Category: Specialist underwriting · Reviewed by Mark Fox, Broker · Renewals · Last reviewed May 2026
An independent financial advice practice in the Midlands received 40 substantively similar complaint files within an 18-month period. The complaints arose from advice given by a single adviser over a period of several years, all relating to a particular structured product range and a common pattern of unsuitability findings under the Financial Ombudsman’s redress methodology. The firm’s PI policy carried a £2m each-and-every-claim limit and a £6m aggregate. Forty complaints at an average redress figure of £45,000 produced gross exposure of £1.8m. On an each-and-every basis, every complaint sat well within limit and the firm’s retention applied once per file. On aggregation — if the 40 complaints could be treated as arising from a single originating cause — the entire exposure sat below one limit, but a single retention applied to the whole. The firm’s premium had been £42,500. The aggregation analysis was worth roughly forty times that figure in cash. This article sets out how UK PI aggregation works in practice, the leading cases, and why specialist brokers spend more time on aggregation language than on almost any other clause.
A PI policy is written with two limit constructs: an each-and-every-claim limit and an aggregate limit. The each-and-every limit is the maximum the insurer will pay for a single claim. The aggregate is the maximum the insurer will pay across all claims in the policy period.
Aggregation is the mechanism that determines whether a series of related matters is treated as one claim, eroding limits once, or as multiple separate claims, each eroding limits independently. The economic consequences move in both directions. For the firm, aggregation under a single retention means one excess applies to the whole — a single £25,000 retention rather than forty individual £25,000 retentions on the IFA example above. For the insurer, aggregation under a single each-and-every limit means a series of related losses is capped at the each-and-every figure rather than potentially multiplying out under the aggregate.
In practice, for any mid-tier programme, the aggregation language carries more cash significance than the difference between two premium quotes. A 10% premium saving means little if a multi-million-pound related-matter is then aggregated under one £1m primary limit when broader aggregation language would have applied a different — and more favourable — structure. The reverse also applies: where the firm benefits from a higher each-and-every limit and a relatively low retention, narrow aggregation may be preferable.
UK PI wordings use a spectrum of aggregation language. From narrowest to broadest:
“Single act or omission” is the narrowest formulation in common use. Under this test, claims aggregate only if they arise from one literal act of negligence. A single advisory letter giving incorrect guidance to ten clients may aggregate. A pattern of similar advice given separately to ten clients on ten occasions almost certainly will not. Insurers writing narrow-aggregation policies effectively underwrite each claim independently, which produces lower limits-exposure per claim but multiplies retention application.
“Series of related acts or omissions” broadens the test. Claims aggregate where they form a series, and where the acts are related to one another. The test depends heavily on the factual matrix: how closely related must the acts be, and what relationship is required? The case law tends to require a real causal or substantive link, not merely similarity of subject matter.
“Originating cause” / “any one originating cause” is broader still. The aggregation test focuses on the source of the loss, not on the acts themselves. Claims arising from a single original mistake, decision, system failure or course of conduct may aggregate even where the operative acts are individually distinct.
“Originating from a single source” and similar phrases sit at the broadest end. These tests have generally been read by the courts as catching claims that share a common root, even where the chain of causation is long.
The breadth of the test is not just a question of words but of judicial construction. The leading UK authorities have made it clear that subtly different wording can produce materially different outcomes.
The leading modern UK authority on PI aggregation is AIG Europe Ltd v Woodman [2017] UKSC 18, the Supreme Court’s decision arising from the collapse of the Midas escrow scheme. The underlying facts involved a firm of solicitors who had acted in setting up an escrow arrangement for investors in property development schemes. The escrow protections failed, investors sustained losses, and a large number of separate claims were brought against the solicitors. The question for the court was whether the claims aggregated under the Minimum Terms and Conditions wording then in force for solicitors’ PI, which provided that “all claims against any one or more insured arising from one act or omission; one series of related acts or omissions; the same act or omission in a series of related matters or transactions; similar acts or omissions in a series of related matters or transactions” should be treated as one claim.
The Supreme Court, in a judgment given by Lord Toulson with Lord Mance and Lord Sumption concurring, focused on the “related matters or transactions” element. Two points are central to the decision and remain the touchstones for current UK practice.
First, “matters or transactions” is wider than “contracts”. The test was not whether the claims arose from a single contract but whether they arose from related transactions in a broader sense. The court was prepared to look at the substantive commercial reality of the underlying arrangements rather than their precise legal form.
