What is points liability?
Points liability is the amount a company owes its members in unredeemed loyalty points, recorded on the balance sheet. When a member earns points, the business has taken on an obligation to provide future value: a discount, a free product, a reward, at a later date. Accounting treats that obligation as a liability until it is settled.
The logic is that the points are not a marketing giveaway sitting outside the financial statements. They are part of what the customer paid for. A shopper who spends and earns points has bought two things: the product in the basket now, and the right to a reward later. The cash received covers both. Recognizing all of it as revenue at the moment of sale would overstate current income and ignore the reward still owed to the member.
This is also why the points are not booked as a simple expense or a provision at issuance. They are not a cost the company chose to incur. They are consideration the company has already collected for a good it has not yet delivered, which is a different item entirely and belongs on a different side of the statements.
So the liability answers a specific question: of the cash already collected, how much belongs to points the company has not yet delivered on. That amount is held back from revenue and carried forward. It grows as members earn and shrinks as they redeem or as points expire. For a large program the balance runs into significant sums, which is why revenue standards address it directly rather than leaving it to management judgment.
How do ASC 606 and IFRS 15 treat points?
Two standards govern the treatment: ASC 606 in United States GAAP and IFRS 15 internationally. They converge on the same core idea. Loyalty points a customer earns through a purchase are a separate performance obligation, often described as a material right, because the points give the customer something they would not have received without the transaction.
This was a deliberate change from the older approach. Earlier guidance let some companies account for loyalty awards at the incremental cost of providing the reward, a cost-accrual method. Under the current standards, points are treated as a deferred-revenue obligation instead, which generally holds back more of the sale than a cost accrual did and ties the release of that revenue to the customer actually using the reward.
The material-right test is what determines whether points are a separate obligation at all. Points a customer earns by buying create a right they would not have had otherwise, and that right is treated as a distinct promise the company must satisfy. This differs from a general discount available to everyone, which is not a separate obligation because it grants no incremental right. The distinction decides whether revenue is deferred. Points that meet the material-right test carry deferred revenue. A blanket price reduction does not, because there is no future performance owed. Programs that blur the two, treating ordinary discounts as loyalty obligations or the reverse, misstate the liability in one direction or the other.
Both standards apply a five-step model to the sale. The company identifies the contract with the customer, identifies the distinct performance obligations in it, determines the transaction price, allocates that price across the obligations, and recognizes revenue as each obligation is satisfied. In a loyalty sale, the goods bought today and the points earned today are two distinct obligations sharing one transaction price.
The consequence follows directly. Revenue assigned to the goods is recognized now, when the goods change hands. Revenue assigned to the points is deferred, held as a contract liability, and recognized later, when the points are redeemed or when the right to redeem them lapses. The standard does not treat the points as a cost of sale at issuance. It treats them as revenue the company has received but not yet earned, which is the distinction that puts them on the balance sheet rather than straight through the income statement.
How is the deferred amount calculated?
The amount deferred is set by allocating the transaction price across the obligations in proportion to their standalone selling prices. Standalone selling price is what each element would sell for on its own. The goods have an observable price. The points need an estimated one, because points are not sold separately in an open market.
That estimate reflects the value of a point to the customer, adjusted for the likelihood it will be redeemed. Two inputs shape it. The first is the redemption value of a point: what a member receives when they burn it, expressed per point. The second is the probability of redemption, since a point that will never be used carries a different expected value than one that will. The standards require the estimate of standalone selling price to consider both the value and the chance of redemption, rather than assuming every point issued will be honored.
Once each obligation has a standalone selling price, the transaction price is split on a relative basis. If the points represent a given share of the combined standalone value, that same share of the price collected is deferred into the liability, and the remainder is recognized against the goods. The mechanism is proportional allocation, not a flat carve-out, so the deferred amount moves with the estimated value of the points rather than sitting at a fixed figure.
Because the estimate depends on program-specific behavior, the standalone selling price of a point is derived from the program's own data: the mix of rewards members choose, the value those rewards carry, and how many points members tend to use. Different programs land on different amounts from the same mechanics, because the inputs differ, not because the rule differs.
How does breakage affect the liability?
Breakage is the portion of points a company expects will never be redeemed. It matters to the liability because a point that will never be burned still had revenue deferred against it at the sale, and that revenue has to be recognized eventually rather than deferred forever.
The standards handle this through the expected redemption pattern. A company estimates how many of the points issued will ultimately be redeemed, which by implication estimates how many will break. Revenue in the liability is then recognized in proportion to the pattern of actual redemptions, measured against total points expected to be redeemed rather than against total points issued. Because the denominator excludes points expected to break, the deferred revenue is released in step with redemptions and is not left stranded on the balance sheet after the member base has effectively stopped using it.
The estimate is not fixed. It is revisited as the program accumulates history and as member behavior changes, and revisions flow through as the expectation updates. On breakage the two standards are converged, and the wording tracks almost exactly. ASC 606-10-55-48 and IFRS 15 paragraph B46 both require expected breakage to be recognized in proportion to the pattern of rights the customer exercises, and both require the company to apply the constraint on variable consideration when estimating it, so revenue is released only to the extent that a significant reversal is not expected to occur later. A company that cannot conclude it is entitled to a breakage amount does not estimate one at all. It recognizes that revenue only when the likelihood of the member exercising the remaining rights becomes remote. Both standards also require a company to check whether unclaimed-property law in its jurisdiction obliges it to remit unredeemed value to the state, in which case the amount is a liability owed to the government rather than revenue.
The practical effect across both frameworks is the same. The liability reflects points the company still realistically owes, and the value of points it no longer expects to honor is returned to revenue on a disciplined, evidence-based basis rather than in a single discretionary write-down. The quality of that estimate rests on the quality of the program's redemption history.
How does the liability move to revenue?
The liability moves off the balance sheet in three ways, and each has a defined trigger.
Redemption. When a member burns points for a reward, the deferred revenue allocated to those points is recognized. The obligation has been satisfied, so the amount held against it becomes earned revenue. This is the primary path, and it is why the timing of redemptions, not issuance, drives when program revenue lands.
Expiry. When points lapse under the program's rules, the obligation ends without a reward being given, so there is no performance obligation left to satisfy. What lands at that moment is smaller than it looks, and this is where programs most often misread the standards. A company that estimates breakage has already been releasing the revenue on points it expected to lapse, in proportion to the redemptions of the points that were used. Expiry confirms an estimate that was recognized along the way rather than triggering a fresh block of revenue, and only the residual between estimate and outcome lands at the date itself. The full amount is recognized at expiry only in the other case, where the company could not conclude it was entitled to a breakage amount, and so waits until the chance of redemption becomes remote.
Remeasurement. As redemption expectations and estimated point values are updated, the carrying amount of the liability is adjusted so it continues to reflect the obligation outstanding. Disclosure requirements then ask companies to explain the contract-liability balance and the revenue recognized from it, so the movement is visible to readers of the accounts.
All three depend on one thing: an accurate, auditable record of every point issued, redeemed and expired, at the member and transaction level. A loyalty platform is where that record lives. GRAVTY®, Loyalty Juggernaut's platform, maintains a complete transaction-level ledger of earn, burn and expiry events tied to member identity, which is the source data finance and audit teams reconcile the liability against. The accounting treatment is set by the standards. The ability to apply it faithfully depends on whether the underlying system can produce the granular record the standards assume already exists.
