The insurance value chain has been pulled apart into specialists. The cost of running it that way is higher than most people think.
The disassembled chain consumes 35 cents of every premium dollar before a single claim is paid. The economics hold only if loss ratios stay well below the industry average.
The insurance value chain, disassembled.
The insurance value chain used to be vertically integrated. A carrier underwrote the risk, issued the paper, held the capital, managed claims, and often even ran distribution, all under one roof. Over the past decade, this value chain has been pulled apart into separate specialists, especially in complex commercial and specialty lines. A broker originates. A wholesaler routes. An MGA underwrites. A fronting carrier provides paper. A reinsurer provides the capital. A TPA handles the claims. Each specialist says it does its piece better than the integrated carrier did.
That may be true. But every specialist also takes a profit margin, and these margins compound. This disassembled chain only works for the most profitable programs.
Most of the $128 billion MGA market is concentrated in complex and specialty risk: cyber, construction, environmental, professional liability, specialty casualty, catastrophe-exposed property, and high-risk GL, most of it in E&S. Standard risk doesn’t need this elaborate modular chain. The premium flows through direct carriers and increasingly embedded distribution. The fee structure I lay out below applies to the complex end of the market, where most of the growth has happened in the past decade, and where the economics now get tight.
This is not an argument against MGAs. Building specialty underwriting inside a carrier is expensive, slow, and often fails. Renting it through an MGA is faster, and that tradeoff is rational. My question is whether the economics sustain themselves at scale, especially for heavily intermediated programs with long-tail risk.
The fee stack
The reinsurer finances the entire operating chain through a ceding commission, typically 30-40% of premium. That pays for distribution, underwriting, administration, claims, compliance, and the fronting carrier.
The MGA’s own economics are thin. After paying producers and covering operating costs, its net margin is typically 5-10% of premium. A $5 million book barely covers one experienced underwriter. The MGA’s real upside is contingent profit commission, typically 10-30% of underwriting profit, but the formula is complex and takes time to materialize.
Let’s take a typical specialty casualty program. $100 of premium enters the chain. The retail broker takes $12, the wholesale broker $8, the MGA retains $7 net, the fronting carrier takes $5, and premium tax, compliance, and claims handling cost $3. The chain takes $35 before a single claim is paid.
The reinsurer also has to account for its own internal costs (actuarial, reserving, overhead) - let’s assume another $8. Together, that’s $43. That leaves only $57 for losses.
The reinsurer’s margin depends entirely on the loss ratio:
- With a 40% loss ratio, the reinsurer keeps $17. The model works.
- With 50%, the reinsurer keeps $7. Still works, but getting thin.
- With 55%, the reinsurer keeps $2. That's not adequate, especially for long-tail risk.
Now compare that with an integrated carrier writing the same risk. The carrier still pays a retail broker, premium tax, and claims handling. But it might skip the wholesale layer, there certainly is no fronting fee, no external MGA profit margin, and no reinsurer overhead on top. Its all-in expense load is closer to $27, not $43.
At a 55% loss ratio, the integrated carrier keeps $18 on every $100. The disassembled chain keeps only $2. That’s the cost of disassembly.
Kinsale Capital, an E&S specialist insurer, writes $2 billion in premium. It doesn’t delegate underwriting to MGAs. It runs a 55% loss ratio and posts a 76% combined ratio. Same loss ratio as in our scenario, different model, very different result.
What return does a reinsurer actually need to commit capacity to an MGA program, especially with long-tail risks? For specialty casualty or professional liability programs, losses develop over 5-10 years. Initial gross accident-year loss ratios have developed adversely over the past seven years across the fronting sector (Conning, August 2026). A 2% profit margin is not adequate for that kind of risk. At a 50% loss ratio, the reinsurer’s profit margin is 7%. One adverse development year can wipe that out. The disassembled model needs loss ratios to stay below 50% to prosper. In specialty casualty, that is not an easy target. It only works for the share of risks that genuinely outperform the market. As the MGA market keeps growing from 10% to 20% of premium, a larger share of the industry is required to beat the average. That gets harder, not easier.
Premium growth has been the headline. The fee stack above shows how little of it reaches the reinsurer.
Why nobody coordinates the value chain
Every player in this disassembled chain optimizes its own piece. That observation isn’t new (I wrote about the accountability gaps in the MGA ecosystem in an earlier article). What I didn’t examine then is why the coordination gap persists. The answer lies in the fee structure.
Let’s look at the players and their incentives. The MGA makes the underwriting call and often handles claims. That means the MGA coordinates the risk decision - the most important decision in the value chain. But the MGA doesn’t control its paper or capital. MGA platforms coordinate the MGAs they own and provide shared infrastructure and carrier relationships. Their coordination stops at their portfolio boundary. Fronting carriers coordinate regulatory infrastructure. Reinsurance brokers coordinate up to the placement. Reinsurers coordinate capital allocation.
