UK & Western Europe · Financial Services M&A

The Toll-Booth, Not the Toll-Road

Off-consensus investment themes in the UK lower mid-market where the alpha in financial services has actually migrated: the DB pension de-risking wave and the MGA layer of insurance distribution

12 min read  ·  Capital-light B2B enablers · Buy-and-build · Hands-on operator mandate

The obvious trades are done.

The retail IFA platform roll-up that worked in 2018 now clears at 8–11× EBITDA with three PE-backed bidders in any competitive process. The commercial insurance broker consolidation is cooling — UK distribution M&A ended 2025 at its quietest fourth quarter since 2016, with PE-backed buyers at multi-year lows in deal share, while the largest platforms are struggling to find exits at the multiples their entry prices require. Xeinadin’s roughly £1 billion sale process, reported at 15–16× EBITDA, has stalled. The spread between what quality businesses were acquired for and what integrated platforms command at exit — the entire engine of the broker and IFA roll-up thesis — has compressed substantially at entry while remaining open only at exit. The arbitrage still exists. It has just migrated one layer down the value chain.

When a sponsor backs a bulk-annuity insurer, the durable, capital-light, fragmented margin pool is not the insurer — it is capital-intensive, regulated to the bone, and unacquirable at any sensible mandate size. The durable margin pool is the scheme administrator who must make every transaction-ready, the data remediation specialist untangling thirty years of GMP records, the professional trustee who governs the process through to wind-up. When a PE house backs an insurance distribution platform, the capital-light toll-booth is not the retail broker — it is the managing general agent who generates commission and proprietary underwriting data without putting a penny of balance sheet at risk.

This is the toll-booth, not the toll-road. The institutions building the toll-road are also the natural exit buyers of the toll-booth — at their multiple, not yours. Two specific expressions of that logic follow.

The UK defined benefit pension de-risking market has been running at extraordinary volume for three consecutive years. In 2024, £47.3 billion of DB liabilities transferred to insurers via bulk purchase annuities, covering 332,500 scheme members. LCP projects 2026 volumes could reach £55 billion — a potential record — with over 150,000 members receiving individual annuity policies that year alone. The H1 2025 transaction count already exceeded H1 2024 by 20%, representing 23% compound annual growth in deal count over the three years to June 2025.

The wave is not slowing. It is broadening.

Three UK bulk annuity insurers were acquired by North American asset managers in 2024 and 2025 — PIC by Athora, Just Group by Brookfield, and L&G partnered with Blackstone. LCP called this a “robust vote of confidence” in the UK market, noting it would provide additional capacity precisely as smaller schemes enter the pipeline. The Association of British Insurers now reports that 54% of UK DB schemes are in surplus on a full buyout basis. A decade ago, that number was a fraction of this figure, and most of those schemes would not have been able to contemplate a transaction at all.

The critical structural shift for this thesis is what is happening at the small end of the market. Insurers that previously only quoted schemes with liabilities above £100 million are now actively pursuing the sub-£100 million tail. XPS Group described this market segment as “buoyant” into 2026, noting particular momentum in small and mid-sized scheme activity. There are estimated to be thousands of UK DB schemes in this bracket, many now approaching transaction-readiness for the first time in their history.

Every one of those schemes must pass through the same mandatory gateway before any insurer will quote them.

The Data Problem Nobody Talks About

The bottleneck in the de-risking market is not insurer appetite or pricing — both remain competitive. The bottleneck is data.

GMP equalisation — the legal requirement to equalize guaranteed minimum pension benefits between men and women for service between 1978 and 1997 — has been grinding through the UK scheme universe for years. For schemes seeking to buy in or buy out, reaching “insurance-ready data” means resolving GMP calculations, completing benefit specification audits, tracing missing members, and connecting to the Pensions Dashboards ecosystem by the statutory 31 October 2026 deadline. LCP noted that “2026 is likely to be another busy year for specialist pensions data and GMP equalisation teams.” Hymans Robertson described data quality as “the queen of all endgames.” Multiple professional bodies have flagged administrator resource constraints as the binding bottleneck on the whole pipeline.

This is structurally non-discretionary spend. A scheme that wants to transact — and is funded to do so — must complete these projects. It cannot defer indefinitely, and it cannot complete them without specialist third-party help. The Pensions Regulator has itself flagged the risk of large TPA exits causing systemic disruption, noting the market is concentrated and that TPA capacity constraints are a material systemic risk.

Concentration at the large end of the TPA market creates white space in the middle. That white space is where a buy-and-build operator builds.

The Three Target Types

The estimated addressable fee pool for the enabling layer — administration, de-risking advisory and data, professional trustee — is order £2.5–4 billion per annum and growing structurally. Estimated 30–60 independent TPAs and boutique pensions consultancies sit in the £2–6 million net profit band, plus a fragmented professional trustee field with a longer tail. The universe comfortably supports a 6–10 acquisition platform.

