Carry phase

Units-of-production amortisation keyed to token consumption

Carry, live question

A units-of-production profile allocates the amortisable amount in proportion to forecast token consumption rather than evenly over time. IAS 38 permits it where it better reflects the pattern in which the asset's benefits are consumed IAS 38 §98. It measures use of the asset, so it is distinct from a prohibited revenue-based method, but it stands or falls on the credibility of the consumption forecast.

How the profile works

The amortisable amount is spread across periods in proportion to the tokens the asset is forecast to consume in each. A period expected to carry more of the workload takes more of the charge. The forecast total tokens over the life is the denominator; each period's forecast tokens over that total gives its share of the charge. The token ledger provides the actuals against which the forecast is later trued up.

Carrying-value ladderIllustrative example

A front-loaded token-consumption forecast (5:3:2) pulls more charge into the early periods. Illustrative example.

When it beats straight-line

A consumption profile is the better reflection of benefit where usage is genuinely uneven and forecastable: an asset whose adoption ramps and then plateaus, or one built for a workload that tapers. The test is whether the pattern of benefit can be determined reliably; if it can and it is uneven, a units-of-production method is not just permitted but preferable to straight-line IAS 38 §98.

The risks of keying to consumption

A speculative forecast is worse than straight-line, because it manufactures a pattern the evidence does not support.
Front-loading the charge can look like earnings management if the consumption case is thin, so the forecast basis must be documented.
The method needs truing up against actual token consumption, adding reconciliation work each period.

An honest position

Whether a token-consumption profile better reflects benefit than straight-line is a matter of judgement on which preparers reasonably differ. The defensible route is to use it only where the consumption forecast is evidenced and material to the pattern, and to fall back to straight-line where the pattern cannot be determined reliably, exactly as the standard directs IAS 38 §98.