Delegated authority should follow data, not volume

I think an insurer should grant binding authority only to partners who give it timely, risk-level information about what they have bound, and should pay commission in a way that rewards that information more than it rewards volume. That sounds obvious when written down. It is not how delegated authority has worked across much of the European commercial market, and the gap between the principle and the practice has widened as broker consolidation has concentrated distribution power.

Delegated authority is a valuable model. Managing general agents and brokers with binding authority reach customers that a large insurer serves poorly: specialist trades, small businesses in particular regions, niches where local knowledge matters more than balance sheet size. Many of them underwrite well. The problem is not delegation as such. It is delegation in which the insurer carries the loss but cannot see the risk until it becomes a claim.

The pen belongs to whoever carries the loss

When an insurer grants binding authority, it lends its balance sheet, its licence and its name to someone else's underwriting judgement. The partner chooses the risks, sets the terms within agreed limits and often handles the customer relationship. The insurer pays the claims and holds the capital. In a well-run arrangement, the insurer knows what has been bound shortly after binding and can monitor quality, accumulation and compliance with its appetite.

In a poorly run arrangement, the insurer receives summary information months later, sometimes as a spreadsheet of premium totals with incomplete addresses. It cannot tell how much flood exposure it has in a particular river catchment, whether the partner is writing outside agreed limits, or whether pricing has drifted since the last audit. It learns about the risk when the loss arrives.

At Talvermark in 2022, we wrote €1.1 billion of premium through 64 binding partners. The loss ratio on that business was 71%, against 61% on business we underwrote directly. Risk-level data often arrived late or incomplete. In some property schemes, we could not reliably tell which buildings we insured until after a claim. That made our catastrophe accumulation reporting less complete than our board papers suggested, and I count that as my responsibility, since I had been chief underwriting officer for part of that period.

There is a regulatory dimension as well. Under the Insurance Distribution Directive, an insurer that manufactures a product remains responsible for its product oversight and governance, including whether the product reaches the customers it was designed for. That responsibility does not transfer to a partner with a binding authority agreement. An insurer that cannot see its delegated business cannot credibly explain to its supervisor how it meets that obligation.

Pay for information, not for premium

Most delegated commission structures reward volume. Higher premium earns a higher commission tier, and profit commissions, where they exist, are often calculated over periods long enough to let a poor year be offset by a good one. Under that structure, a partner has every reason to grow and little reason to send detailed data quickly. Data costs money to collect and clean, and nobody pays for it.

We changed the conditions over 2023 to 2025. Binding authority now requires risk-level data within thirty days of binding, in a defined format, including full addresses and every field we need for accumulation. For property risks above a set size, partners use our pricing rather than their own. Commission no longer rises with volume alone. It is linked to data timeliness and quality, with a component based on loss experience over a defined period.

The effect was significant. The share of risk data received within thirty days rose from 54% to 93%. Our accumulation reporting now covers the delegated book properly. The loss ratio on delegated business improved from 71% to 63% by the end of 2025, though some of that improvement came from the partners who left rather than from changed behaviour among those who stayed. The number of binding partners fell from 64 to 38, and delegated premium from €1.1 billion to €860 million.

We made mistakes in the transition. I announced the new conditions before we had built the shared data portal partners needed to meet them, and several capable small agencies nearly left because our systems lagged behind our policy. We extended their transition to twelve months and built the portal with them. Some of those agencies are now among our fastest-growing partners.

What a consolidated broker is owed

Broker consolidation has changed the bargaining position on both sides. Large consolidated brokers, many of them backed by private equity, now control significant shares of commercial placement in several European markets. They can move substantial volume between insurers, and they increasingly ask for higher commissions, profit shares or placement fees in return for continued support.

I do not think consolidation is bad for customers in itself. Larger brokers can invest in expertise and technology that small firms cannot. But an insurer should be clear about what it owes a broker and what it does not. It owes fair terms, reliable service, prompt claims payment and honest explanations of price. It does not owe higher commission in return for volume alone, and it should not grant binding authority to a partner unwilling to share what the insurer needs to oversee that authority.

In 2024, one of the largest consolidated brokers in our markets asked for a higher commission and a profit share to continue placing business with us under delegated authority. The business was worth €140 million of annual premium. Our head of broker distribution warned, correctly, that competitors would accept the terms. We declined, offered the same data-linked structure we offered everyone else, and let the business go. I do not know whether the insurer that took it will be pleased in five years. I do know that we could not have explained the arrangement to our supervisor in terms of product oversight.

My suggestion to other insurance boards is practical. Ask for a list of every binding authority arrangement, the premium written under it, the time it takes to receive risk-level data and the loss ratio over five years. If the board cannot see that list in one page, the insurer does not know what it has delegated. If the arrangements paying the highest commission are also the ones providing the least information, the incentives are pointing in the wrong direction.