Methodology

How we arrive at a defensible number

Breakage is a behavioral question, not an accounting one. Our job is to measure the behavior precisely and document it well enough to withstand review.

The core mechanic

Redemption decays, then it flattens

Stored value is redeemed fastest immediately after issuance. As a cohort ages, the redemption rate collapses toward an asymptote — and the gap between that asymptote and full redemption is Breakage.

Finding the asymptote credibly is the entire technical problem. Everything else in the methodology exists to make that estimate defensible.

Cumulative outstanding share by cohort age

Observed decay Terminal asymptote

Principles

Six rules we do not relax

Cohort-based, never blended

Every analysis is built at issuance-vintage level. A blended portfolio rate averages together cohorts that behave nothing alike and hides the aging effect entirely.

Empirical before parametric

We fit the observed curve first and only then apply a functional form for extrapolation, so the shape comes from your data rather than an assumed distribution.

Reconciled to the ledger

Modeled outstanding balances are tied back to your reported liability. If they do not reconcile, the segmentation is wrong and we fix it before proceeding.

Sensitivity, not point estimates

Every rate is delivered as a low, base, and high case with the drivers of the spread identified explicitly.

Exposure separated from opportunity

Potentially escheatable value is quantified and reported apart from breakage, so nothing is double-counted as recognizable.

Reproducible by construction

Inputs, transformations, and assumptions are documented. The same extract run again produces the same result.

Sequence

The analytical pipeline

Curve construction

For each cohort we compute cumulative redemption as a share of issued value by months since activation, producing an empirical decay series.

Segment discovery

Segments are selected where redemption behavior genuinely diverges — denomination, channel, product tier — rather than by convenience of reporting.

Terminal extrapolation

Mature cohorts anchor the terminal asymptote; immature cohorts inherit it with widened bands proportional to their remaining uncertainty.

Aging overlay

Outstanding balances are aged against the fitted curve to distinguish value still inside the active window from behaviorally settled value.

Exposure testing

Settled balances are tested against dormancy characteristics to size potential unclaimed-property exposure by jurisdiction profile.

Reconciliation and documentation

The final position is reconciled to the reported liability and packaged with full assumption documentation for audit review.

Scope boundary

What our analysis is not

CardBreakage produces quantitative behavioral analysis. We do not issue accounting opinions, legal conclusions, or unclaimed-property filings. Recognition and compliance decisions remain with your accounting policy, your auditors, and your counsel — our output is built to inform those decisions, not to replace them.

Engagement

Run your breakage audit

A confidential, analyst-led review of your stored-value liability — expected breakage, redemption behavior, and escheatment exposure quantified against your own data.

No engagement commitment required. Scope is confirmed before work begins.

Run My Breakage Audit Review Methodology