Insurance Distribution: Underwriting Through a Softening MarketBy John Yeager, Managing Director, WhiteHorse Capital

September 29, 2026 – Insurance distribution platforms financed between 2021 and 2024 grew into a hard market. That market has turned, though not evenly, and the unevenness is the useful part: it lets a lender read exposure off revenue mix long before it reaches reported earnings. We cover the sector closely and expect the opportunity set to widen as pricing normalizes. We lend to commission- and fee-based distributors across the middle market, usually as agent and often as sole lender, with the hold size to carry a scaled or growing platform and its acquisition program through a full cycle. Our coverage runs through more than 70 investment professionals in 12 North American offices.

An Uneven Turn

Commercial premiums fell 2.0% across all accounts in the second quarter, the second straight quarterly decline and the first back-to-back declines since 2017.¹ The headline is the least interesting number in the survey. Commercial property fell 6.3%, a fourth consecutive decrease, while umbrella rose 5.3% in a 35th consecutive quarter of increases.¹ Marsh has U.S. property down 13% and U.S. casualty up 7% over the same period.² The two lines are moving for different reasons. Property has absorbed several years of new reinsurance and catastrophe bond capacity. Casualty is still pricing against social inflation and adverse development on prior accident years.

Rate is not the whole story, and the distinction matters before anyone draws conclusions from it. Premium is rate multiplied by an exposure base, and payroll, revenue, vehicle counts and insured values have kept growing, which offsets part of the decline. A good deal of larger-account revenue is fee-based rather than commission-based and does not move with rate at all. What rate did provide was operating leverage. Strip it out and our own estimate is that organic growth at the platforms financed in 2022 and 2023 ran closer to mid single digits than to the low double digits reported.

Two things temper the picture. Softening has been size-graded, with large accounts down 3.7%, medium down 1.9% and small down 0.5%,¹ so a book weighted to small and mid-sized accounts carries less of this than the aggregate suggests. And the gap between property and casualty is closing. Casualty decelerated from 9% to 7%,² and umbrella at 5.3% is half the 11.5% it printed a year ago.¹

More Transactions, Less Premium

The excess and surplus market shows the margin mechanic most clearly. Surplus lines premium across the 15 stamping-office states grew 2.8% in the first half to $47.6bn, against 13.2% a year earlier. Item filings grew 16.9% to 4.28 million over the same period. Property premium fell 13.7% to $13.6bn. In Florida, premium fell 5.6% while filings rose 14.4%.³

Commission scales with premium. Cost scales with transactions. Seventeen percent more filings against three percent more premium is margin compression, whatever else it resembles. Ryan Specialty’s second-quarter filing puts the pullback in its property book down to two things: continued rate decline, and retailers placing coverage directly.⁴ The second is structural and will not reverse when rate does. Surplus lines licensing and diligent-search requirements cap how much of that flow can actually move, but the direction is set, and this is the channel automated placement reaches first.

Reported results already show it. Organic growth across the listed brokers in the second quarter ran from negative 0.7% at Brown & Brown, or positive 0.7% counting contingents, to 6.0% at Arthur J. Gallagher, with Aon and Marsh & McLennan at 5.0% and Ryan Specialty at 6.3%, down from 11.8% in the first quarter.⁵ Brown & Brown is the cleanest illustration: total revenue up 30.4% on acquisition, organic down, adjusted EBITDAC margin off 100 basis points.⁵ Acquisition can carry reported revenue for a long time. It cannot carry organic growth or margin.

Same Market, Different Outcomes

Agency M&A is at a decade low. There were 292 announced transactions in the first half of 2026, down 15% year over year, 24% below the five-year average, and the slowest first half since 2016. The trailing twelve-month count of 646 is the lowest since the first quarter of 2019, a fourth consecutive year of decline.⁶

Part of that is straightforward. Financing costs and a slower exit environment have made buyers more disciplined, and after a decade of consolidation there are fewer independent agencies of scale left to buy. Neither explains the dispersion inside the cohort, which is the part worth looking at. Hub International, Acrisure, Keystone Agency Partners and Highstreet Insurance Partners each cut acquisition pace by more than half. BroadStreet Partners went the other way, closing 67 transactions on a trailing twelve-month basis against its own five-year average of 60, with Inszone completing 33 in the half. Across eight quarters, 143 distinct buyers were active; 43% did one deal and only 53 did four or more. Valuations reportedly held for larger, well-run agencies and softened everywhere else.⁶

Funding explains some of it. BroadStreet repriced its term loan 25 basis points tighter in January and raised a $350 million add-on, upsized by $200 million, reported as intended to pre-fund tuck-ins.⁷ It then bought above its own historical pace into the slowest market in a decade. It does not explain all of it. Highstreet raised an incremental $550 million delayed draw term loan in August 2025 to support its acquisition and innovation strategies⁸ and still slowed by more than half. Something other than access to capital sets cadence, and in our experience it is integration bandwidth, seller price expectations, or a sponsor deciding this is not the moment.

