Note

21 March 2026 · Lim Wei Jun

Maker-checker gaps that keep appearing in origination files

Dual control on paper is not the same as dual control in the application. Sampling still finds the same three patterns.

Credit officer reviewing an application pack at her desk

Most origination policies in Malaysia still say that one person prepares and another person checks. The origination application is supposed to enforce that split. In file samples, the split fails in three ways that look minor until you follow the money.

The first pattern is shared credentials, or a checker who is also the maker on the next screen because the application allows the same officer to complete both steps in one sitting. The log shows two roles. The timestamps show one person. We report that as a control failure, not as a training anecdote.

The second pattern is a checker who can approve without seeing the evidence the maker attached. If the application stores images in a side panel that is optional to open, dual control is ceremonial. We test this by asking the checker, on a live walk-through, to reconstruct the file from the application alone. Hesitation is not a finding. A path that never presents the evidence is.

The third pattern is post-approval editing. A file is approved, then payee details, fees, or a special condition are changed without returning the file to a checker. If the origination application does not lock those fields, or does not force a re-approval, the original checker has signed something other than what was disbursed.

None of this requires a laboratory. It requires a sample, the user-access list for the period, and enough time on the operations floor to watch a file move. We ask for that access in the engagement letter. Where access is refused, the limitation is stated in the report rather than filled with a polite assumption.

If your internal audit plan already includes origination, this review is often commissioned as a focused piece of fieldwork rather than a full control audit. The output is a short findings paper and a workbook of the files in the sample. It is written so a credit manager can change the matrix next month, not so a brochure can claim that dual control exists.

All notes · Request a scoping meeting