Comparison
Internal audit found nothing. Here is why that proves little.
Sampling detects deviations from recorded state; this leak class is unrecorded state. A clean opinion and material silent leakage coexist comfortably unless the audit explicitly tests reversal completeness — a control we hand over ready-made, pro-audit posture included.
Mechanics of the miss
Sample-based testing compares records against records. Unrecorded movements appear in neither column, hence pass every sample by construction.
The control that would catch it
Re-performance of reversal completeness on a refund sample: expected-versus-actual recomputation per sampled charge. Procedure documented in our checklist; auditors may lift it directly.
Point-in-time versus continuous
Annual control testing versus event-driven checking — frequency itself is the finding. Continuous trails satisfy evidentiary appetites modern audit functions increasingly express.
Common questions
Is this a SOX control?
We never claim compliance outcomes — we describe the control precisely and let your auditors map it into their framework.