In sum, nobody coordinates end-to-end across the full chain because nobody’s fee justifies it. The broker earns its commission whether the MGA’s reserves develop well or poorly. The MGA earns its base commission whether the book performs or not. The front earns its fee whether the reinsurer is satisfied with the data or not. The reinsurer earns its return whether the program’s total economics are sustainable or not. The coordination gap is economic, not organizational.
Where the fragility shows
The challenge is that the incentives don’t just fail to align. They pull in opposite directions.
The MGA’s base commission scales with premium. The contingent profit commission is deferred and can take years to pay out. For an MGA, growing premium is the rational short-term move. The reinsurer expects low loss ratios but also wants to hit its own premium target – objectives that can conflict.
Reserves for long-tail risks expose this misalignment. Three parties touch the same reserve estimate with different economic consequences. The MGA sees the losses first and in the most detail, but its base commission doesn’t depend on reserve accuracy. The fronting carrier retains regulatory incentives for reserve discipline, but its direct exposure is small. The reinsurer bears the most risk but sees the data last and with the least granularity. When the party with the most information has the least incentive for accuracy, optimism tends to persist. The result: initial gross accident-year loss ratios across the fronting sector have developed adversely in each of the past seven years (Conning, August 2026). That is primarily a long-tail phenomenon, exactly where the MGA market has been growing fastest.
When loss ratios stay below 50%, every participant gets paid handsomely and the coordination gaps remain invisible. But when loss ratios climb toward 55%, the reinsurer’s margin thins to where one bad year erases it.
An MGA’s core value proposition is isolating the profitable niches within broader classes - picking better risks than the carrier’s book-level underwriting can identify. But if the MGA systematically pulls the best risks from the pool, the carrier’s retained book deteriorates. The MGA takes the 40-50% loss ratio risks. The carrier keeps the rest, and its combined ratio worsens, pushing it to cede more.
The carrier is paying the MGA to make its own book worse.
And the pool of above-average risks is finite. As the market keeps growing toward $150 billion or even $200 billion, the marginal risks must get worse. PE investors are valuing MGAs on EBITDA multiples that look strong precisely because the MGA doesn’t bear the risk. The fee stack math shows who does, and what’s left. At some point, the MGA market’s aggregate loss ratio converges toward the industry average, and the fee stack that worked at $100 billion stops working at $200 billion. Nobody in the chain is paid to find that ceiling. The adverse development data may be telling us that parts of the market have already crossed it.
What happens when the market turns
At a full-stack carrier, one management team owns underwriting, claims, reserves, and capital allocation. That feedback loop isn’t perfect either (anyone who has worked inside a carrier knows the internal fights). But when loss ratios deteriorate, one team decides what to do: raise rates, tighten guidelines, adjust reserves, redeploy capital. The response is coordinated because the economics sit inside a single P&L.
The disassembled chain has no equivalent coordination mechanism. When the loss ratio on a fronted program moves from 45% to 55%, no single entity decides what to change. The MGA wants to keep writing. The front’s retention is small enough and doesn’t bother. The reinsurer can non-renew, but only at the next treaty renewal - months later. And a different reinsurer, hungry for new premium, may step in. Capital filling a vacuum is not coordination.
One logical response to this pressure is re-integration.
The evidence is already visible. Hagerty moved to retaining 100% of its US book in January 2026, keeping the paper arrangement with Essentia but taking back all the economics. Munich Re, which had provided capacity to At-Bay since the program’s early years, agreed in August 2026 to acquire the company outright. From the carrier side, Allianz transferred its entire standalone commercial cyber portfolio to Coalition, an MGA that had started on carrier paper and acquired its own admitted carrier in 2022. Re-integration happens from both directions.
Successful MGAs don’t wait for a coordinator - they move toward greater integration. Large reinsurers don’t coordinate MGAs - they acquire them. Whether driven by success or by stress, the trajectory bends back toward the vertical integration the disassembled chain was built to replace.
The MGAs that can’t vertically integrate - too small, too early, or not profitable enough to attract their own capital - need a coordinator the most and are least likely to get one. They depend on the platform model. Whether the major MGA platforms can deliver real coordination under stress, not just shared services in a benign market, is still untested.
The questions that matter: What is the breakeven loss ratio on each program after the full fee stack, and how much room does that leave before the reinsurer walks away? Who has real-time visibility into loss development across the value chain - not quarterly bordereaux, not annual stewardship meetings, but current data from someone whose compensation depends on getting it right? And at what premium volume does building the capability internally cost less than renting it through six intermediaries? Most boards are not asking these. The fee stack makes them urgent.
In their 2026 MGA study, Conning asks whether MGAs are “reconfiguring the insurance value chain.” They are. Whether the reconfigured chain can sustain itself, with a fee stack that needs exceptional loss ratios, without a coordinator, when the market eventually turns, is the part the celebration hasn’t reached yet.
