Excess layer PI insurance explained

Reviewed by Matthew Bartlett, Director, Apex Insurance Brokers Limited · Last reviewed 2026-08-10

Most professional firms start life with a single PI policy from a single insurer. For many, that remains the right structure indefinitely. But once a firm's contractual obligations, claim scenarios or regulatory exposure push the required limit beyond what one insurer is comfortable writing — or beyond what one insurer will write economically — the market's answer is layering: a primary policy at the bottom, with one or more excess layer policies stacked above it. This article explains how those programmes actually work, why they are usually structured the way they are, and where the traps sit.

How does a layered PI programme actually work?

Picture the total limit as a tower. The primary policy occupies the ground floor — say the first £2m or £5m of cover. It is the working layer: it deals with notifications, appoints defence lawyers, handles the day-to-day claims activity, and pays first. Above it sits the first excess layer, which attaches only when the primary limit is exhausted by a covered claim. A second excess layer attaches above that, and so on, until the tower reaches the total limit the firm needs.

Each layer is a separate contract of insurance, typically with a different insurer (or a different mix of insurers), its own premium, and its own policy document. The excess insurer's obligation is defined by its attachment point — the amount of loss that must erode the layers below before it pays anything. A £5m excess layer attaching at £5m, written above a £5m primary, gives the firm £10m in total: the excess insurer contributes nothing to the first £5m of a loss and everything from £5m to £10m.

Two mechanics matter in practice. First, whether the programme operates on an any one claim or aggregate basis at each level — a distinction that changes how much cover is actually available across a policy year, and one we cover in detail in our guide to any one claim versus aggregate PI limits. Second, how the excess layer treats erosion of the underlying limit: most excess wordings are drafted so the layer drops down and attaches once the underlying insurers have paid (or been held liable to pay) their limits, but the precise trigger language varies between wordings and deserves scrutiny rather than assumption.

What does "follow form" mean — and does it really follow?

The phrase you will hear most in this market is follow form. A follow-form excess layer adopts the terms, conditions and exclusions of the underlying policy rather than setting out its own full wording. Conceptually, this is what a buyer wants: one set of coverage terms running vertically through the tower, so that a claim covered at the bottom is covered all the way up. If the primary insurer pays, the excess insurers should pay on the same basis once their attachment points are reached.

The caveat — and it is an important one — is that "follow form" is rarely absolute. Most excess wordings follow the underlying form except as otherwise stated, and that exception clause is where the drafting lives. Excess layers commonly retain their own provisions on limits, attachment, notification, cancellation and premium, which is unobjectionable. Less benign are excess wordings that quietly impose their own exclusions, their own claims control conditions, or a different definition of when the layer is triggered. A tower with misaligned wordings can leave a firm covered for a claim at primary level but facing an argument at excess level — precisely at the loss sizes where the firm can least afford a coverage dispute.

The alternative to follow form is a standalone excess wording: a full policy in its own right, drafted by the excess insurer. These can be perfectly sound, but they demand a line-by-line comparison against the primary, because nothing about them is aligned by default. In either case, the broker's job is verticality: making sure the coverage a firm thinks it has at £10m is the same coverage it has at £2m.

Why is layering often cheaper per million than one big limit?

The economics of a tower reflect where the risk actually sits. Claims frequency is concentrated at the bottom: most PI claims, even against substantial firms, settle within the primary layer. The primary insurer prices for frequency and severity — it expects to handle notifications every year and to pay claims with some regularity. An insurer writing a layer attaching at £5m or £10m is pricing something quite different: the probability that a single claim (or, on an aggregate basis, a year's claims) burns through everything beneath it. That probability falls as the attachment point rises, and the rate per million falls with it.

This is why a firm buying £10m as a tower — a primary layer plus excess layers — will typically pay less in total than the same firm would pay for a £10m limit from a single insurer, if a single insurer would offer it at all. It is also why the rate on the top layer of a well-built tower is usually a fraction of the rate on the primary. Layering converts one insurer's reluctance to concentrate that much exposure into several insurers' willingness to take a slice they can each price comfortably.

There is a second, structural benefit: diversification of counterparty risk. With the total limit spread across several insurers, no single carrier's appetite change, withdrawal from the class, or solvency problem takes the whole limit with it. At renewal, a broker can re-market individual layers without dismantling the entire programme.

