What Does a Claims-Made Clause Actually Cover?
Archive analysis of Italian case law from 2016–2017. The translation preserves the original argument and historical legal context; it is not an update on subsequent developments.
Italian Supreme Court of Cassation, Joint Divisions, No. 9140, 6 May 2016
Milan Court, No. 7149, 17 June 2016
Italian Supreme Court of Cassation, Third Civil Division, No. 10506, 28 April 2017
Udine Court, No. 613, 3 May 2017
CONTENTS: 1. Claims-made clauses as insurers’ solution to reserving uncertainty. 2. Unlimited retroactivity in claims-made cover. 3. “Impure” claims-made clauses. 4. The Joint Divisions and review of whether impure clauses deserve legal protection. 5. The Supreme Court’s objection to clauses without post-expiry notification. 6. Udine Court: a rebuttable presumption of invalidity. 7. Conclusions.
1. The starting point for discussing claims-made clauses is decisive. This short article aims to clarify the complex but crucial issue of effective professional indemnity cover—professional policies being, in this account, universally claims-made. We must therefore begin almost at the beginning, ensuring that the arguments lead to defensible conclusions.
That beginning is the Italian Civil Code. Article 1917 governs liability insurance around an event, the insured loss, necessarily occurring “during the period of insurance”.
If insurance covers a year, say 2016, an event occurring within that year is covered regardless of when the claim is notified to the insurer, subject to limitation periods. From the insurer’s perspective, the 2016 policy therefore represents a potential claim exposure until the injured third party’s claim becomes time-barred. The injured party, in turn, must be aware of the harm before the relevant limitation period begins.
An injured party might therefore report an event that occurred in 2016 many years later. The wrongdoer/insured would then notify its insurer years later, and the insurer would still be required to respond.
This traditional “loss occurrence” approach made insurers’ calculations difficult. How could an insurer determine whether a product sold in 2016 was profitable or disastrous? Having sold policies and collected premiums, it had to await payment of claims: broadly, premium income less claims payments determines the result. If payments can follow many years later, final accounting is delayed correspondingly, impairing forecasting, reporting and accurate pricing—too expensive to compete, or too cheap to meet future claims?
Claims-made insurance addresses this specific problem: identifying, within a defined period, the claims opened so that calculations can be more precise. This requires a closed population of claims, with no further indemnity demands outside the period allowed for opening them.
The fundamental shift is from the damaging event to notification of the claim. We must distinguish (A) the event, such as a professional error; (B) the injured party’s discovery and formal or informal challenge to the alleged wrongdoer, the relevant circumstance; and (C) notification by the actual or alleged wrongdoer to the insurer.
Under the claims-made model discussed here, the relevant insurance period concerns notification (C), while when the underlying event (A) occurred is essentially immaterial—even if before inception.
2. This leads to the conclusion that, in relation to the damaging event, a pure claims-made clause has unlimited retroactive effect by definition.
The reason is that the relevant event is notification rather than the original loss. Notification is a “new” event within the insurance period, although based on harm occurring long before.
To complete the picture, the insured must also be unaware, when taking out cover, of the underlying event (A). Substantial lack of awareness, generally presumed, is required; necessarily, the injured party must not already have challenged the insured formally or informally (B).
Consider a lawyer who unknowingly makes an error harming a client. The following year the lawyer obtains insurance, perhaps for the first time. Another year later, the error emerges, for example in a judgment, and the lawyer notifies the insurer under the renewed policy. The underlying event occurred while uninsured, but notification falls within the policy period; as that is the relevant element, cover operates.
What cannot happen is inception after the lawyer has become aware of the potential claim: the so-called known-circumstances issue.
From this perspective, pure claims-made insurance is neutral for the insured, functioning like occurrence cover once the shift from damaging event to notification is understood.
If an event occurs during the policy period but the insured remains unaware, timely notification is impossible. On discovery, the insured must notify the insurer covering that later period. Under the contract, expiry ends the opportunity to notify under the expired policy, regardless of when the event occurred.
3. Problems multiply when restrictions are placed on this model by limiting retroactive cover. Many insurers adopted what the original article characterises as an Italian modification of pure claims-made cover: limits on retroactive effect.
These “impure” or mixed clauses require not only notification within the relevant period but also the damaging event within a specified interval. At their most restrictive, that interval is the policy period itself, with no retroactivity. In the argument advanced here, even a few years’ retroactivity is inconsistent with the model: unlimited retroactivity is needed for proper operation. Otherwise, unpredictable circumstances can leave a person who believed they were adequately insured without cover.
The first conclusion is therefore that a pure claims-made clause is essentially neutral—neither better nor worse than traditional Article 1917 cover—whereas impure clauses require careful examination of the actual, potentially very limited, cover offered.
4. The Joint Divisions expressed this well in Judgment No. 9140 of 6 May 2016: in liability insurance, a clause requiring both the wrongful event and the compensation claim to arise within the policy period, or within predetermined intervals—a mixed or impure claims-made clause—is not an onerous standard term requiring special approval merely on that basis; nevertheless, in certain circumstances it may be void because the interests it serves do not merit legal protection.