Second, the word “related” requires that the transactions have an “intrinsic” relationship with each other — not merely that they share a common feature such as a single solicitor or a single document type. The Supreme Court rejected the broader test that would have aggregated claims simply because they shared similar characteristics. The relationship had to be intrinsic to the transactions themselves.
The practical effect of Woodman is that aggregation under the “related matters or transactions” formulation requires a real, substantive connection between the underlying transactions, and is not satisfied by surface similarity. For PI wordings that adopt the Minimum Terms language or similar drafting, the case is the starting point of any serious aggregation analysis.
The earlier authority on “any one originating cause” is Lloyds TSB General Insurance Holdings Ltd v Lloyds Bank Group Insurance Company Ltd [2003] UKHL 48. The case arose from claims by customers of Lloyds Bank for negligent pensions advice given by the bank’s representatives over an extended period. The insurance policy aggregated claims arising from “any one originating cause”.
The House of Lords held that the breadth of “originating cause” was such that the claims aggregated. Lord Hoffmann’s analysis identified that the test looked back to the underlying root from which the claims sprang. The court found that the failure to give adequate training, supervision and compliance support to the bank’s representatives constituted a single originating cause from which the individual instances of mis-selling flowed.
The case is frequently cited for the proposition that “originating cause” is the broadest aggregation language in common use, and that it permits aggregation across an extended series of substantively similar but factually distinct claims where they share a common organisational or systemic root. Wordings adopting “originating cause” language are favoured by insurers seeking to cap their exposure across a book of similar claims and, conversely, are sometimes favoured by insureds where the breadth of aggregation simplifies the application of retention and limits.
The case law is best understood through worked examples drawn from common professional services scenarios.
A single-design error across multiple projects. An architectural practice reuses a structural detail across ten projects. The detail is later found to be defective. Ten separate claims are made by ten different developers. Under “single act or omission”, the claims may aggregate as one act if the detail was drawn once and reused without modification; the position is fact-sensitive. Under “originating cause”, they almost certainly aggregate — the single design decision is the originating cause. Under “series of related acts or omissions”, the analysis turns on whether the reuse of the detail constitutes a series, which depends on the underlying conduct of the practice.
A software error in a tax practice. An accounting firm uses bespoke software to file VAT returns on behalf of 200 clients. A coding error causes incorrect figures to be submitted on behalf of all clients in a given quarter. Under any of the aggregation tests, this is likely to aggregate: the single software error is a clear common root. The harder analysis arises where the error affected different clients in different ways depending on their underlying transactions.
Missed renewal advice across an IFA’s book. A financial adviser fails to recommend annual reviews to a cohort of pension clients over an extended period. Claims arise from individual clients whose portfolios suffered as a result. Aggregation here is highly fact-sensitive. Under “single act or omission”, the claims will probably not aggregate — there is no single act, there is an ongoing pattern of omission across many clients. Under “originating cause”, aggregation may follow if the failure can be characterised as flowing from a systemic absence of review processes. Under Woodman’s “related matters or transactions” test, the question would be whether the underlying client relationships were intrinsically related — which, for an IFA book, they generally are not.
Two firms identical in every respect except aggregation language can produce dramatically different recoveries on a related-matters loss.
Firm A is insured with a £2m each-and-every / £6m aggregate policy that aggregates on “single act or omission”. A series of forty related but factually distinct complaints produces forty separate claims. Each claim attracts a retention of £25,000 and each erodes the each-and-every limit independently. Forty retentions equal £1m of uninsured exposure. Each claim sits inside the £2m each-and-every limit, but the aggregate is consumed by total settlements and defence costs across all forty claims.
Firm B has the same headline limits but aggregates on “originating cause”. The forty complaints, sharing a common adviser and product range, aggregate as a single claim. One retention of £25,000 applies. The combined exposure sits within the £2m each-and-every limit, leaving the rest of the aggregate intact for any other matters during the policy year.
Which structure is better depends entirely on the firm’s claims profile. A firm with predominantly individual, unrelated matters benefits from narrow aggregation: separate retentions on small claims are tolerable; each-and-every limits are not eroded by aggregation. A firm with concentrated claims activity around a single adviser, system or product line benefits from broad aggregation: one retention, one limit-erosion event.
