Insurance and legal commentary, not advice on your specific position. Aggregation outcomes are highly fact-sensitive — consult your broker and legal advisors. Apex Insurance Brokers Limited is authorised and regulated by the Financial Conduct Authority, FRN 724952.

AIG Europe Ltd v Woodman & Others [2017] UKSC 18 — the leading PI aggregation case

If you only read one case on aggregation, read this one. AIG Europe Ltd v Woodman is the Supreme Court's leading authority on what the phrase "a series of related matters or transactions" means in a professional indemnity policy. That phrase is in the SRA Minimum Terms, in nearly every commercial PI wording sold in the UK, and in many international wordings that have been imported into the London market. The construction the Supreme Court adopted in Woodman now governs how UK insurers and brokers think about aggregation across every regulated profession.

This article is Spoke 1 of the Apex hub on aggregation and series clauses in PI insurance. It assumes you have read the hub or are comfortable with the four aggregation triggers and the basic mechanics of aggregation. It goes deeper than the hub on the case itself: facts, procedural history, the four arguments the Supreme Court heard, Lord Toulson's reasoning, the three-step test, and how the test has been applied since 2017.

Plain English explanation

A firm of solicitors looked after escrow money for two property development schemes. The firm released the money to the developers without taking proper security over the development land. Both schemes failed and 214 investors lost roughly £10 million. The firm's PI policy had a £3 million per-claim limit. The insurer said all 214 investors aggregated into one claim — so its total exposure was £3 million. The investors said they were 214 separate claims, or at most two groups of about 107. The Supreme Court split the difference: the schemes were not related to each other, but investors within a scheme were related, so there were two aggregated claims, not 214 and not one.

The court did not, however, just decide the case. It laid down a test of general application for the words "series of related matters or transactions" — and that test, the Woodman three-step, now governs aggregation analysis under SRA Minimum Terms and analogous wordings everywhere in UK PI.

The case: facts and procedural history

The litigation arose from two Midas International Property Development Plc schemes promoted between 2005 and 2008. One scheme acquired a development site in Turkey; the other, in Morocco. Both were marketed to UK retail investors as opportunities to acquire fractional interests in completed holiday resorts. Investors paid their money into client accounts held by the International Law Partnership LLP ("ILP"), a small firm of solicitors. ILP's role was to act as escrow agent: it was to release the investor money to the scheme promoters only when the scheme developer had provided adequate security over the development land, sufficient that investors' money would be protected if the developer defaulted.

Both schemes collapsed. The Turkey site was sold by ILP and the proceeds returned (in part) to investors; the Morocco site was lost altogether. ILP went into administration. The 214 investors brought third-party rights claims against ILP's professional indemnity insurer, AIG Europe Ltd, under the SRA Minimum Terms.

ILP's policy was an SRA Minimum Terms compliant primary policy. The aggregation clause was the SRA's standard clause 2.5(a):

"The insurance may provide that, when considering what may be regarded as one Claim... all Claims against any one or more Insured arising from: (i) one act or omission; (ii) one series of related acts or omissions; (iii) the same act or omission in a series of related matters or transactions; [or] (iv) similar acts or omissions in a series of related matters or transactions ... will be regarded as one Claim."

AIG argued aggregation under rung (iv): similar acts or omissions (the release of escrow money without adequate security) in a series of related matters or transactions (the two schemes, viewed as one series). If that was right, all 214 investors aggregated into one claim and AIG's liability was capped at the £3 million per-claim limit. The investors argued that the matters or transactions were not "related" within the meaning of rung (iv) and that, at most, each scheme stood alone. On that argument, there were either two aggregated claims (about £5 million each) or 214 separate ones.

Teare J at first instance ([2016] EWHC 2615 (Comm)) sided largely with AIG. He held the matters were sufficiently related. The Court of Appeal ([2016] EWCA Civ 367) reversed. It adopted a narrow "intrinsic relationship" test — the matters or transactions had to be related to each other in some way that did not depend on a third common factor, and on the facts the schemes were not. AIG appealed to the Supreme Court.

