, LGD, EAD

Probability of Default (PD) — the chance a borrower defaults within a specified time horizon (usually 1 year).
Loss Given Default (LGD) — the percentage of exposure you expect to lose if default occurs after recoveries and enforcement.
Exposure at Default (EAD) — the expected outstanding exposure at the moment of default, including undrawn facilities where appropriate.

These feed the core expected loss relationship:

EL = PD × LGD × EAD

Example: For a AUD 100,000 SME loan with PD = 3% and LGD = 40%:

EL = 0.03 × 0.40 × 100,000 = AUD 1,200 expected loss

Additional measures

Credit scoring and rating scales: Ordinal scales and scores convert financial and non-financial indicators into PD estimates.
Vintage and migration matrices: Track rating changes across time to estimate migration risk.
Portfolio metrics: Credit Value-at-Risk (Credit VaR), expected and unexpected loss, concentration indices and factor models to capture correlation effects.

Common credit risk models and tools

Lenders use a mix of statistical models, rule-based systems and external ratings.

Credit scoring models: Logistic regression, decision trees and scorecards for consumer and SME lending. They map applicant attributes (income, repayment history) to PD.
Internal Ratings-Based (IRB) approaches: Banks with sophisticated risk systems may use IRB methods under prudential frameworks to estimate PD, LGD and EAD for capital calculation.
Machine learning: Gradient boosting and random forests are increasingly used to augment traditional models, particularly when alternative data is available. Governance and explainability are key.
Credit rating agencies: External ratings provide issuer-level assessments used by bond investors and some lenders.
Model validation and governance: Independent validation, back-testing, benchmarking and documentation are essential to meet prudential expectations.

How lenders manage and mitigate credit risk

Lenders use layered controls that reduce exposure and improve recoverability:

Underwriting and screening: Robust affordability checks, industry analysis and stress testing borrower cashflows.
Collateral and security: Mortgages, fixed and floating charges, pledged receivables. Recovery depends on asset quality and enforceability.
Covenants: Financial (DSCR, leverage), affirmative and negative covenants that give early rights to act as borrower condition deteriorates.
Guarantees and credit enhancements: Personal guarantees, parent company guarantees, letters of credit.
Netting and set-off: Reduce gross exposure across contracts.
Credit derivatives: Credit default swaps (CDS) provide transfer of credit risk when used appropriately.
Diversification: Limit concentration by setting sector, obligor and product caps.
Monitoring and early-warning systems: Payment monitoring, watchlists, covenant triggers and automated alerts.

For SME borrowers, practical mitigation can include restructuring receivables via invoice finance to improve liquidity, or asset-backed facilities via asset finance to reduce LGD.

Credit risk and pricing

Credit risk is a core input to how lenders price loans and structure terms.

Risk premium / spread: Lenders add a credit spread above a funding or benchmark rate to cover expected loss, cost of capital and operational costs.
Term and security: Higher PD or LGD typically leads to shorter tenors, higher margins, tighter covenants, or required collateral.
Capital allocation: Under prudential rules, higher risk exposures attract more capital — raising the lender's required return and therefore borrower pricing.
Provisions and impairment: Anticipated credit losses (see IFRS 9 / expected credit loss) affect profitability and hence pricing decisions.

In practice, pricing models convert PD × LGD × EAD into an annualised cost per loan, incorporating a capital charge and target return on equity.

Regulatory, accounting and reporting context

Regulatory and accounting frameworks shape how you measure, provision and manage credit risk.

Prudential supervision (APRA): APRA sets prudential expectations for capital, risk governance and model validation for deposit takers and regulated institutions. APRA publications outline stress testing, IRB model standards and concentration risk management.
Market conduct and credit licensing (ASIC): ASIC enforces credit provider licensing, responsible lending obligations and disclosure requirements.
Monetary and stability oversight (RBA): The central bank monitors financial stability risks arising from credit cycles and lending standards.
Accounting — AASB / IFRS 9: The expected credit loss (ECL) model under IFRS 9 (AASB adoption) moved provisioning from incurred loss to forward-looking expected loss.

Under IFRS 9, you must estimate lifetime ECL for exposures that have experienced a significant increase in credit risk; otherwise a 12-month ECL is recognised. This change affects provisioning volatility and management reporting.

Portfolio credit risk and stress testing

Portfolio-level management captures concentration, correlation and tail risk.

