Conveyancing fraud and aggregation across multiple properties

Two very different fact patterns

It helps to separate the cases before applying the clause. In the first pattern, an insider steals: a partner or employee takes money from the client account across many unrelated matters over a period of years. Each victim is a different client, each theft a different transaction, and the only common feature is the thief. In the second pattern, an external fraud is worked through the firm: an identity fraudster poses as the registered owner of several properties, or a single fraudulent scheme runs multiple purchases financed by the same lender, and the firm’s alleged failing is the same due diligence step missed on each. Our conveyancing fraud entry sets out how these frauds are typically executed.

The second pattern has a much better prospect of aggregating than the first, because there is a real prospect that the matters are related to one another and that the same or similar omission runs through them. The first pattern is where aggregation arguments most often fail.

Why serial theft does not aggregate

The Court of Appeal addressed this directly in Baines and others v Dixon Coles & Gill (a firm) and others [2021] EWCA Civ 1211, decided on 6 August 2021. A partner in a small firm had misappropriated client money over a long period across many unconnected matters before the thefts were discovered, and was later convicted. The insurer argued that the resulting claims aggregated, relying principally on the second limb of the SRA minimum terms — one series of related acts or omissions — and alternatively on the first.

The Court refused to aggregate. The first limb failed because each theft was a separate act; a single dishonest objective running through a course of conduct does not turn many acts into one. The second limb failed because, applying Lloyds TSB General Insurance Holdings v Lloyds Bank Group Insurance Co Ltd [2003] UKHL 48, each claim has to arise from the series taken as a whole, and each client’s loss here arose from the theft that affected them, not from the others. Having the same source — the same dishonest individual — was not enough.

The consequence for a firm in that position is severe and worth stating plainly. Non-aggregated claims each attract their own excess, and where the aggregate of the individual claims is large, the firm carries the excess many times over even though a single limit would comfortably have covered the total. Insurers, of course, face the opposite arithmetic: many separate claims each get their own limit.

Where a multi-property fraud can aggregate

The third and fourth limbs are the ones that do the work when the fraud is external and systematic. If the firm acted on several purchases forming part of a single scheme — the same fraudulent vendor, the same portfolio, the same lender relying on the same certificate — then there is a serious argument that the matters or transactions are related, and that the same or similar omission (the identity check that was not done, the source of funds enquiry that was not made) ran through all of them. That is the territory AIG Europe Ltd v Woodman [2017] UKSC 18 governs: the transactions do not have to be dependent on one another or intrinsically related, but they must fit together and be connected in significant ways when looked at objectively and in the round.

What tends to decide it in practice is whether the properties and the parties are genuinely linked or merely alike. Several purchases within one development, for one fraudster, funded by one lender under one facility, look connected. A dozen unrelated remortgages of unrelated properties for unrelated clients, all defeated by the same weak identity procedure, look far less so — the omission is similar but the matters are not obviously related to each other. Similarity of method is precisely what the Supreme Court held was insufficient when it declined to aggregate across the two developments in Woodman.

The lender dimension

Property fraud claims usually come from a lender rather than a buyer, and lender claims add their own layer. A single lender may bring many claims arising from many advances, which raises the question whether they are one matter or many. The final part of the SRA aggregation wording — that all claims arising from one matter or transaction are regarded as one claim — means that all the claims flowing from a single conveyance, from buyer, lender and anyone else, come together as one. That does not, by itself, join separate conveyances.

Note too that under the minimum terms an insurer must generally still meet claims arising from the dishonesty of an individual, subject to the exclusion of the dishonest person’s own claim, which is why these disputes are about how many limits respond rather than whether any does. The SRA minimum terms guide covers that structure, and the solicitors’ PI page the placement.

What to do about it before it happens

Since the wording cannot be changed on the compulsory layer, the work is in limit setting and notification. Set the limit against a scenario in which claims do not aggregate as well as one in which they do, because the excess exposure in the non-aggregating case is a real and often overlooked cost. Notify a suspected fraud as a single circumstance, described broadly and early, rather than reporting individual files as they surface: how a circumstance is framed at notification has a real bearing on how the resulting claims are treated. And keep the evidence that links transactions together, because that evidence is what an aggregation argument later runs on.

Frequently asked questions

If one dishonest employee stole from many clients, is that one claim?

On the authorities, generally not. In Baines v Dixon Coles & Gill [2021] EWCA Civ 1211 the Court of Appeal held that each theft was a separate act, so the first limb did not apply, and that each claimant’s loss arose from their own theft rather than from the series as a whole, so the second limb did not apply either. A shared dishonest source is not enough.

Can purchases of several different properties aggregate?

They can, where the transactions are genuinely connected — one scheme, one fraudster, one lender facility, one development — and the same or a similar omission runs through them. Following AIG Europe v Woodman, the matters need not be dependent on each other but must fit together and be connected in significant ways viewed objectively. Similarity of method alone was held insufficient.

Why does non-aggregation hurt the firm if the limit is large enough?

Because the excess applies claim by claim. Twenty separate claims mean twenty excesses, and in volume conveyancing that uninsured layer can exceed what the firm could absorb, even though a single limit would have covered the total loss comfortably. This is why the limit and the excess structure should both be tested against the non-aggregating scenario.

Does how we notify affect whether claims aggregate?

It can matter a great deal. A suspected fraud notified early as one broadly described circumstance, with the evidence of what links the matters, gives a much better foundation than a trickle of individual file notifications spread across policy years. Notification framing does not change the wording, but it shapes the facts an aggregation argument later has to run on.

This page is general insurance information about how aggregation provisions in UK professional indemnity policies are commonly drafted and argued. It is insurance information, not legal advice, and it is not a statement of what any particular policy covers. Position stated as at August 2026. Where a claim, a circumstance or a limit is actually in issue, read your own wording and take advice on your own facts.

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