Point program unit economics (JP loyalty funding, float, breakage, CPA)

Wiki route

This entry sits under loyalty index as the unit-economics / funding-mechanics page. It is the economic complement to point liability accounting boundary: that page asks how a point is classified, this page asks whether the program makes money and who pays. It reads alongside the program map in Japan points and loyalty landscape, the data-monetisation flywheel in retail-media points data loop, and the inter-operator transfer mechanics in point exchange network risk. The downstream beneficiary of favourable point economics is group finance: Rakuten FG, NDFG, PayPay FG.

TL;DR

A point program’s economics reduce to a small equation: funding source − redemption cost − operating cost + breakage gain + float yield + monetisation (retail media / finance cross-sell) = program margin. The two structurally interesting terms are breakage (points granted but never redeemed are a near-pure margin gain whose timing is governed by ASBJ Statement No.29 / IFRS 15) and float (the gap between granting a point liability and paying for redemption is a zero-interest funding balance). Whether a program is profitable turns less on the headline reward rate than on who funds the point (merchant vs operator), how high breakage runs, and whether the ID graph is monetised. A “1% back” program is not one economic object; it is at least four very different ones depending on those answers.

The unit-economics equation

For a representative unit of points granted, the operator’s economics are:

Term	Sign	Driver
Funding inflow	+	Merchant-funded points: the merchant pays the operator to issue. Operator-funded: pure cost.
Redemption cost	−	Value of points actually redeemed (the operator must honour them)
Operating cost	−	Platform, settlement, fraud control, data infrastructure, partner management
Breakage gain	+	Points never redeemed → liability released to revenue (timing per accounting standard)
Float yield	+	Outstanding liability is an interest-free balance held until redemption
Monetisation	+	Retail-media + finance cross-sell revenue attributable to the ID graph

The headline “reward rate” only sets the gross redemption cost. Margin is decided by the other five terms — which is why two programs advertising the same percentage back can have opposite economics.

Funding: who actually pays for the point

The single biggest economic split is the funding source, and it does not match what the consumer sees.

Funding model	Who pays	Operator economics	Typical case
Merchant-funded	Partner merchant pays operator per point issued	Operator earns a margin on issuance + keeps breakage/float	Common-point acceptance at partner retailers
Operator-funded (campaign)	Operator’s own marketing budget	Pure acquisition cost; only pays back via data / cross-sell	Wallet “100% 還元” promotions
Issuer-funded (card)	Card issuer funds reward from interchange / fees	Defends card economics; reward is a cost of interchange	Credit-card reward points
Self-funded (group retail)	The group’s own retail margin	Retention spend; pays back as repeat purchase	Store-group points (grocery / convenience)

A common point typically runs a merchant-funded core (margin-positive on issuance, plus breakage and float) with an operator-funded campaign overlay on top. Conflating the two is the classic misread flagged in the accounting-boundary page: the campaign overlay hits P&L immediately and makes the program look loss-making, while the merchant-funded base is quietly margin-positive.

Breakage: the margin engine

Breakage — points granted but never redeemed — is the closest thing in loyalty to a pure-margin gain. Mechanically:

Points the operator never has to honour are a liability that is eventually released to revenue.
Under ASBJ Statement No.29 and IFRS 15, that release is recognised in proportion to the pattern of redemption over the expected redemption period — not all at once on expiry — and only when breakage is a reliable estimate from history. Where it is not estimable, revenue waits until further redemption is remote.
Higher breakage = higher structural margin; but it is also the most estimate-sensitive and review-exposed number in the whole program. Overstating it pulls revenue forward; the Cashless Promotion Council / Payments Japan disclosure norms (comparable redemption-rate, expiry, outstanding-balance reporting) exist partly to stop breakage being used as a revenue-smoothing valve.

Design choices push breakage directly: short expiry, limited-use points (期間・用途限定), and high minimum-redemption thresholds all raise breakage — which is one (rarely stated) reason limited-use points are so prevalent. The accounting framing of this is bucket 3 in the accounting-boundary page.

Float: the interest-free funding balance

Between granting a point and paying for its redemption, the operator holds an outstanding liability it has not yet had to settle. In aggregate this is a large, slow-moving, interest-free balance — analogous to insurance float or prepaid-instrument float:

The longer the average time-to-redemption, the larger the float relative to issuance.
For a large common point or wallet, the outstanding point liability is a standing balance funded entirely by customers at zero interest.
Float is separate from breakage: breakage is points that never come back (a margin gain); float is the carry on points that will eventually come back (a funding benefit while held).