The legal issue

The Supreme Court was asked the apparently simple question: what does "in a series of related matters or transactions" mean? The answer was important well beyond this case. The same phrase is in nearly every UK PI wording. If the Court adopted the Court of Appeal's narrow test, insurers would face significantly more fragmented claim exposures and would re-price accordingly. If it adopted Teare J's broader approach, policyholders would face more frequent aggregation and would (rightly) demand higher limits.

The Insurance Act 2015 did not bear directly on the construction of the aggregation clause. But the Act's duty of fair presentation under section 3 sat in the background. If aggregation could be triggered by features of the firm's work that the insurer did not know about, those features were arguably material circumstances that the firm should disclose. We explore that link in our Insurance Act 2015 overview.

The Supreme Court's decision

The Supreme Court unanimously allowed AIG's appeal in part. Lord Toulson gave the leading judgment. Lords Mance, Clarke, Sumption and Reed agreed.

The key passage is at paragraph 22:

"The words 'in a series of related matters or transactions' require an examination of the matters or transactions to see whether they are connected in some real way. There must be some 'inter-connection' between them, and that inter-connection must be one which, viewing matters realistically and in the round, would lead an objective observer to say that the matters are 'related' for the purposes of the clause."

The Court rejected the Court of Appeal's "intrinsic relationship" test as too narrow. It also rejected any test that required the transactions to be "dependent on each other" or "conditional upon each other". The test is open-textured and fact-sensitive: are the matters or transactions connected in a real way, taking a realistic view in the round?

Applied to the facts, the Turkey and Morocco schemes were not related to each other. Each had its own site, its own investor pool, its own trust deed, its own scheme documentation, and its own escrow account. The fact that ILP made similar alleged errors across the two schemes did not, by itself, make the schemes themselves related. Within each scheme, however, the investors' claims did aggregate. They were participating in a single project, in a common trust structure, with a common security failure. The case was remitted on that basis. The practical effect was two aggregated claims, not one and not 214.

The principle distilled — the three-step test

Drawing the threads together, the Woodman analysis can be applied as a three-step test, which we set out in the hub and reproduce here for completeness.

Step 1 — Are there matters or transactions at all? The aggregation clause aggregates by reference to matters or transactions. So you have to be able to identify the unit of work that gave rise to each claim. For a solicitor, that is usually a file. For a surveyor, an instruction. For an IFA, an advice event. If a claim cannot be tied back to a matter or transaction, aggregation under this trigger fails.

Step 2 — Do the matters or transactions form a series? A "series" requires more than one matter or transaction and some connecting thread. The thread can be temporal (same period), structural (same scheme), contractual (same client or counterparty) or documentary (same template documents). It does not need to be intrinsic to the transactions themselves.

Step 3 — Is the series "related" in a real way? This is the open-textured question. The test is objective ("an objective observer"), realistic ("realistically") and global ("in the round"). The Court was deliberately non-prescriptive. Connections that have usually been held to satisfy the test include: common scheme, common trust deed, common security failure, common counterparty, common funding source, common documentation. Connections that have not, on their own, included: common solicitor, common species of error, common professional discipline, common period.

A negative answer at any step ends the analysis: no aggregation under "series of related matters or transactions".

Worked example with numbers

Take a firm with a £2 million primary PI policy and a £25,000 excess. The firm acts on 40 conveyancing transactions for a single developer over twelve months, all on the same new-build estate, using the same template purchase contract, the same template lease, and the same lender. A title defect is later discovered that affects every flat on the estate. Each buyer claims £75,000 of loss.

Without aggregation: 40 claims of £75,000 each = £3 million total loss. Excess applied 40 times = £1 million paid by the firm. Insurer pays £2 million up to limit per claim (£50,000 net of excess on each, capped at £2 million in total — assuming no separate aggregate). Total firm cost: at least £1 million in excesses plus any limit shortfall.

