Reading the Redress Signal: Underwriting UK MGA Conduct Risk Through Complaint Patterns

Complaints are often treated as a customer-service metric: a number to be logged, trended and reported to the board. For a UK managing general agent, that treatment is too narrow. Complaint patterns, and particularly Financial Ombudsman Service outcomes, are a forward-looking signal of conduct risk. They reveal how delegated authority is being exercised at the point of sale, how claims and cancellations are handled when a customer pushes back, and whether the MGA’s own oversight is catching problems before the ombudsman does.

For capacity providers and acquirers, this matters because conduct risk is rarely visible in the loss ratio at the point of signing. It accumulates quietly in the complaint file. A 2026 underwriting approach should treat the complaints ledger as an intangible exposure with measurable characteristics: velocity, uphold rate, root-cause mix and the adequacy of redress reserves. None of these is a substitute for prudential underwriting. Together, they form a conduct-risk overlay that can be priced, monitored and, where necessary, used to shape delegated authority terms.

Why complaint data belongs in the schedule

A UK MGA operates inside a regulatory architecture that gives complaint data a formal lifecycle. The Financial Conduct Authority’s dispute resolution rules require firms to handle complaints promptly and fairly, to issue final responses within defined periods, and to report complaints data. Where a customer remains dissatisfied, the Financial Ombudsman Service can review the case and make an award against the firm. That sequence creates a paper trail that underwriters can read.

The signal is not simply the volume of complaints. It is the relationship between complaints received, complaints upheld and the reasons behind them. An MGA with a modest complaint count but a high uphold rate may be signalling a systemic issue in product design, sales scripts or claims triage. An MGA with a larger complaint count but low uphold rates may be experiencing friction rather than misconduct. The distinction matters when a capacity provider is deciding whether to extend, tighten or reprice delegated authority.

In due diligence, complaint data also tests the quality of the MGA’s own management information. If the MGA cannot produce complaint categories, root-cause analysis and redress calculations in a consistent format, that is itself an underwriting finding. Conduct risk that cannot be evidenced is conduct risk that cannot be monitored.

Four metrics that repay underwriting attention

Complaint velocity. This is the rate at which complaints arrive relative to policies in force, new business and claims. Absolute counts are easily distorted by growth. Velocity normalises for scale and helps distinguish a book that is growing into a problem from one that is growing away from it. Underwriters should ask for monthly data over at least two years, segmented by product, distribution channel and policy year.

Uphold rate. The proportion of complaints resolved in the customer’s favour, whether by the MGA or by the ombudsman, is a direct indicator of whether the original decision was defensible. A rising uphold rate is more concerning than a rising complaint count, because it suggests that the firm’s first-line response is not reliably correct.

Root-cause concentration. Complaints should be categorised by cause: mis-selling, disclosure, pricing, claims delays, cancellation terms, vulnerable customer handling. Concentration in one category is an underwriting signal about where the MGA’s processes are weakest. A long tail of miscellaneous causes is often less informative than a single category that accounts for a third of upheld complaints.

Redress reserve adequacy. Redress is a liability, not a cost of doing business. The reserve should reflect known complaints, expected uphold rates and the current award limit. Underwriters should test whether the reserve is calculated on a consistent basis and whether it is reviewed at the same cadence as claims reserves. A thin redress reserve in a book with rising uphold rates is an intangible liability waiting to be recognised.

A 2026 diligence sequence for capacity reviews

Start with the complaint register, not the summary slide. Request the underlying data and reconcile it to regulatory returns and board papers. Then test the MGA’s root-cause taxonomy against a sample of actual files. Read the final responses and, where available, the ombudsman decisions. The language of those documents often reveals more about conduct culture than any policy document.

Next, map complaint causes to delegated authority. If complaints cluster around a particular product or distribution channel, the underwriting response may be a tighter product schedule, enhanced monitoring or a specific exclusion. If they cluster around claims handling, the question becomes whether the MGA’s third-party administrator is part of the problem. Conduct risk and operational dependency frequently overlap.

Finally, reflect the findings in the capacity agreement. That may mean a conduct-risk monitoring condition, a periodic complaints data warranty, or a redress reserve adjustment in the MGA’s capital modelling. The point is not to penalise complaint volume. It is to ensure that the capacity provider is paid for the risk it is actually carrying.

The limits of the signal

Complaint data has known limitations. Volumes can be small, especially in specialist lines, which makes statistical inference unreliable. Consumers may complain to the distributor rather than the MGA. The ombudsman’s approach evolves. And a well-run MGA can generate complaints simply by selling a product that customers do not read carefully.

Those limitations argue for triangulation, not neglect. Complaint data should sit alongside Consumer Duty evidence, quality assurance results, claims disputes and regulatory correspondence. It is one lens on conduct risk, but a lens that is unusually close to the customer’s actual experience.

For UK MGAs in 2026, the complaints ledger is not a back-office artefact. It is an underwriting input. Capacity providers that read it carefully will be better placed to distinguish friction from misconduct, and to price the difference.