We Didn't Start The Fire
Library  /  Insights

The Playbook - The Portfolio, not the Pipeline

How the Broker-to-MGA Bridge Captures the Largest Unclaimed Pool of Specialty Premium — and the Four Rungs That Decide Who Owns It

Arijit Sarkar10 September 2026 · 3 min read

Decide Who Owns It

Every large retail broker carries a book it would rather not discuss: small-commercial E&S — thousands of policies, average premium in the low thousands, serviced by handlers whose time is priced against programmes a hundred times larger. Modelled on a representative book — 15,000 policies, $108m GWP, $7,200 average — retail commission at 12% earns $12.96m; fully loaded cost to serve at $1,050 a policy consumes $15.75m. The book loses $2.79m a year — $186 on every policy. Yet it cannot be released: small E&S policies sit alongside the property programme, the casualty tower, the management liability. Handing one back puts the anchor relationship in play. A book that cannot be fixed and cannot be released is not a problem. It is the largest pool of already-written premium in the market.

The Blindspot: Premium Rented, Never Owned

MGAs have spent a decade competing for submissions on appetite, speed and rate — growth rented one quote at a time — while the largest available block of premium sits inside broker books the brokers themselves cannot make work. This is not a pipeline. It is a portfolio: already written, already renewing, with a counterparty who has a financial reason to move. The deal is won on four dimensions — servicing capability, genuine underwriting share, duty of care, and broker economics. Price is the last of the four, not the first.

The Same Blindspot, Three Times Over

Consolidation backlog. Buying is fast; absorbing is slow. A decade of aggregation left estates of 30+ unabsorbed brokerages — 120,000 policies, ~£10m a year running 30 of something you need 1 of. Finishing the job costs ~£6m; payback under seven months.

It sits undone because the only people who can migrate are carrying revenue: pulling 40 handlers off client work forfeits ~£8m of commission. The constraint is capacity, not capital.

Broker IQ. FCA value-measures data puts renewal commission at £36–50 per account — less than a team lunch against a gap review's hours. So the hunt never happens: cross-sell sits near 15%, three in four SMEs are underinsured. An external intelligence layer ingests every AMS as-is, benchmarks each SME against its peer cohort, and drops pre-renewal advisory packs into the handler's workflow — lifting cross-sell revenue 20–30% inside the £36–50 budget.

Where The Premium goes: Mirrored US–UK Routing

From the broker's distressed book, 70–80% of risks fit existing MGA programmes and are underwritten directly; 20–30% require new MGA vehicles purpose-built around the book's residual niches; the remainder is pushed to Lloyd's and the London Market (US residual) — mirrored in the UK, where the open market takes the lead, the coverholder market absorbs the delegated share, and the US E&S market takes what neither can hold. US E&S direct premiums surpassed $100bn in 2025; delegated authority now carries ~40% of Lloyd's £58bn GWP; the US MGA market reached ~$128bn. The rails exist. The books do not.

Four Rungs, Not A Spectrum - and the jump that decides Everything

Rung 1 — Service only / Migration factory. Fee income under the broker's brand. Fastest to stand up, lowest resistance — a contract, not an asset.

Rung 2 — Service + place / Migrate and run the queue. Bind on own paper where appetite fits; discount the fee against commission. Alignment that shows up in the numbers.

Rung 3 — Full transfer / Own the back office. Broker of record moves; renewal economics in perpetuity — the only rung that creates enterprise value, and the only one where the duty of care moves.

Rung 4 — Hybrid by segment. Where most arrangements land, and the most commonly mispriced: a third operating model, not a midpoint.

Three hard truths.

First, everything hinges on the share of the book you would genuinely underwrite — measured on the broker's in-force book as a distribution, never an average; below break-even, service it, above it, own it.

Second, Rung 3 always earns less than Rung 2 in year one — licences, onshore advisory, PI cover and client consent are paid up front while renewal income builds.

You do Rung 3 because you end up owning the business. Third, liability does not creep up rung by rung — it jumps in one step, the moment you become the client's broker. Price the jump; insure it; be the supplier, not the defendant.

The Move

Blueprint all four rungs. Enter at Rung 2 — supplying the capacity the broker cannot free up — with a contractually pre-agreed escalator to Rung 3. Build the segmentation gate (underwrite / new vehicle / Lloyd's–London; migrate / absorb / retire) even if the pilot is single-model: it is close to non-retrofittable afterwards. Layer Broker IQ across every acquired book from day one.

Follow on LinkedInDiscuss this with ArijitSee the working demos

More from the library