The pitch to MGA founders is similar: preserve your team, brand, culture, and carrier relationships – and don’t touch the underwriting. This strategy works as long as the program performs.
Doxa, K2, Novacore, Amynta, Bishop Street, Baldwin/MSI, ANV, and others run variations of this playbook. These platforms collectively account for over 150 programs and tens of billions in premium. The question nobody can answer cleanly: what is the platform itself worth?
Photo: Guillaume Bolduc / Unsplash
Structural limitations
“Buy and preserve” is what makes founders sell. An underwriting team that built a $100 million book doesn’t want to be absorbed into somebody else’s operating model. The promise of autonomy – keep your brand, your team, your carrier relationships, and run your program independently – closes the deal at 12-15x EBITDA (MarshBerry).
However, this autonomy also creates a structural tension. Preserving specialist underwriting authority is not the problem – several platforms describe centralized underwriting governance and portfolio-wide oversight. But do those capabilities change decisions or consolidate reporting? If loss patterns in one program don’t inform pricing in another, and deteriorating performance in one line doesn’t trigger a review across the portfolio, the underwriting stays behind the program door regardless of what the org chart says.
Social inflation hits casualty programs. Natural catastrophes hit property. Litigation trends hit D&O. Each line deteriorates on its own schedule, and no one coordinates the response. A carrier CUO with the same diversification sees the trends and tightens guidelines before losses materialize. In an MGA platform, each program has to find the problem on its own.
And the autonomy has an expiration date. Most earn-outs tie the founder for 3-5 years. After that, the program leader can leave. I’ve seen what happens when the key person walks: the book is still there, the infrastructure is still there, but the judgment and the relationships that built the underwriting are gone. What’s left is a hollowed-out vessel – a platform servicing a book it didn’t build and doesn’t fully understand.
Conglomerate discount
In 2022, Carlyle acquired NSM Insurance Group for $1.8 billion – programs across specialty P&C, accident & health, and reinsurance. Three years later, Carlyle split NSM into two entities: the commercial programs were sold to New Mountain Capital and rebranded as Novacore, the niche specialty businesses were retained as Ignyte, and the NSM brand was retired. Six months after the split, Ignyte sold its collector vehicle division to Philadelphia Insurance Companies for $615 million. The platform is being monetized as parts.
$1.8B
Ignyte retained, NSM retired
vehicles to PHLY — $615M
NSM’s separation raises a question every MGA platform should be able to answer: which businesses benefit from common ownership, and what capabilities make that ownership valuable?
A back of the envelope illustration: Assume an MGA platform with 20 programs of similar size. A buyer looks at the programs individually. Fourteen are performing well and get valued at 15x EBITDA. Three have deteriorating loss ratios and get valued at 7x. Two have earn-outs expiring within 12 months with no clear successor – without the team, they only get valued at 4x. One is in a capacity dispute and gets excluded from the bid.
The gross sum-of-parts comes to roughly 12x across the portfolio. Then the buyer subtracts the platform management layer and applies a conglomerate discount because the programs operate independently and could be owned separately. Net to the buyer: 10-11x. The sponsor assembled the portfolio at 12x over seven years.
| Programs | Count | Multiple |
|---|---|---|
| Performing well | 14 | 15x |
| Deteriorating loss ratios | 3 | 7x |
| Earn-out expiring, no successor | 2 | 4x |
| Capacity dispute | 1 | Excluded |
| Gross sum-of-parts | ~12x | |
| Less: platform overhead, conglomerate discount | ||
| Net to buyer | 10–11x |
Of course, this is simplified, not a valuation model. But it shows how program-level diligence can disaggregate a portfolio story that looks compelling in aggregate.
For most PE-backed platforms assembled over the past decade, the arithmetic doesn’t work. Sponsors built portfolios at 12x. At exit, the gross sum-of-parts comes in near the same number – meaning the platform created no portfolio premium. Subtract the management layer and cost of capital over a seven-year hold, and the return on the platform itself is at or near zero. The programs inside may have performed. The wrapper didn’t earn its keep.
Shared infrastructure has value, but it’s a cost center, not a source of underwriting returns. The fee stack already consumes 35 cents of every premium dollar before a claim is paid (The Cost of Disassembly). The platform’s management layer sits on top of that. A buyer will subtract the overhead, not add a premium for it.
When the capacity providers start asking questions
Every MGA depends on borrowed capacity – that’s the Achilles heel of the model.
If a capacity provider non-renews a program inside a PE platform, the signal doesn’t stop at the program boundary. Other capacity providers look at their own exposure to the same platform. The diversification that investors treat as protection also works in reverse: programs are associated through the platform, and a problem in one raises questions about the rest.
I’ve seen a smaller version of this firsthand. A deductible buyback program lost its capacity not because of underperformance, but because the capacity provider reorganized and changed its appetite. The underwriting was sound, but this single dependency was fatal.
I’ve been on the MGA side. When a carrier starts asking questions, you answer them. When those questions are about someone else’s losses on the same paper, you have a problem you didn’t create.
Three questions for anyone evaluating an MGA platform
MGA platforms have been valuable, especially for smaller programs. Shared infrastructure, carrier access, growth capital, compliance support – they solve real problems that would otherwise pull time and attention away from writing business.
But does the platform create underwriting intelligence across programs, or just house them? Most PE-backed aggregators have not built that.
The underwriting intelligence that makes an insurance company resilient through cycles is built over years, especially in long-tail lines. It requires people who stayed long enough to see their own reserves develop. Preserving an underwriting business is not the same as institutionalizing its capability.
“Buy and preserve” is an honest strategy. It is not a platform.
MGA Economics Series
Previous: The Two-Track MGA Market
