The retroactive date: the clause that decides whether your PI still covers the work you did last year

Most of the attention at a professional indemnity renewal goes to the limit and the premium. For a larger firm, the clause that most often decides whether a claim is paid is neither of those. It is the retroactive date: the single line in the schedule that sets how far back in time the policy reaches. It rarely comes up in the quote conversation, it is easy to lose without noticing, and when it goes wrong the firm usually finds out years later, when a claim lands in respect of work nobody is thinking about any more.

This is a programme-level issue, not a project one. The retroactive date protects every piece of work the firm has ever done, in aggregate, against claims that have not yet been made. It is the finance director, the managing partner or the general counsel who carries that exposure, and it is at their level — at renewal, at a broker move, and above all at any change of structure — that continuity is either preserved or quietly broken.

Why “claims-made” changes everything

Professional indemnity is written on a claims-made basis. That is the root of the whole question, so it is worth stating plainly: the policy that responds to a claim is the one in force when the claim is first made against you — or when you first become aware of the circumstance that leads to it — not the policy you held when you did the work. The work might have been done five years ago under a different insurer; what matters is which policy is live on the day the letter of claim arrives.

That is the opposite of how most business insurance works. Your office or liability cover responds to the year in which the event happened, and once that year is closed it stays closed. Claims-made cover is a rolling promise: this year’s policy carries not just this year’s work but every prior year still capable of producing a claim. Which is why the one-year policy cannot be read in isolation — the continuity between one year and the next is the cover.

The retroactive date, in one sentence

The retroactive date is the earliest date of work the policy will pay a claim for. A claim made this year is covered only if the work it relates to was carried out on or after that date; anything earlier is excluded, however blameless the firm, and however live the current policy. A policy written with “full retroactive cover”, or a retroactive date set at the firm’s inception, reaches back over the whole of its working history. A policy with a later retroactive date draws a line, and everything before it is on the firm’s own account.

While a firm renews continuously with cover that preserves the date, this sits quietly in the background doing its job: each year’s policy picks up the previous year’s work, and the chain holds. The danger is not the date itself but the moments when the chain can break — and those moments are nearly always the ones a firm chooses, not accidents that befall it.

Where larger firms lose continuity

The first is a change of insurer. When you move your cover to a new market, the new insurer sets its own retroactive date. If it matches your existing one, continuity holds and your back history stays covered. If the new insurer will only offer cover from the date it comes on risk — or imposes a retroactive date later than your true inception — a gap opens for the earliest years of the firm’s work, and no one may spot it until a claim falls into that gap. A new insurer is not obliged to inherit your old one’s start date; it has to be asked for, and checked on the schedule when the policy is issued.

The second is a change of broker made carelessly. Switching broker does not, by itself, change your cover — but a broker who does not understand the firm’s history can present the risk in a way that loses the retroactive date, or fail to insist the incoming insurer preserves it. A move handled properly keeps the date intact; handled as a price exercise, it can quietly reset the clock. The retroactive date is one of the first things a considered switch should protect.

The third, and the one that catches the biggest firms, is a change of structure. Incorporating a partnership, bringing several practices under a holding company, acquiring another firm, or reorganising a group all create a new or different insured entity. The policy covers the entity named in the schedule; it does not automatically follow the work of a predecessor partnership, an acquired business, or a merged practice. Unless the policy is specifically arranged or endorsed to pick up the prior entity’s past work — or that work is separately run off — a restructure can leave years of liability uninsured even though the people and the clients carried straight on. Where a firm is acquiring or merging with another, the target’s past work is a live due-diligence question in its own right: who is standing behind it, on what basis, and for how long.

The fourth is the simplest and the most avoidable: a gap in cover. Let a policy lapse, even briefly, and the claims-made chain breaks at that point. A claim made during the gap — or a circumstance that should have been notified during it — may have no live policy to answer. Continuity means uninterrupted, not merely renewed.

The known-circumstances trap at a switch

Preserving the retroactive date is necessary but not sufficient. A new policy almost always excludes claims and circumstances the firm was already aware of before the cover began. So even with the date carried across, anything the firm knew about at the point of the switch is the outgoing insurer’s responsibility, not the incoming one’s — and it only stays the outgoing insurer’s responsibility if it was notified to them before the policy ended.

That makes the weeks around a switch the highest-risk moment in the cycle. The discipline is to notify every known circumstance to the current insurer before you move, and to present the incoming risk fully and fairly, as the duty of fair presentation under the Insurance Act 2015 requires. A circumstance known but not notified before the change can end up covered by neither policy: too old for the new one, never reported to the old one.

When run-off is the answer instead

Sometimes there is no continuing policy to carry the past work — a practice closes, a partnership dissolves without a successor, a business line is discontinued. Then the past is protected not by continuity but by run-off cover: a policy that keeps answering claims made after the firm has stopped trading, for work it did while it traded. Regulated firms are generally required to arrange run-off when they cease; unregulated firms are not, which is why it is so often forgotten. The principle is the same as the retroactive date — a claims-made world only protects work that a live policy, continuing or run-off, still stands behind.

What a finance director or managing partner should actually check

None of this requires technical insurance knowledge to govern — only a handful of deliberate checks, made by someone senior enough to see the whole firm. Confirm the retroactive date on every renewal schedule, and satisfy yourself it still reaches back to the firm’s inception. Before moving insurer or broker, make preserving that date an explicit condition of the move, not an afterthought — and get the preserved date confirmed on the new schedule in writing. Notify every known circumstance to the outgoing insurer before any switch completes. On any restructure, acquisition or incorporation, treat the past work of every predecessor or acquired entity as a separate cover question and resolve it before the deal closes. And never let cover lapse, even for a day, to save a few weeks’ premium timing.

The common thread is that the retroactive date is managed, not bought. It survives as long as someone with a view of the whole firm makes sure each change preserves it, and it is lost the moment a renewal, a broker move or a reorganisation is handled as a one-off transaction rather than a link in a chain. The firm that treats continuity as the thing it is actually paying for is the one still insured for its own history when a claim finally tests it.

Frequently asked

We have renewed with the same insurer for years. Do we still need to worry about the retroactive date?
Less so, but it is still worth a look each year. Continuous cover with the same insurer usually preserves the date automatically, so the chain holds. The point to watch is any year the wording, insurer or structure changed — a mid-term insurer substitution, a group reorganisation, a gap while a renewal was finalised. Confirming the date on the schedule each year takes a minute and closes off the one thing that quietly undoes years of continuity.

If we switch to a cheaper insurer, is our past work automatically covered?
No. The new insurer sets its own retroactive date, and it is not obliged to match your previous one. Unless preserving the date is made an explicit condition of the move and confirmed on the new schedule, the earliest years of your work can fall outside cover. A switch that looks like a saving on premium can carry a hidden cost in lost continuity, which is why the retroactive date belongs at the front of the conversation, not the end.

We are acquiring another practice. What happens to its professional indemnity?
Treat the target’s past work as a cover question in its own right. Your policy covers the entity named in it; it does not automatically answer for an acquired firm’s earlier work. That work needs either to be picked up by endorsement on the enlarged programme or run off separately — and who stands behind it, on what basis and for how long is a point to settle before completion, with your broker involved, rather than discover afterwards.

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.