Exit references frame the arithmetic: XPS Group (FTSE 250 pensions consulting and administration), Isio (advising on £300 billion+ of client assets), and Broadstone (~800 staff across pensions and employee benefits). The XPS listing provides a live public-market comparable. Global brokers — Gallagher, WTW, Aon, Mercer — have established pattern behaviour in this sector and represent the natural trade-sale buyer pool.

The Operator Playbook

This is a process business — which means the operator edge is real and durable. Industrialise manual administration through workflow tooling and member-data automation. Centralise actuarial and technical capability that sub-scale TPAs cannot afford to carry individually. Cross-sell de-risking readiness into the existing administration book: a scheme that uses your administration services is a natural buyer of your GMP remediation and trustee services. Professionalise the trustee proposition across the portfolio.

A hands-on operator out-earns a passive financial sponsor in this market by a wider margin than in almost any other theme. The reason is simple: the value creation is in the process improvement, not the multiple re-rating. The multiple re-rating follows the process improvement. A financial engineer who buys TPAs and holds them without integrating them does not get the strategic exit multiple — they get the sub-scale entry multiple at exit.

The UK retail commercial broker roll-up has had a remarkable decade. It is now the late-cycle expression of a thesis, not the early-cycle one.

Platform commercial brokers are trading at 12–16× EBITDA where platforms are genuinely integrated. PE-backed consolidators that loaded leverage in 2020–2023 are discovering that exits at those multiples require size, technology, and integration discipline that many of them never built. MarshBerry’s February 2026 update noted that “larger PE-owned businesses are having difficulties securing high-multiple exits” and described a growing scepticism of “pile ’em high” M&A strategies that characterised the recent record-breaking years of UK deal activity. UK insurance distribution M&A in 2025 was the quietest since 2016, with PE-backed buyer share at multi-year lows.

The off-consensus trade is not the retail broker. It is the managing general agent.

Why an MGA Is a Structurally Better Asset

An MGA holds delegated underwriting authority from an insurance carrier. It does not hold the risk — that stays with the insurer — but it controls the underwriting decision, owns the broker relationship, and generates commission and fee revenue on every policy it writes. There are over 300 UK-based MGAs in operation (MGAA, 2024), underwriting roughly 10% of the £47 billion in total UK general insurance premiums. Some estimates put the universe at 350 MGAs underwriting more than £20 billion in gross written premium.

The structural case for an MGA over a retail broker is not marginal — it is significant across three dimensions:

Product control and proprietary data. An MGA differentiates on the coverage it writes, not just the price it shops. Proprietary product is a moat a retail broker rarely has. Underwriting data accumulates as the book seasons — a mature MGA with years of loss experience in a niche class has a genuine informational advantage that compounds with time and cannot easily be replicated by a carrier quoting cold.

Capital-light economics. No balance sheet exposure. Revenue is commission and fee income, recurring in nature and high-margin at the operating level. This is a fundamentally better cash flow profile than a sub-scale retail broker competing on price in a competitive commoditised market.

Carrier alignment. Carriers increasingly prefer to use MGAs for specialist classes rather than underwriting directly — it is cheaper and faster to access niche expertise this way than to build it in-house. Inflexion’s head of financial services stated explicitly in March 2026 that “insurers would rather rely on MGAs for specialist expertise than underwrite these risks directly,” framing insurer demand as a structural rather than cyclical driver. The MGAA described MGAs as “taking a growing share of underwriting” with insurer appetite increasing through 2025 and into 2026.

Where the Arbitrage Actually Lives

Platform MGAs are now fully discovered as an asset class. The best-in-class are commanding 18× EBITDA in some transactions — up from 10× five years ago. Deloitte noted that PE competition has “intensified, pushing valuations higher — so much so that it’s now common to see deals closing at double-digit EBITDA multiples.” At that entry point, the financial engineering argument for aggregation is materially weaker.

The arbitrage lives in the single-class, owner-managed MGA. A firm running one or two lines of specialty business, with carrier relationships established, a seasoned book of recurring commission income, and no institutional capital or management ambition to scale independently. There are estimated to be 100–200 such firms across the UK and Western Europe in the £2–6 million net profit band, most of which have never received a formal approach.

Entry for these assets runs at 5–8× EBITDA. A genuinely integrated, multi-class MGA platform — with pooled carrier capacity negotiated at scale, common compliance and delegated-authority governance, unified data and pricing infrastructure, and cross-class distribution — trades in the low-to-mid teens. The integration is the value creation mechanism. A pile of single-class MGAs sharing no infrastructure is not a platform; it is a holding company with a discount to its components, and it will exit like one.

