Payment controls

Can a payment run be reconstructed from its approvals?

A proposed protocol for testing whether a payment run record shows the batch, the approvers, and the bank confirmation.

This research is general operational information, not legal, tax, accounting, employment, benefits, privacy, cybersecurity, insurance, or financial advice.

Can a payment run be reconstructed from its approvals? evidence review illustration

Question and scope

This is a proposed research protocol, not a report of completed findings. It asks whether a payment run can be reconstructed from its approvals and bank record. The result would describe the run record. Name the owner in the same line as the action.

The protocol does not measure whether each payment was correct. It measures whether the control evidence is complete. Review the first few notes together before the process expands. Write the deadline as a date and a local time.

Population and unit of analysis

Define the population as payment runs during a stated window. The unit is one run, with the batch, the approvers, the bank file, and the confirmation recorded. Ask whether a new reviewer could act on the record without a conversation. Record the exception and the person who approved it.

Link the run to the invoices so the same payment is not counted twice. Test the rule on one real transaction before applying it to the run. Update the template when the same question appears twice. Keep the boundary visible in the place where the work is assigned.

Definitions and measures

Define reconstruction as the ability of a reviewer to state the batch total, the approvers, and the release time using only the run record. Record the date and the source so the next reviewer can repeat the check. Keep the original note beside the corrected version. Keep the evidence with the record rather than in a separate log.

Report reconstruction and completeness separately. A complete file can still be missing the approval. Ask whether a new reviewer could act on the record without a conversation. Keep the evidence with the record rather than in a separate log.

Evidence standard

Use the bank file, the approval record, and the bank acknowledgement. A note that says approved without a name is weak evidence. Keep the evidence with the record rather than in a separate log. A control that depends on memory is not yet a control.

Keep the original run record and the corrections. A later adjustment should show the old value. Review the first few notes together before the process expands. A control that depends on memory is not yet a control.

Sampling and collection

Use a census of the runs if the volume is manageable. If sampling is necessary, stratify by amount and approver. Select the sample before reviewing reconstruction. A short note with four fields is easier to use than a long narrative.

Pilot the review on a few runs and compare the reviewers. Clarify what counts as a complete approval. Update the template when the same question appears twice. Write the deadline as a date and a local time.

Analysis and reporting

Report reconstruction and completeness by run size and approver. Show the runs that could not be reconstructed. Record the date and the source so the next reviewer can repeat the check. If the source is unclear, route the question instead of guessing.

Present the results as a description of the record, not a judgment about the team. Keep the original note beside the corrected version. A control that depends on memory is not yet a control. Ask whether a new reviewer could act on the record without a conversation.

Limits and sources of error

The bank record may not show the internal approval. The approval may be in a separate system. If the source is unclear, route the question instead of guessing. Record the exception and the person who approved it.

The run may have been corrected after release. Note the date. Keep the boundary visible in the place where the work is assigned. Test the rule on one real transaction before applying it to the run.

Review cadence

Repeat the review after a process change or at the same interval each quarter. Compare the same run types. Test the rule on one real transaction before applying it to the run. A control that depends on memory is not yet a control.

Assign the follow-up to the treasury owner and set a date. The protocol should produce a specific decision about the run record. Record the date and the source so the next reviewer can repeat the check. Keep the original note beside the corrected version.

Useful internal handoffs

Sources