Concentration limits: Set caps by obligor, sector and product to reduce single-point failures.
Correlation modelling: Factor models (macro factors, industry exposures) reflect common drivers of default.
Scenario analysis and stress testing: Model severe but plausible macro scenarios and map PD migrations and LGD increases to capital and provisioning outcomes. APRA expects regulated entities to conduct scenario analysis and reverse stress tests.
Migration matrices: Estimate probability of rating transitions and their impact on expected and unexpected losses.

For non-bank lenders, stress testing highlights liquidity strains as defaults rise; for banks, it informs capital adequacy and buffers.

Practical examples and case studies

SME loan default: You underwrote a AUD 100,000 working-capital facility with PD = 4% and LGD = 50%. EL = 0.04 × 0.50 × 100,000 = AUD 2,000. If a local downturn raises PD to 8%, EL doubles to AUD 4,000, prompting provisioning and potential covenant enforcement. Consider invoice finance or staged drawdowns to reduce EAD and LGD.
Corporate bond downgrade: A rated issuer is downgraded two notches. Market spreads widen, CDS premia rise and you re-evaluate PD via migration matrices; trading desks mark-to-market while credit teams reassess expected loss and collateral triggers.
IFRS 9 provisioning impact: A lender recognised historically low incurred losses. Under expected credit loss accounting, economic forecasts and forward indicators raise lifetime ECL for a subset of its loan book, requiring higher provisions and reduced current period profit.
Collateral enforcement: A defaulted asset-backed facility secured by plant and equipment results in recovery actions; effective PPSR filings and timely repossession reduce LGD materially compared with unsecured exposure.

Emerging considerations and trends

Credit risk assessment is evolving.

Climate and ESG risk: Physical and transition risks shift sectoral credit profiles (e.g., fossil-fuel exposed firms).
Alternative data & AI: Non-traditional signals (transaction data, utilities payments) plus machine learning improve PD estimates, but add governance and explainability challenges.
Cyber risk: Operational shocks and data breaches can induce sudden credit deterioration.
Post-pandemic trends: Sectoral re-allocations and heightened SME vulnerabilities make forward-looking surveillance essential.

Model governance must adapt to data science tools while meeting prudential validation expectations.

Key takeaways

Credit risk = default risk + recovery loss; managed by estimating PD, LGD and EAD and using the expected loss relationship. Measurement combines borrower scoring, rating migration and portfolio analytics. Mitigation uses underwriting, collateral, covenants, guarantees, netting and diversification. Regulatory and accounting frameworks (APRA, ASIC, RBA, AASB/IFRS 9) require robust governance, forward-looking provisioning and stress testing. Keep models validated, stress scenarios realistic, and monitoring systems responsive to early warnings.

FAQ

What exactly causes credit risk?

Defaults caused by cashflow shortfalls, business model failure, macro shocks or legal and operational failures. Collateral quality and enforcement regimes determine recoveries.

How do lenders estimate the probability of default (PD)?

Through credit scoring, statistical models (logistic regression), rating migration studies and benchmark data. External ratings and market signals also inform PD.

What is the difference between PD, LGD and EAD?

PD is the chance of default; LGD is the share of exposure lost after recoveries; EAD is the exposure amount at default. Together they give expected loss.

How does IFRS 9 change credit loss provisioning?

IFRS 9 requires expected credit losses (ECL) using forward-looking information, often leading to earlier provisioning compared with the incurred loss approach.

How can small businesses reduce their credit risk to lenders?

Maintain timely financials, reduce undrawn commitments, provide collateral, use invoice finance or asset finance to reduce unsecured borrowing, and agree realistic covenants.

What is concentration risk and why does it matter?

Overexposure to one obligor, sector or geography increases correlated default risk and potential large losses.

How do credit ratings affect borrowing costs?

Lower external ratings imply higher PD and typically require higher spreads, shorter tenors or additional covenants.

When do lenders use credit derivatives?

To hedge or transfer specific credit exposures (e.g., via CDS) or to synthetically manage portfolio risk when direct asset sale is impractical.

How is credit risk monitored after a loan is made?

Through payment monitoring, covenant testing, periodic financial reviews, watchlists and automated early-warning alerts tied to behavioural or macro indicators.

What are the most important model governance steps?

Clear documentation, independent validation, back-testing, data lineage checks and regular performance monitoring — all aligned with prudential expectations from APRA and accounting rules under AASB/IFRS 9.