Expected credit losses on government and mega-project receivables

A low historical default rate on government and mega-project receivables tempts many contractors to apply a token or zero expected credit loss, but IFRS 9 requires a forward-looking lifetime ECL from initial recognition for trade receivables and contract assets without a significant financing component (the simplified approach), regardless of counterparty type; where a significant financing component exists, as it may for long-dated retentions, lifetime ECL from day one is a policy choice, otherwise the general model applies; understating ECL overstates net receivables and profit, and is especially risky where oil-price or budget-cycle stress could plausibly delay (even if not ultimately default on) government payments.

What is the historical pattern of payment delay, not just actual default, on receivables from government ministries, municipalities, and mega-project owners? Are receivables and contract assets segmented into a provision matrix by counterparty type reflecting genuinely different credit risk profiles? Does the simplified lifetime ECL approach apply because there is no significant financing component in these balances? Are forward-looking macroeconomic factors (e.g. fiscal budget cycles, oil price) incorporated into the loss rate rather than relying solely on historical experience?

1. Apply the simplified approach mandatorily to trade receivables and contract assets without a significant financing component: recognize lifetime expected credit losses from initial recognition, with no need to track significant increases in credit risk.
2. Segment the population into a provision matrix by shared credit risk characteristics, commonly counterparty type (government, semi-government, private developer), geography, and payment history, rather than a single blanket rate.
3. Incorporate reasonable and supportable forward-looking information (macroeconomic outlook, oil price/budget cycle indicators, sector-specific payment trends) as an overlay to historical loss rates.
4. Retentions and other long-dated receivables or contract assets may contain a significant financing component. For these, lifetime ECL from initial recognition is a policy choice: either elect the simplified approach for that class and apply it consistently, or apply the general model and assess staging (12-month versus lifetime ECL) based on whether credit risk has increased significantly since initial recognition. Document which approach applies to retentions.
5. Reassess and update the provision matrix and overlay at each reporting date.

Build the provision matrix from historical loss and delay experience (default rates or, absent actual defaults, the credit-equivalent cost of chronic late payment) segmented by counterparty type and adjusted for current conditions and reasonable forward-looking macroeconomic indicators; document the basis for any near-zero loss rate applied to government counterparties, since 'government' is not itself a risk-free credit rating. Measure the loss allowance as the probability-weighted present value of expected cash shortfalls over the expected life of the receivable, discounted at the original effective interest rate (or an approximation for short-duration receivables where discounting is immaterial).

Disclose: (1) the loss allowance reconciliation (opening balance, charge, write-offs, releases, closing balance) by class of receivable; (2) the provision matrix or methodology, including loss rates applied by counterparty segment; (3) credit risk concentration by counterparty type (government, semi-government, private); (4) significant assumptions and forward-looking adjustments applied; (5) receivables written off during the period and amounts still subject to enforcement activity.

• Applying a zero or token ECL rate to government receivables solely because '
• government does not default,'
• ignoring that chronic delay itself represents an economic loss requiring a time-value adjustment
• using a single company-wide loss rate instead of segmenting by counterparty type and risk profile
• basing the provision matrix purely on historical data without any forward-looking macroeconomic overlay
• failing to decide and document the approach for longer-dated retentions that contain a significant financing component (the simplified approach elected as a policy, or the general model with staging)
• not updating the provision matrix when a specific counterparty'
• s payment behavior deteriorates mid-year.