The reverse also occurs. Insureds sometimes prefer broad aggregation because it simplifies the claims process — a single notification, a single excess, a single defence team. Insurers sometimes prefer narrow aggregation because it limits their maximum exposure on any single related-matters event to the each-and-every limit rather than the aggregate.
A serious aggregation analysis at quote stage involves three steps.
First, a scenarios test. The broker takes the firm’s actual book of work — services, products, client types, fee patterns — and constructs hypothetical loss scenarios that reflect the most plausible claim patterns. For a quantity surveying practice, this might include a methodology error across a portfolio of similar developments. For an IFA, it might include an unsuitable-advice pattern across a single product range. For a law firm, it might include a missed limitation date pattern from a single practice area. The aggregation language is then tested against each scenario: would the matter aggregate as one claim or as many?
Second, a fact-pattern stress test. The broker takes the firm’s last five years of claims activity and applies the proposed wording’s aggregation language retrospectively. Would past claims have aggregated differently? Would historic recoveries have been better or worse under the proposed structure?
Third, a side-by-side wording comparison. Where the firm has more than one quote, the aggregation language is compared word for word, not at headline level. “Single act or omission, or related series” and “any one originating cause” are different tests with different case law and different outcomes.
For firms with a tower of cover, the analysis extends to whether the primary and excess wordings aggregate on the same basis. Mismatched aggregation across a tower is a recurrent source of disputes.
Aggregation is the mechanism by which two or more related claims are treated as a single claim under the policy. The effect is that one retention applies, one each-and-every limit is consumed, and the aggregate is eroded once rather than multiple times. Aggregation language sits at the centre of any PI policy and frequently has more cash significance than headline limits or premium.
AIG Europe Ltd v Woodman [2017] UKSC 18 is the leading modern Supreme Court authority. The case arose from the Midas escrow scheme and turned on the “related matters or transactions” language in the solicitors’ Minimum Terms wording. The court held that “related” required an intrinsic relationship between the underlying transactions, not merely surface similarity.
“Single act or omission” is narrow — claims aggregate only if they arise from one literal act. “Originating cause” is broad and looks back to the underlying root from which the claims sprang, as analysed in Lloyds TSB v Lloyds Bank Group Insurance [2003] UKHL 48. The two tests can produce materially different outcomes on the same fact-pattern.
No. Broad aggregation means one retention applies to a series of related claims, which is cash-positive for the insured. But it also means the each-and-every limit caps the combined exposure rather than each claim having its own limit. For firms with concentrated claims activity around a common cause, broad aggregation is generally better; for firms with diverse, unrelated matters, narrow aggregation can be preferable.
Where the primary policy aggregates on a broader basis than the excess, related claims may consume the primary as a single claim but be treated as separate claims by the excess insurer. The result is that the excess attachment point may not be reached even where the primary appears to have exhausted. Mismatched aggregation across a tower is a recurrent source of coverage disputes.
Yes, although the breadth of any change depends on the market appetite. Some insurers have standard wordings they do not depart from; others will adjust aggregation language for specific risks. Where the firm’s claims profile clearly favours one aggregation basis over another, the broker should test the available market for the most appropriate language.
Woodman established that aggregation under “related matters or transactions” requires an intrinsic relationship between the underlying transactions. Practical effects include that not all claims involving the same fee-earner, the same document type, or the same client sector will aggregate; the underlying transactions must be substantively linked. Wordings adopting Minimum Terms or similar language should be read in light of the case.
A specialist broker should run a scenarios test against the firm’s actual book of work, a stress test against the firm’s last five years of claims, and a side-by-side comparison of the wording across competing quotes. Headline limits and premium are secondary to the aggregation language itself.
Apex Insurance Brokers Ltd is an independent UK insurance broker based in Bristol, advising professional services firms on professional indemnity insurance and related covers. The firm is authorised and regulated by the Financial Conduct Authority (firm reference 724952) and registered at Companies House (company number 07014570).
This commentary reflects market conditions as at May 2026 and is provided for general information. Insurance market conditions, policy wordings and regulatory positions change frequently; firms should obtain advice specific to their circumstances rather than rely on general commentary.
Apex Insurance Brokers serves UK professional services firms and commercial businesses. Call 0117 325 0027, email hello@apexinsurancebrokers.co.uk, or request a quotation.
Get a quote