When Compliance and Finance Report Different Numbers

A board pack arrives containing two figures for redress paid in the quarter. The compliance report says £184,000. The management accounts say £211,000. Nobody noticed until a non-executive director asked which one was right, and the answer took eleven days to produce.

Variations on this happen constantly in FCA-regulated firms, and they are becoming more common rather than less. The reason is structural: compliance monitoring has become data-led, and almost all the data it runs on is owned, produced or reconciled by finance. Two functions are now reporting on the same underlying transactions to the same board, using different systems, different definitions and different cut-offs.

When the numbers diverge, the question lands on the financial controller. This is a guide to why it happens, where the gaps come from, and what to do about it.

Why this is a newer problem than it looks

Compliance monitoring used to be a process discipline. A reviewer took a sample of files, checked whether the procedure had been followed, and wrote up findings. Finance barely featured, because the question was whether people did the right things rather than what the numbers said.

Two shifts changed that. The Consumer Duty moved the test from process to outcomes, which means monitoring now has to evidence what actually happened to customers — what they were charged, what they received, what it cost them, how complaints and redress resolved. Those are financial quantities. Separately, monitoring functions have adopted data analytics in place of pure sampling, which means they query populations rather than inspecting files.

The practical consequence is that the compliance function now routinely produces numbers. It produces them from source systems, extracts and reports that finance either owns or reconciles — and it produces them without the month-end discipline that governs finance’s own figures. SMF Capital’s guide to compliance monitoring sets out how a modern monitoring programme is designed and what the data-led approach involves, which is useful background for understanding where your figures are being drawn from.

Where the gaps actually come from

In almost every case we see, the divergence is not an error. Both numbers are defensible. They answer slightly different questions, and nobody wrote the questions down.

Cause	What it looks like	Why both numbers are arguably right
Cut-off	Compliance counts when the case was closed; finance counts when the payment cleared	Different events, days or weeks apart
Accruals	Finance includes a provision for redress agreed but unpaid; compliance counts cash out	One is the obligation, the other the settlement
Scope	Compliance counts all complaints; finance counts only those hitting a redress cost code	Goodwill gestures, waived fees and write-offs may sit elsewhere
Gross versus net	Compliance reports the customer’s total; finance reports net of recoveries or insurance	Both correct, different question
Entity boundary	Compliance reports the regulated entity; finance reports the group or consolidated position	Appointed representatives and unregulated subsidiaries fall differently
Source system	Compliance pulls from the CRM or case system; finance from the ledger	The two were never reconciled and drift over time
Restatement	Finance adjusts prior periods after a correction; compliance keeps its original report	Finance’s figure moves after compliance’s is published

The last two are the most corrosive, because they compound quietly. A case management system and a general ledger that were aligned at implementation will diverge over several years through manual adjustments, cancelled transactions, reclassifications and corrections that only one side applies. By the time anyone compares them the gap is large enough to be embarrassing and old enough to be hard to explain.

Why it matters more than an internal tidying exercise

It would be easy to treat this as housekeeping. It is not, for three reasons.

It undermines both reports

A board presented with two figures for the same thing does not conclude that one function is right. It concludes that it cannot rely on either, and that scepticism carries across to every other number in the pack. A finance function’s authority rests on its figures being the ones people stop arguing about.

It is a finding waiting to happen

Data integrity sits squarely within what monitoring tests. A reviewer who establishes that the firm cannot reconcile its own customer outcome data to its ledger will rate that highly, because it means the firm cannot evidence the outcomes it is reporting. That is a problem for the firm’s FCA obligations as much as a finance one, and it attaches to the senior manager accountable for the relevant area rather than to a spreadsheet.

It surfaces at the worst moment

Nobody discovers this calmly. It emerges during a skilled person review, an FCA information request, a due diligence process or an audit — when there is no time to investigate and the inconsistency itself becomes the story. The firms that handle it well are the ones that found it themselves, eighteen months earlier, during a routine comparison nobody was required to do.

The reconciliation that should exist and usually does not

The fix is not complicated. It is a periodic reconciliation between the compliance function’s reported figures and finance’s ledger, performed with the same discipline as any other control account reconciliation.

What it needs:

1. An agreed definition for each reported metric — written down, signed off by both functions, specifying the event that triggers inclusion, the entity boundary, whether figures are gross or net, and the cut-off convention.
2. A named owner on each side. In most firms this is the financial controller and the compliance officer or monitoring lead. Not a committee.
3. A stated frequency tied to the reporting cycle — quarterly for most firms, monthly where volumes are high or the firm is under supervisory attention.
4. A documented reconciling items schedule. Differences are expected and fine; unexplained differences are not. The schedule is the deliverable, not a zero difference.
5. An escalation threshold above which an unexplained variance goes to the risk committee rather than being absorbed.

This is ordinary control account discipline applied to a boundary nobody owns. Our guide to internal controls for growing businesses covers the underlying mechanics, and the designing financial controls piece covers making them evidenceable. The only unusual feature here is that the counterparty is another function rather than a bank statement.

Who should own it

The honest answer is finance, and finance is usually reluctant.

The reluctance is understandable. The data lives in systems finance does not administer, the definitions were set by compliance, and the whole thing feels like absorbing somebody else’s problem. But the ledger is the firm’s authoritative record of what money moved, and when two accounts of the same transactions disagree, the function that owns the authoritative record is the one that has to explain the difference. That is true whether or not the task has been formally allocated.

There is also a positional argument. A financial controller who establishes this reconciliation becomes the person who can answer a regulator’s question about customer outcomes with a number that holds up. That is a materially stronger position than being the person who produced one of two conflicting figures. The FCA’s expectations of a finance function increasingly run in this direction — towards finance as the evidential backbone rather than the statutory reporting corner.

A first pass you can run this quarter

Pick the three numbers that appear in both the compliance report and the management accounts. In most firms they are redress or remediation paid, complaint-related costs, and fees or charges by customer segment.

For each one, ask four questions. What event causes a transaction to be counted? Which entities are in scope? Is the figure gross or net? What is the cut-off? Then run both figures for the same period and compare.

In our experience of regulated finance functions, the first comparison produces a variance in roughly two cases out of three, and the variance is almost always explicable once the definitions are written down. The value is not in the number — it is in discovering that two functions had been using the same word to mean different things, before someone external asks which one the board relied on.

The hiring angle

This problem is a reasonable proxy for regulated experience. An FC who has worked inside an authorised firm will recognise the scenario immediately and will usually have a view on who should own the reconciliation. An FC from unregulated industry — however technically strong — tends to hear it as an odd question about data cleansing.

It is one of the reasons the regulated FC market is genuinely thin rather than merely competitive, and why we treat it as a distinct search discipline on our FC recruitment for FCA-regulated firms desk. The interview question is simple enough to ask in any first conversation: if compliance and finance reported different redress figures to the board, whose number is right and who fixes it?

The short version

Compliance monitoring now runs on financial data, which means two functions are reporting on the same transactions to the same board. Divergence is normal, explicable and almost never anyone’s fault — but unreconciled divergence is a data integrity weakness that will be rated highly when it is found, and it will be found at the least convenient moment.

The remedy is a written definition for each shared metric, a named owner on each side, and a periodic reconciliation with a documented schedule of reconciling items. It is a modest amount of work, and the firms that do it are the ones that can answer a supervisor’s question about customer outcomes in an afternoon rather than eleven days.

For the wider picture of how monitoring programmes are built and what they test, SMF Capital’s compliance monitoring guide is the fullest treatment we know of for UK regulated firms.