The Joint Divisions thus recognised claims-made insurance’s departure from the Article 1917 model: its subject shifts from the damaging event to the notified claim, which necessarily presupposes that event. At the same time, they identified potential invalidity where an impure clause’s effects fail the Article 1322 Civil Code test of deserving legal protection.
Following the 2016 judgment, litigation therefore focused on whether impure clauses satisfied that test. The list of decisions continued to grow.
Two decisions upheld claims-made clauses: Naples Court, 20 June 2016, and Bologna Court, 12 August 2016. The first known decision finding a clause undeserving of protection was Milan Court, No. 7149 of 17 June 2016.
What clause did Milan invalidate? A particularly impure one, providing no retroactivity. The court found it incompatible with the professional liability involved: given the intellectual service’s characteristics and the inevitable gap between performance and manifestation of harm, it was virtually impossible for both the professional’s act or omission and the third party’s claim to occur in the same year.
Ultimately, the issue is how that test affects the validity of the insurer’s contractual terms.
5. A subsequent decision, Supreme Court Third Civil Division No. 10506 of 28 April 2017, sought to guide the assessment through three criteria drawn from case law.
According to the Court, a contract or term does not deserve legal protection:
1) when it seeks to confer an unjust, disproportionate advantage on one party without a corresponding benefit for the other;
2) when it places one party in an indefinite position of subjection to the other;
3) when one party compels the other to behave contrary to constitutionally imposed duties of solidarity. These criteria are certainly defensible.
The Court’s further conclusions in the particular case are less persuasive. It located the problem not in retroactivity, as in the Joint Divisions’ decision, but in the absence of post-expiry cover. It stated that a claims-made clause excluding later claims appeared undeserving of protection because it gave the insurer an unjust, disproportionate advantage without a corresponding benefit.
That is not so: the counterpart to pure claims-made cover is, as explained, unlimited retroactivity. Review should therefore examine whether that extension exists, at least on the Joint Divisions’ approach. To say the absence of post-expiry cover necessarily defeats the clause is effectively to say no claims-made clause, even a pure one, can fit the Italian legal system—a question already extensively addressed and resolved in the opposite direction.
Examining the facts, however, reveals that this policy also lacked unlimited retroactivity: it provided only three years. In outcome, therefore, the 2017 decision did not depart from the 2016 Joint Divisions. It invalidated an impure clause offering insufficient protection because its temporal scope was too narrow.
Nevertheless, expressions such as the one quoted deserve criticism because they are likely to support arguments that any claims-made clause is void, pure or otherwise, simply for excluding post-expiry notification. Moving the focus from retroactivity to later notification jeopardises legal certainty and may mislead readers unfamiliar with the subject into conclusions inconsistent with its contractual and judicial development. The Joint Divisions focused on the notified claim; the 2017 decision appears to assume the damaging event is the clause’s subject. On the argument advanced here, that is wrong: applying an occurrence-based framework will by definition make every claims-made clause appear undeserving of protection.
A leading argument illustrates the point: without later notification, a loss occurring on the policy’s last day would supposedly be uninsurable, because a compensation demand could hardly arise that same day. That is wrongly framed around the damaging event rather than notification. On the last day of a claims-made policy, notified claims may relate to any earlier period, even an event ten years old. Cover on the last day is therefore the same as on the first: claims first received by that date are covered. Saying a claims-made policy generally offers no real protection at the end of its period is incorrect. Insureds must understand the fundamental shift: the policy responds to previously unasserted claims, irrespective of when the underlying events occurred, unlike Article 1917’s focus on an error or injury within a defined period.
6. Finally, consider Udine Court Judgment No. 613 of 3 May 2017, involving two insurers.
Both policies were claims-made. The damaging event occurred in 2008. The first insurer’s policy, first taken out in 2011, was pure claims-made with unlimited retroactivity. However, the insurer established that the insured already knew material facts relating to the event when entering the contract. Those facts could reasonably call future liability for the advice into question: the financial police’s investigation concerned precisely the contribution relief on which the insured had advised. A future client compensation claim was therefore reasonably foreseeable.
The claim against the first insurer was dismissed. The second insurer’s policy was impure, with neither retroactivity nor later notification. It argued that notification first occurred in 2014, although the contractual relationship ended in 2011. Relying on the Joint Divisions, the court held that particularly restrictive clauses requiring both loss and claim within the insurance period, without any extension backwards or forwards, were very difficult to regard as deserving protection. It inferred a rebuttable presumption under Article 2729(1) of the Civil Code, subject to specific terms or circumstances supporting a different conclusion. None did in that case, so the impure clause was replaced by the statutory Article 1917 model.
7. The conclusions can be summarised as follows: pure claims-made policies are legitimate and deserve legal protection. As regards impure policies only, the absence of retroactivity, or very limited retroactivity, may support a rebuttable presumption that the clause is void for failing the test of deserving legal protection.
Comments