Build a sales cohort
Take an illustrative dispatch of 100 items priced at €100 each; customers pay €10,000 immediately. If ten items are expected to come back for a full refund, recognised revenue is €9,000 in this simplified IFRS 15 example and the refund liability is €1,000. Ignore tax and other adjustments here. The original cash receipt still totals €10,000. Treating all of that as repeatable operating cash generation overlooks the amount expected to leave the account.
Estimate what comes back
Suppose each item cost €60 and each expected return costs €5 to recover, with no additional loss in saleable value. The separate recovery asset for ten units is €550. It cannot settle a €1,000 refund: the customer needs cash, while the returned item needs inspection and another buyer. If a garment is damaged or will need a markdown, revise its expected recovery value rather than assuming every returned item immediately restores its original economics.
Follow the calendar
A holiday sales surge can bring collections into December and refunds into January. Compare cohorts by dispatch month, return window and product category. Separate a rising return rate from longer processing delays, then check whether the outstanding refund liability has moved consistently. Reported inventory days can also mislead when stock is in transit from customers. This analysis does not identify a problem from one year-end balance: follow the refunds paid and resale proceeds in the following weeks.
Practical check
Reconcile the sample cohort to cash received, expected cash refunds and saleable units recovered. Never add the recovery asset to cash available for refunds.