Accounting for Customer Returns and Refunds

Under accounting for customer returns and refunds, sellers record two items at the point of sale: a refund liability (for the obligation to repay money) and a corresponding asset (the right to recover the unsold goods). ASC 606 requires both to be recognized in the same period, even before the return window closes.

The Dual Recognition Requirement

When a customer buys an item with a return privilege, the seller faces two questions: how much revenue to recognize now, and how to account for the goods that may come back. Under ASC 606, the answer is to record both a refund liability and a corresponding asset—simultaneously, at the moment of sale.

The refund liability is straightforward: it is the amount the seller expects to pay out as refunds. If a retailer sells 100 shirts at USD 50 each and estimates that 10 percent will be returned, the refund liability is USD 500 (100 × USD 50 × 10%). This liability reduces revenue recognized. The retailer books USD 4,500 in gross revenue and a USD 500 refund liability, netting USD 4,000 in revenue recognized.

The return asset is the flip side. When those shirts come back, the seller regains possession of inventory that still has value. Under ASC 606, sellers must record an asset reflecting the right to recover the returned goods. This asset is typically valued at the amount the company expects to receive from either reselling the returned item or salvaging it, less any costs to recover and resell it.

This dual-recognition approach prevents the financial statements from becoming distorted. Without it, the balance sheet would show only a liability with no offsetting recovery value, understating asset position.

Measurement: The Probability-Weighted Estimate

Companies must estimate the proportion of units they expect customers to return before measuring the refund liability and asset. This is where ASC 606 permits two methods: the expected value method and the most likely amount method.

Expected value method (weighted probability) is more common in large, diverse customer bases. A company selling a thousand SKUs to millions of customers calculates the return rate for each product category or region, then weights them by volume. If one product has a 5% return rate and another a 15% return rate, the company computes a weighted average and applies it to total sales.

Most likely amount method applies when there are only two possible outcomes—high likelihood of no return or high likelihood of full return. For example, a limited-edition collectible with a 2% return rate or a garment with a 40% return rate might fit this model, though practice varies.

Both methods require judgment. A company with a 10-year sales history and stable return patterns can rely on that data. A new product line with no history must forecast returns based on product category benchmarks, customer segment behavior, or management estimates. The precision of the estimate directly affects the accuracy of revenue and asset valuations.

Estimates must be updated each reporting period if new information emerges—a change in return policy, a quality issue, or a shift in customer behavior. When estimates change, companies restate prior periods or adjust the current period’s refund liability and asset accordingly, under the relevant revenue recognition rules.

Recording the Journal Entries

At the point of sale, assume a retailer sells USD 10,000 of goods that it estimates will have a 12% return rate. The cost of goods sold is USD 6,000.

Journal entry at sale:

This entry recognizes USD 8,800 net revenue (USD 10,000 gross less the USD 1,200 refund liability) and USD 8,800 in cost of goods sold (USD 6,000 cost adjustment for expected returns).

At the same time, the company records the return asset:

The return asset is valued at USD 720 (12% × USD 6,000 cost). This reduces reported cost of goods sold to USD 5,280 (USD 6,000 − USD 720), reflecting that the company expects to recover goods worth that amount.

When goods are actually returned within the return window, the company reverses the refund liability and decreases the return asset (or removes it entirely if all expected goods have arrived). If fewer items than forecast actually return, the company adjusts the refund liability downward in the period when actual returns underrun expectations. If more return, the reverse occurs—though in practice, updated estimates account for this before the return window closes.

Return Assets and Valuation Risk

The return asset valuation depends on the probability that returned goods are resalable. For most new, unopened products (books, electronics, clothing in good condition), the recovery value is close to the original cost. For perishable goods, fashion items past their season, or damaged merchandise, the recovery value is lower.

A smartphone seller might assume returned phones (if unopened) are worth 95% of cost; a fashion retailer might assume returned seasonal clothing is worth 50% of cost if it cannot be resold at regular price and must be marked down or sent to outlet stores. This judgment directly affects the return asset amount and, by extension, the reported gross margin.

Companies that lack visibility into the actual resale value of returns must rely on historical data or conservative estimates. Underestimating the return asset overstates cost of goods sold; overestimating it overstates assets and net income. Some retailers track actual recovery rates by product category and update their estimate model annually.

Impact on Key Metrics

The presence of refund liabilities and return assets subtly alters common financial metrics. A company that records USD 1 million in gross revenue and a USD 100,000 refund liability shows USD 900,000 in net revenue. Gross profit margin calculated on net revenue (after refund liability) differs from a margin calculated on gross revenue. Investors comparing gross margins across retailers must check whether each company recognizes refund liabilities in revenue or as a separate line item.

Similarly, the return asset inflates total assets on the balance sheet, which affects asset turnover and return on assets ratios. These effects are usually immaterial for mature, stable-return-rate businesses but can be significant for retailers introducing new product lines or entering markets with unfamiliar customer bases.

Under IFRS 15 (the international equivalent of ASC 606), the same rules apply. The standard is principles-based, emphasizing that both the liability and asset must reflect the “transaction price” of goods expected to be returned. The structure of the journal entries may differ slightly by company and accounting system, but the dual-recognition principle is universal.

See also