The float reading is what makes point liabilities interesting to a financial group: a point operator inside a bank / telco group (Rakuten FG, NDFG, PayPay FG) contributes a customer-funded balance and a daily-frequency touchpoint at once.

Balance-sheet view: the liability as customer-funded float

Stepping back from the per-unit equation, the aggregate outstanding point balance has a distinct profile on the balance sheet. Once granted, unredeemed points are a standing liability, and four properties together give it the textbook shape of a float:

Property	Implication
Granted before settled	The operator holds value it has not yet paid out
Customer-funded	The balance exists because customers earned, not because the operator borrowed
Interest-free	No coupon is paid on the outstanding point liability
Slow-moving	For a large common point, the aggregate balance is sticky and replenished faster than it drains

How it differs from prepaid stored value

A loyalty point balance and a prepaid stored-value balance (electronic money, gift balance) both look like customer-funded float, but they are not the same instrument:

Dimension	Loyalty point liability	Prepaid stored value
Origin	Granted as a reward (no cash in)	Customer loaded cash
Regulatory home	Loyalty / revenue-recognition accounting	Payment Services Act prepaid regime (asset-preservation, registration)
Cash-equivalence	Usually limited-use, lower cash-likeness	Closer to cash; redeemability / transferability matter
Breakage logic	Expiry-driven, estimate-heavy	Constrained by prepaid rules / unused-balance treatment

The line between them is precisely the funds-transfer vs prepaid boundary — a point that becomes cash-charged or freely transferable can cross into the prepaid regime, changing both its regulation and its float treatment.

The risk side of the float

Customer-funded float is not free money; it carries balance-specific risks an analyst should price:

Risk	What it is	Read against
Redemption surge	A campaign or expiry change accelerates redemption → the float drains and cash goes out faster than modelled	[[loyalty/point-program-unit-economics
Breakage misestimate	Optimistic breakage pulls revenue forward; a true-up reverses it	[[loyalty/point-liability-accounting-boundary
Reclassification	A point that becomes cash-like migrates into the prepaid regime, raising preservation / registration duties	[[payments/funds-transfer-vs-prepaid-boundary
Cross-program leakage	Exchange into other operators / mileage moves value off-balance at a settlement rate	[[loyalty/point-exchange-network-risk

Monetisation: where thin programs turn profitable

A point program judged on funding − redemption − cost alone can look marginal. The terms that flip it are breakage, float, and monetisation of the ID graph — the retail-media + finance cross-sell flywheel set out in the retail-media data-loop page. This is the reconciliation for “loss-making” wallets:

Campaign-heavy wallets book operator-funded grants as immediate expense (P&L looks bad).
The same spend buys the richest ID graph and the deepest daily-frequency funnel.
Retail-media margins and finance cross-sell — higher-margin than the thin retail/payment business the point subsidised — are where it pays back.

So the profitability question is never “what is the reward rate”; it is “merchant-funded or operator-funded, how high is breakage, how large is the float, and is the graph monetised.”

Why this matters for JapanFG / financial analysis

Reward rate is the wrong headline. Two “1% back” programs differ entirely on funding source, breakage, float, and monetisation. Compare those, not the percentage.
Breakage is the quality-of-earnings flag. A program leaning on optimistic breakage to show margin is pulling revenue forward; check whether “ポイント引当金” has migrated to “契約負債” and how redemption assumptions are disclosed (per accounting boundary).
Float + ID graph are why financial groups want point operators. A point inside a group supplies an interest-free customer-funded balance and a cross-sell funnel — the economics that make SMFG / V-Point, NDFG / dポイント, and Rakuten FG internal integration rational beyond marketing.

Sources

Rakuten Point Club official guidance — point program structure and redemption terms.
ASBJ Statement No.29, “Accounting Standard for Revenue Recognition” (収益認識に関する会計基準) — breakage / deferred-revenue timing.
Rakuten Group IR — contract-liability and point-related deferred-revenue disclosures.
NTT docomo IR — dポイント redemption assumptions and revenue allocation.
Payments Japan Association — code-payment redemption / breakage disclosure norms.