Mortgage broker file review checklist

File review is one of those activities that either genuinely improves a firm or wastes a day a month. The difference is almost never the checklist itself. It is whether the review looks at the reasoning in the file or only at whether the fields are populated.

A file where every box is ticked and the recommendation makes no sense is a worse file than one with a missing address history line.

Decide what the review is for

There are three distinct purposes and they need different approaches.

Supervision of an individual adviser looks at a sample of one person's work over time, to judge competence and identify coaching needs. It is about the person.

Process assurance looks across advisers at one aspect — say, how affordability is evidenced — to find out whether the firm's process holds up. It is about the system.

Pre-submission checking looks at one live case before it goes to a lender, to catch errors while they are still cheap. It is about the case.

Firms that run one review and expect it to serve all three get a checking exercise that catches typos and misses patterns. Run them separately, even if the same person does all three.

How to select the sample

Random sampling is defensible and dull. A mixed approach usually finds more.

Take some at random, so nobody can game it. Add some deliberately chosen for risk characteristics: adverse credit, complex income, later-life lending, interest-only, cases where the client is potentially vulnerable, cases where the fee was unusual, cases where the adviser departed from what the firm normally recommends. Add any case that generated a lender query or a client complaint, because those are free evidence.

Include cases that did not proceed. Files where the client walked away are reviewed almost nowhere and often contain the most interesting information about how the firm sells.

The reasoning tests

These are the questions worth spending your time on. Everything else is bookkeeping.

• Reading only the file, can you say what the client was trying to achieve and by when?
• Is there a stated reason for the recommendation that goes beyond the product being cheapest or the lender accepting the case?
• Were alternatives considered, and is it apparent why they were rejected?
• Where the client's circumstances were unusual, does the file show that the adviser noticed?
• If the client's stated priority and the recommendation appear to conflict, is that conflict addressed anywhere?
• Where a fee was charged, is it clear the client understood it before committing?
• If the term extends materially beyond retirement, or the product is interest-only, is the plan documented?
• Is anything in the file inconsistent with anything else in the file?

That last question catches more than any other single check.

The evidence tests

• Do the figures used in the application match the underlying documents rather than the fact find?
• Is identity and anti-money-laundering evidence present and current?
• Is the source of deposit evidenced, including any gifted element?
• Is the sourcing evidence contemporaneous rather than reconstructed?
• Are the client-facing documents issued at the right points, with the dates to show it?
• Are file notes contemporaneous, dated and attributable to a named person?
• Where a vulnerability was recorded, is there any indication of what was done differently as a result?

Grading and what happens next

A binary pass or fail wastes information. Three or four grades work better, with clear meanings agreed in advance: acceptable; acceptable with observations; requires remediation on this file; requires remediation and a review of other cases by the same adviser.

The critical part is what the grades trigger. If a grade produces nothing except a record, advisers learn quickly that the review is theatre. Each outcome should have a defined consequence — feedback within a set period, a re-check of an agreed number of subsequent cases, additional training, or in serious cases contacting the client.

Feedback given promptly and specifically changes behaviour. Feedback given six weeks later as a summary statistic does not.

Look for patterns, not just faults

At the end of a review cycle, spend an hour on the aggregate. The individual findings matter less than the answer to: which issue came up most, and is it concentrated in one person or spread across everyone?

An issue spread across everyone is a process or training problem and blaming individuals for it will not fix it. An issue concentrated in one adviser is a supervision matter. Getting this distinction wrong is the most common failure in broker file checking.

Practical points that make reviews stick

Fix the time. A review that happens when there is a quiet week never happens.

Keep the checklist short enough that a reviewer can hold it in mind. Forty items produces mechanical ticking. Twenty focused ones produce reading.

Have someone other than the usual reviewer do a batch occasionally. Reviewers develop blind spots as reliably as advisers do.

Review your own reviews once a year. Are the findings changing? If the same three issues appear every cycle, the feedback loop is broken somewhere.

Let advisers see the aggregate findings. Most people would rather not be the one who keeps causing a firm-wide issue.

The regulatory boundary

What your firm must check, how often, how the results must be recorded and how long they must be retained is not something to take from a generic checklist. It depends on your permissions and on whether you are directly authorised or an appointed representative. Networks typically operate their own file-checking regime with its own standards and frequencies, and yours must sit alongside it rather than duplicate or contradict it.

Confirm the requirements with the FCA Handbook or your compliance support and build your checklist from those. The material above is about making the review useful, not about defining what is mandatory.