When does a firm outgrow a single-carrier policy?

There is no fixed threshold, but the signals are consistent. The most common trigger is contractual: a client, funder or framework agreement requires a PI limit above what the firm currently carries, and often above what the incumbent insurer will extend to. Firms in this position should read our guide on what to do when a contract requires a higher PI limit before agreeing terms, because the drafting of the requirement affects what needs to be bought.

The second trigger is exposure-led rather than contract-led. A firm advising on transactions or projects where a single error could plausibly generate a loss well beyond its current limit — high-value property work, corporate transactions, structural engineering on major schemes, tax planning at scale — may conclude that its realistic worst case, not its contractual minimum, should set the limit. For regulated professions the compulsory layer is only a floor: solicitors' minimum terms cover, for instance, sits at £2m any one claim (£3m for recognised and licensed bodies), figures that bear no relationship to the size of claims a substantial practice can face. The layered market is how the gap gets bridged.

The third trigger is capacity itself. Insurers manage line size — the maximum they will commit to one risk — and many are simply unwilling to put their full theoretical capacity on a single professional firm. Once your required limit exceeds the line your insurer will deploy, layering stops being an option and becomes the only route.

How does a broker build and place a layered programme?

Placement starts at the bottom, because everything above inherits from it. The primary layer is the hardest to place and the most consequential: its wording becomes the reference document for the tower, its insurer usually controls claims, and its pricing anchors the rates above. A broker will place the primary with an insurer whose wording is strong, whose claims handling is proven in the relevant profession, and whose appetite for the risk is durable — a primary carrier that exits the class at the first renewal destabilises the whole structure.

With the primary agreed, the broker markets the excess layers. Sensible practice at this stage includes:

• Structuring attachment points deliberately — layer sizes are chosen to match where different insurers' appetites and pricing are most efficient, not carved arbitrarily.
• Aligning wordings vertically — securing genuinely follow-form terms where possible, and documenting every departure where not.
• Creating competitive tension per layer — each layer is a separate placement, so each can be marketed separately; a strong price on one layer can be used to discipline pricing on the others.
• Checking the joins — confirming that each layer's attachment language meshes with the exhaustion language below it, so no gap opens between layers, and that notification given to the primary insurer is treated as (or duplicated to) notification up the tower.
• Stress-testing renewal alignment — common renewal dates and consistent claims information across all layers, so the tower renews as a programme rather than as unrelated policies.

On larger towers, individual layers may themselves be shared between insurers by subscription, each taking a percentage of that layer. The mechanics are the same; the coordination burden is simply higher, which is one reason high-limit placement rewards brokers who do it regularly.

Where do layered programmes go wrong?

Almost every problem with a tower traces back to one of three failures. The first is wording misalignment — an excess layer with a broader exclusion, a different retroactive date, or its own notification condition that nobody complied with because everyone assumed follow form meant follow form. The second is a gap at the joins: exhaustion and attachment language that does not quite meet, or an underlying limit eroded in a way the excess wording did not anticipate — for example, where defence costs erode the primary limit but the excess layer's trigger contemplates payment of damages. The third is programme drift: layers renewed at different dates, with different information, until the tower's insurers hold materially different views of the same risk. None of these problems is visible until a large claim arrives, which is exactly when they become expensive.

The discipline that prevents them is unglamorous: read every wording, reconcile every layer against the primary at each renewal, and keep the whole programme moving on one timetable with one consistent presentation of the risk.

What should we bring to the table before approaching the market?

High-limit placements are underwritten on information. A firm seeking a layered programme should expect to present, and benefits from presenting well: a clear account of its work split and fee income; its largest engagements and the contractual limits of liability it negotiates; its claims and circumstance history with context, not just numbers; and its risk management framework — supervision, peer review, engagement letter discipline. Underwriters at excess level are pricing tail risk, and a firm that can demonstrate it understands and manages its own tail risk gets meaningfully better outcomes, in terms and in premium, than one that submits a bare proposal form.

Timing matters too. Towers cannot be assembled well in the final week before renewal. Starting the process early gives the broker room to market layers competitively, resolve wording points before they harden into take-it-or-leave-it positions, and structure the tower rather than merely fill it.

Apex Insurance Brokers Limited is authorised and regulated by the Financial Conduct Authority (FRN 724952). This article is general information, not advice on a specific policy.