Once an obligation qualifies for over-time recognition, the work is measuring how far along it is. This lesson tests the three over-time criteria more closely, compares output and input methods, adjusts cost-to-cost for costs that don't show progress, and recognizes losses on contracts.
The three criteria, more closely
The five-step lesson lists the criteria. Three details decide most cases:
- Customer consumes the benefit as the entity performs. When that isn't obvious, ask whether a replacement provider taking over midway would need to substantially redo the work done so far. Assume the contract could be handed over, and that the replacement wouldn't get any asset the entity keeps. If no redo is needed, the criterion is met.
- Customer controls the asset as it's built. Work on the customer's own land or equipment, such as an addition to its building, usually qualifies.
- No alternative use, plus a right to payment. A contractual restriction counts only if it is substantive (the customer could enforce it); a practical limitation means redirecting the asset would cause a significant loss, as with a design unique to one customer. The test is made at inception and not revisited unless a modification substantively changes the obligation. The right to payment must exist throughout the contract if the customer cancels for any reason other than the entity's failure to perform, and it must cover costs plus a reasonable margin, not only lost profit. A payment schedule alone doesn't prove the right exists.
Choosing a measure of progress
One method is used for each obligation, applied consistently to similar obligations, and progress is remeasured at every reporting date. A revised measure is a change in estimate, so the cumulative effect is recognized in the period of the change.
- Output methods measure value delivered to the customer: milestones reached, units delivered or produced, appraisals, surveys of work done, time elapsed. Units delivered misstate progress if the customer already controls work in process that hasn't been counted.
- Input methods measure the entity's effort: costs incurred, labor or machine hours, resources consumed, time elapsed.
- Right to invoice: if the amount the entity can bill matches the value of its work to date, such as a fixed fee per inspection completed, it may recognize revenue equal to the amount it can invoice.
If progress can't yet be reasonably measured but the entity expects to recover its costs, it recognizes revenue equal to costs incurred until it can.
Cost-to-cost and costs that don't show progress
Under cost-to-cost, progress is costs incurred to date divided by total expected costs. Two kinds of cost are taken out of the measure:
- costs of significant inefficiency not reflected in the price, such as unexpected waste or rework, which are expensed as incurred;
- uninstalled materials: a good that isn't distinct, that the customer controls well before it is installed, whose cost is significant to the total, and that the entity buys from a supplier without significant design involvement. Its cost leaves both the measure and the price, and revenue equal to its cost is recognized when control passes (zero margin).
Example: an engineering firm integrates a production line in a customer's plant for 2,000,000, one obligation measured by cost-to-cost. Expected costs, which assumed no rework, are 1,500,000, including a 450,000 robotic welding cell bought from a robot maker with no design input from the firm. The cell isn't distinct from the integration work, and the customer controls it on delivery in Year 1; it is integrated in Year 2. Year 1 other costs are 378,000, including 42,000 of unplanned rework.
| Item | Amount |
|---|---|
| Costs that count: 378,000 - 42,000 | 336,000 |
| Expected other costs: 1,500,000 - 450,000 | 1,050,000 |
| Progress: 336,000 / 1,050,000 | 32% |
| Price excluding the cell: 2,000,000 - 450,000 | 1,550,000 |
| Revenue: 32% x 1,550,000 + 450,000 | 946,000 |
| Cost of sales: 378,000 + 450,000 | 828,000 |
| Gross profit (32% x 500,000 - 42,000) | 118,000 |
Loss contracts
ASC 606 has no onerous-contract rule, so construction-type and production-type contracts in the scope of ASC 605-35 keep their own: when expected costs exceed the consideration expected (measured by ASC 606's price rules but without the constraint, and adjusted for credit risk), the entire loss is recognized as soon as it becomes evident. The test is at the contract level (contracts combined under ASC 606 are tested as one unit), or by performance obligation as a policy election applied consistently to similar contracts (ASU 2016-20).
Example: a 5,000,000 contract reports 300,000 of profit in Year 1 (costs 1,200,000 of 4,000,000 expected). In Year 2, costs to date reach 3,120,000 and total expected costs rise to 5,200,000, a 200,000 loss. Year 2 reports a 500,000 loss: the 300,000 already recognized plus the 200,000. Revenue is 60% x 5,000,000 - 1,500,000 = 1,500,000, costs incurred are 1,920,000, and a provision of 80,000 covers the loss on the remaining work (2,000,000 of revenue against 2,080,000 of costs).
On the exam
Traps: a termination payment covering costs only treated as a right to payment, uninstalled materials left in the cost-to-cost ratio, rework counted as progress, and a loss recognized only in proportion to progress.