When an insurer retains more risk, its expense ratio needs a different reading

Kinsale’s October 22, 2026 release put its third-quarter expense ratio at 22.0%, versus 21.0% a year earlier, attributing the rise primarily to lower ceding commissions following higher retention. Company release

That explanation opens a question more useful than whether the ratio moved in the “right” direction: how much of a reported cost change represents running the insurance operation, and how much represents changing the amount of risk it owns? Those decisions intersect in the accounts, but they should not be evaluated as if they were interchangeable.

This is an analytical framework, not a calculation of Kinsale’s optimal reinsurance purchase. A quarterly ratio cannot supply the contract-level exposures and loss distributions needed for that calculation. It can, however, tell readers where a comparison needs to become more demanding.

First hold the operating work constant

Kinsale’s 2025 accounting note says ceding commissions are capitalized and amortized as a reduction of underwriting, acquisition and insurance expenses. Annual report, reinsurance accounting

Consider a hypothetical insurer comparing two ways of financing exactly the same portfolio. Its underwriters assess the same applications, brokers place the same policies, and staff perform the same administrative work. In one arrangement it transfers more business to a reinsurer; in the other it keeps more. The two arrangements can produce different net expenses even though no employee has become more or less productive.

The commission is part of the economics of the transfer. Losing some of that offset need not mean that the underlying acquisition process became more expensive. Conversely, a larger offset is not evidence that the insurer developed a better operating system. Calling either movement “efficiency” before separating the transfer would give an accounting presentation credit for a business improvement it has not established.

The first comparison should therefore start before reinsurance offsets: what did it cost to originate and administer a comparable portfolio? Even that needs care. Smaller policies can require nearly as much work as larger ones; a more complex account may justify a higher service cost. Premium dollars alone are not a complete measure of workload. A credible operating comparison would hold product mix, service demands and the period of expense recognition reasonably constant, rather than reward whichever denominator happens to grow faster.

Reinsurance also changes the premium base remaining with the insurer. A reader cannot adjust only the expense numerator and leave the rest of the comparison untouched. The direction of a net ratio is not mechanically determined by the direction of commissions. Both sides must be reconstructed on a consistent basis, with any other revenue included according to the actual reporting definition.

Then compare ownership of the same losses

Once operating work is separated, the economically harder question appears: what is being retained in exchange for the foregone transfer?

In the hypothetical portfolio, retaining an additional share of every policy is different from accepting a larger loss before protection starts. So is reducing the maximum amount recoverable in an extreme event. These choices can leave similar amounts of premium on the balance sheet while producing very different outcomes in a bad year. “More retention” is a starting description, not a sufficient risk measurement.

The useful counterfactual applies the old and new arrangements to the same underlying portfolio and the same set of loss scenarios. Ordinary claims show the expected earnings trade-off. A difficult underwriting year tests how quickly that advantage disappears. A severe but plausible scenario tests the amount of capital and cash needed to keep paying claims and writing sound business. These are proposed comparisons, not statements about Kinsale’s undisclosed treaty structure or future losses.

This approach also prevents hindsight from taking over. A quiet quarter makes protection that was not used look costly. A damaging quarter makes almost any prior purchase look wise. Neither observation alone establishes whether the original decision was well priced. The relevant assessment uses the information and range of outcomes available when the contract was chosen, rather than judging the decision solely by the realized weather of one reporting period.

Capital is part of the price

NAIC describes risk-based capital as a statutory minimum linked to size and risk, not necessarily all the capital an insurer needs for its objectives. NAIC explanation

That distinction matters when comparing retained earnings opportunities. The insurer may keep additional funds for a time, but shareholders cannot assume that all those funds are available for distribution. Some must remain able to absorb adverse outcomes. The opportunity cost is the alternative use of that supporting capital, not merely the cash price of the reinsurance contract.

A fair comparison would therefore consider incremental expected underwriting profit, associated investment earnings, and the additional capital and liquidity required under the same risk assumptions. Investment income belongs in the analysis, but adding it without the corresponding obligations would favor retention by construction. Timing matters too: cash received early and claims paid later can help earnings while still creating an obligation to meet payments under unfavorable conditions.

There are competing, reasonable interpretations. An insurer with stronger risk information and a more resilient balance sheet may rationally retain business that previously needed external support. Alternatively, expensive reinsurance may make retention the less costly available choice without making the underlying risk any better. A favorable expected margin can coexist with a wider range of possible outcomes. Public headline ratios do not distinguish those explanations by themselves.

What the comparison can actually establish

The practical reading is a bridge, not a verdict. Begin with comparable gross operating costs. Identify the premium and commission effects of the transfer. Compare losses under a common retention structure. Then ask what capital would be committed to support each outcome and for how long. If the necessary disclosures are missing, that part of the result remains unknown; it should not be filled with a confident efficiency claim.

A higher expense ratio can accompany a sensible decision to own more risk. It can also accompany genuine cost pressure. Separating those possibilities protects the analysis in both directions: it avoids penalizing a deliberate financing choice as operational deterioration, and avoids treating extra risk-bearing as a costless improvement. The useful question is whether the complete economics improved at a comparable level of risk, not whether one reported percentage became smaller.