With aggregation under "series of related matters or transactions": 40 claims aggregate into one claim. £3 million total loss is the aggregated claim. One excess of £25,000. Insurer pays up to the £2 million limit. The firm pays £25,000 excess + £975,000 limit shortfall = £1 million. Same net cost, but applied very differently.

Net effect: with aggregation, the firm pays one excess and is exposed to limit shortfall above £2 million. Without aggregation, the firm pays the excess 40 times but in principle has £80 million of limit available (£2m × 40). For a firm carrying excess layers above the £2 million primary, aggregation drives losses up to the primary much faster but also unifies the excess. For a firm with no excess layers, aggregation usually hurts.

Now run the same numbers with 200 claims of £15,000 each:

Without aggregation: £3 million total loss × £25,000 excess applied 200 times = £5 million in excess payments — more than the loss itself. Insurer effectively pays nothing because each claim is below excess.

With aggregation: £3 million aggregated claim × £25,000 excess = £25,000 excess paid. Insurer pays £2 million. Firm pays £25,000 + £975,000 limit shortfall = £1 million.

This is why aggregation cuts both ways. Small claims aggregating up help the policyholder (excess unification). Large claims aggregating up hurt the policyholder (limit acceleration). The wording is fixed; the policyholder's preferred outcome flips depending on the cluster.

Sector implications

Solicitors. Woodman itself was a solicitors' case and the test applies directly under SRA Minimum Terms. See Spoke 4 on SRA aggregation. Particularly affected: scheme work, conveyancing fraud, trust account breaches, will-writing in bulk, residential conveyancing for a single developer.

Surveyors. RICS market wordings use comparable language. Aggregation issues arise where a surveyor repeatedly values for the same lender on the same estate, or repeatedly inspects for the same insurer. See Spoke 5.

Architects. Project-level work tends to aggregate naturally — one project, one matter — but multi-phase developments and framework agreements complicate the analysis. See Spoke 6.

IFAs. Woodman travels into IFA territory through DB transfer mass claims. Common scheme, common advice methodology, common ceding scheme — these can satisfy the Woodman test for aggregation across a hundred advised clients. See Spoke 7.

How insurers have used the case since 2017

In our claims practice we have seen Woodman deployed in three recurring ways since 2017.

Aggregation up to cap exposure. The most obvious use. Where a firm faces a large cluster of related claims, the insurer cites Woodman to argue that the cluster is a single aggregated claim subject to one limit. Common in conveyancing fraud sweeps, scheme failures, and DB transfer mis-selling.

Fair presentation argument. The insurer argues that scheme-level or cluster-level features were material circumstances that the firm should have disclosed under Insurance Act 2015 section 3. If the firm failed to disclose, the insurer pursues a proportionate remedy under Schedule 1 (premium uplift, partial decline, or — for deliberate or reckless breach — avoidance). Cluster-level features include: repeated work for one developer, repeated work in one scheme, repeated use of a particular template, repeated reliance on a particular counterparty. We covered the fair presentation duty in our dedicated piece on the duty of fair presentation and the material circumstance test.

Aggregation down to multiply excess. Less obvious, but real. Where the cluster is small and consists of low-value claims, the insurer occasionally argues that the matters are not related, so each claim attracts a separate excess. The effect is that the firm pays multiple excesses and the insurer pays nothing (because each claim falls below excess). The leverage runs the other way in this scenario, and your broker's job is to flush the inconsistency.

What this means for your firm

Four practical takeaways.

Disclose cluster features at proposal. When you fill in a PI proposal, do not just list the percentages of work by category. List the clusters — schemes, funds, repeat counterparties, template documents — that could become aggregating features under Woodman. The benefit is that the insurer has rated for them and cannot later argue non-disclosure. See our solicitors PI proposal completion guide.

Notify by cluster, not by file. Standard practice for sophisticated PI buyers since 2017. When a single complaint arises out of a connected scheme, notify "circumstances affecting [the scheme]" rather than "circumstances affecting [this file]". This crystallises the aggregation point at notification and protects against the insurer arguing that later-discovered claims are uninsured because they were not notified.