MarshBerry’s analysis of the UK specialty segment (2023 data, the most recent year with full counts) recorded 23 specialty transactions — a record high at 16% of all UK insurance distribution deals. Specialty and MGA targets have been a rising share of UK distribution M&A even as the retail broker market has cooled. The market is not illiquid. It is less competed than retail.

The Cyclical Entry Window

The 2025 rate-softening across several UK commercial lines is a live pressure on MGA EBITDA. When insurers can underwrite directly at competitive pricing, their incentive to maintain delegated relationships can weaken. This is not a theoretical risk — it is currently suppressing organic growth in several specialty segments.

For an operator, this is the entry signal, not a deterrent.

Soft markets compress EBITDA at owner-managed MGAs and make principals willing sellers at reasonable multiples. An operator that builds during the soft cycle — acquiring single-class specialists, integrating them onto shared infrastructure, deepening carrier relationships through scale — and exits when the hard cycle returns captures both the operational improvement and the multiple expansion. MarshBerry’s February 2026 update noted that the sector “still has plenty of capital waiting to be deployed” and that three UK commercial broking consolidators took on new PE capital in 2025, reinforcing exit demand once the cycle turns. By most cycle estimates, the build window is 2026–2028.

The Exit Path

Capstone’s 2025 data places global distribution M&A multiples in the specialty segment above 16× EBITDA on average. The strategic acquirers are global brokers building niche underwriting capability — Aon, Gallagher, Marsh, and their equivalents — who consistently pay for proprietary specialty expertise that they cannot build faster than they can buy. NFP (an Aon company) acquired MGA Bspoke Insurance Group. Allianz Commercial partnered with specialty MGAs explicitly to access niche classes of business. This is a buyer base with both strategic motivation and available capital, operating on a different time horizon than a PE fund.

The Common Thread — and What It Requires

Both themes share a structural logic that is worth making explicit before any capital is committed.

The institutional capital in each value chain is chasing the headline asset: the bulk-annuity insurer, the retail broker platform. The headline asset is either unacquirable at this mandate’s size, or is priced past the point where entry-to-exit arbitrage remains compelling. The enabling layer — the toll-booth that every transaction must pass through — is where the recurring, capital-light, defensible margin sits. It is fragmented, under-institutionalised, and in both cases carrying entry multiples that have not kept pace with what the strategic exit tier is willing to pay.

This gap exists because the enabling layer requires an operator, not just capital. A pension TPA is not a financial model — it is a process, a team, a set of regulatory relationships, and a client book that stays only as long as the service quality justifies it. An MGA is not a ledger entry — it is carrier trust, underwriter expertise, and broker relationships that are entirely personal until they are institutionalised. Financial engineering alone does not create the exit multiple. Operational integration does.

Three cross-theme risks deserve honest treatment and must be underwritten explicitly.

Integration is the whole return. Sector evidence from insurance distribution is unambiguous: non-integrated stacks of acquired firms trade at a discount to integrated platforms — sometimes a steep one. A buy-and-build programme that acquires without integrating is accumulating assets at entry multiples while destroying the exit multiple. The operator edge must be real and must show up in operational outcomes — shared infrastructure, cross-selling, data advantage — not just portfolio scale.

Talent is the binding constraint. Actuaries, pension administrators, specialty underwriters — the asset walks out every evening. Earn-out structures and culture integration are not diligence footnotes; they are the primary value-preservation mechanism in every acquisition. Any acquisition plan that doesn’t lead with the people question is an incomplete plan.

Roll-up regulatory scrutiny is rising. UK competition authorities have flagged serial-acquisition strategies as an enforcement priority. Below-threshold deals can attract review at scale. Building a competition-aware acquisition pathway from day one is not optional.

The Forward View

The UK defined benefit de-risking wave will run for the better part of a decade. The small-scheme long tail is now entering the pipeline in volume for the first time. Data remediation and administration capacity constraints will not resolve themselves without significant investment in specialist capability, and the statutory dashboard deadline of October 2026 creates a near-term non-deferrable spending obligation that is entirely independent of the transaction environment. Whoever owns the enabling infrastructure when the wave crests — and continues long after the last big-scheme BPA is written — owns a compounding, recurring revenue stream with structural demand pulling it forward from insurers, employers, and regulators simultaneously.

The MGA opportunity is narrower in timing but sharper in return profile if cycle-timed correctly. Soft-market entry, genuine operational integration of single-class specialists, and exit into a hard-market environment with multiple strategic acquirers competing for proprietary specialty capability is a well-defined path. The capital is patient; the cycle is not permanent; the window is 2026–2028.

Neither theme offers a passive return. Both require an operator who understands the specific regulatory, commercial, and human texture of the market — can originate off-market from relationships, not from a CIM — and has the discipline to integrate rather than merely aggregate. The institutions funding the headline trade in each value chain will eventually need to buy what has been built in the enabling layer. At their multiple, not ours.