A Second Axis of Risk

Rate is not the only thing repricing this market. On February 9, two AI quoting applications went live inside ChatGPT and the MarshBerry Broker Composite Index fell 8.9% in a single session, with Willis Towers Watson off 12% on its worst day since 2008.⁹ Analysts called the reaction overdone, correctly noting the applications addressed personal lines, and distributor valuations had mostly recovered by spring.¹⁰

The price recovery settles less than it looks like it does. BofA Global Research later sized at least $15 billion of U.S. independent agency commissions and broker fees as exposed to some disintermediation, out of a pool above $100 billion in 2025.¹⁰ That exposure sits in low-complexity business: personal lines and small commercial across six carriers, with national accounts, global specialty and middle market left out of the count. Ryan Specialty is treated as comparatively insulated because of the complexity of what it places.¹⁰

Read that against the rest of this note and an awkward pattern emerges. Rate exposure and AI exposure run in roughly opposite directions. The small-account books least troubled by softening premium are the ones most open to automated placement, while the complex specialty risk most exposed to rate is the hardest to substitute. Rotating away from rate sensitivity toward small-account revenue is not de-risking. It is swapping one risk for another.

Displacement is also not the only channel. BofA expects AI to de-mystify markets and cut pricing power even where the intermediary survives.¹⁰ Cutting the other way, agentic workflows that absorb servicing work create capacity exactly where filings are outrunning premium.⁹

Where We Have Been Active, and What We Underwrite

We have watched this sector closely through the hard market that followed COVID and been selective about where we leaned in. Where we did not participate, deals cleared at leverage or pricing we were not prepared to match, against reported organic growth that leaned heavily on premium rate and forecasts that assumed rate would hold or keep rising.

That is why we expect to be more active from here. As pricing normalizes and sponsors reset expectations, more of this paper should come within range on terms we are comfortable underwriting. The businesses we find most durable are the ones whose revenue is not levered to premium rate alone.

How We Are Thinking About Underwriting

Most of this can be read off revenue mix long before it reaches EBITDA. When we look at insurance distribution credits, we think about:

• Organic Growth Net of Rate: Whether management can bridge reported organic into premium rate and true new business. Not being able to is a diligence finding, not a presentation gap.
• Pro Forma Acquired EBITDA: Books capitalized at run-rate with full synergy credit tend to shrink after close in softening lines, which understates closing leverage and lets it drift up.
• Carrier Capacity: AM Best has flagged fronting carriers for potential adverse development in accident years 2021 through 2024.¹¹ For a concentrated MGA, losing the pen is a going-concern event rather than a margin event, so we look hard at how many carriers stand behind the book and how the authority is held contractually.
• Complexity of the Placement: Whether the business places risk that needs judgment, market access and structuring, or risk that can be quoted off structured data. We treat complexity as a durable feature of the credit, not a technology question.
• Contingent and Profit Commissions: The highest-margin and least durable line in an MGA’s revenue. They track loss ratio and volume rather than rate, so they need their own forecast rather than a scaled version of the commission line.

None of these are gating criteria. A book weighted to rate-sensitive lines is not uninvestable. It may just warrant lower leverage, tighter structure, or different pricing.

Figure 3: Rate and AI Exposure by Borrower Profile

Profile	Revenue basis	Rate sensitivity	AI exposure	WhiteHorse posture
Property-weighted MGA or program	Commission and contingents on softening lines	High	Low; delegated authority and complex risk	Selective; lower leverage and tighter structure
E&S wholesale	Commission, flow-levered	High	Low to moderate; complexity and surplus lines licensing insulate placement, though flow is already moving	Selective; we underwrite the flow assumption
Casualty program	Commission on firming lines	Low near term	Low	Constructive, with focus on reserve development
Retail benefits broker	Fee and commission, headcount-levered	Low	Moderate; small-group and voluntary lines are the most substitutable	Core focus
TPA, claims, and fee-based services	Fee per claim or per life	Negligible	Moderate; automation compresses cost per claim and the fee with it	Core focus

WhiteHorse Capital analysis.

Focused on Revenue Quality

Property-weighted commission books and fee-based service businesses are different credits, and for two years the market has largely priced them the same. We do not pretend to know where premium rate settles, or when. We do think the line between revenue that depends on rate and revenue that does not will matter more over the next several years than it has over the past five. That is where our diligence time goes, and where we expect the better partnerships to be found.