Test the limit against the worst-case cluster. A £2 million primary policy is not £2 million per claim against a 200-claimant scheme. Run the Woodman arithmetic on your top three aggregation exposures. If the worst case exceeds the limit, buy excess layers.

Build the analysis into the renewal process. Aggregation is now a renewal conversation, not a claims conversation. See Spoke 12 on negotiating aggregation at renewal.

How AIG v Woodman intersects with later case law

The Supreme Court's test is now the starting point for every PI aggregation case. Two later cases, both of which we cover in their own spokes, have refined and applied it.

Spire Healthcare v RSA [2022] EWCA Civ 17 — the Ian Paterson case — applied a Woodman-style analysis to the question whether claims by victims of the same surgeon, in the same hospital, over a 14-year period aggregated. The Court of Appeal held they did, by reference to "originating cause" wording (which is a wider trigger than Woodman's "matters or transactions" but the underlying logic of looking for "real" connections was the same). See Spoke 2.

Lloyds TSB v Lloyds Bank Group Insurance [2003] UKHL 48 pre-dated Woodman but laid the template for the three-step approach the Supreme Court refined. It is still cited as the leading authority on "originating cause" aggregation. See Spoke 3.

The Woodman test also features in our broader AIG v Woodman piece in the IA2015 case-law batch, which approaches the same judgment through the lens of fair presentation interactions.

FAQs

Q1. Is AIG v Woodman binding on me as a non-solicitor? The construction adopted in Woodman binds the courts on the wording the case construed. If your PI policy uses materially the same wording — and most UK PI wordings do — Woodman will govern. If your policy uses different wording (e.g. "originating cause" only), the case is highly persuasive but the specific test may differ.

Q2. Does Woodman apply to international PI policies underwritten in London? Yes if the policy is governed by English law. Many international PI policies are, even where the insured firm is overseas, because of London market practice. Check the governing law clause.

Q3. Does Woodman apply retrospectively to claims from before 2017? It applies to the construction of contracts that were already in force. Whether the claim arose before or after the judgment is irrelevant; the wording in the contract was always to be construed as the Supreme Court ultimately held.

Q4. Did the case change the SRA Minimum Terms? No. The SRA Minimum Terms have used the "series of related matters or transactions" formulation for a long time and were not amended after Woodman. What changed is the meaning courts will give that wording.

Q5. Can I argue that a cluster is not "related" if the connecting feature is internal to my firm? Possibly. The Woodman test focuses on whether the matters or transactions are related, not on whether the firm's approach to them is consistent. A common adviser making the same kind of error across unrelated transactions is, under Woodman, weak ground for aggregation. The matters themselves have to be connected.

Q6. What is the strongest aggregation argument for an insurer post-Woodman? A common scheme, fund or trust structure linking the claims. The Supreme Court accepted that investors in one scheme were aggregated because the scheme itself was the connecting feature. Replicating that structural unity is the cleanest path for an insurer.

Q7. What is the strongest non-aggregation argument for a policyholder post-Woodman? That the matters or transactions are structurally independent — different sites, different contracts, different counterparties, different funding sources — even if the firm's role across them was similar. Woodman itself is the precedent: similar errors across two independent schemes did not aggregate the schemes themselves.

Q8. Does Woodman bite on excess as well as limit? Depends on the wording. The SRA Minimum Terms apply the aggregation rules to the limit. Many commercial wordings apply them to both limit and excess; some apply different aggregation rules to each. Woodman governs the construction wherever the words appear.

Q9. How does Woodman interact with run-off cover? Aggregation analysis in run-off uses the same wording and the same test. The practical consequence is different: run-off cover usually has a single aggregate over a fixed period (six years under SRA Minimum Terms), and a Woodman-style aggregation in run-off can exhaust the aggregate early in the period.

Q10. Is Woodman likely to be revisited by the Supreme Court? Not soon. The test is well-bedded. Refinements are likely at Court of Appeal level on specific factual patterns (cyber, mass tort, multi-claimant fraud) rather than at Supreme Court level on the test itself.

AIG v Woodman - full case deep-dive on aggregation