Sizing your PI limit: a decision framework

Step 1: Establish the regulatory floor

Several professions have minimum terms set by a regulator or professional body, and those minimums frequently specify not just a figure but a basis, a maximum excess, and required run-off. Solicitors, accountants, architects, surveyors, financial advisers and others each sit under their own regime, and the rules change. The floor is a starting point, not an answer: minimum terms are designed to protect clients of the smallest firms in the profession, not to reflect the exposure of a firm doing large or complex work.

Write the floor down, with the source, and note the basis it is expressed on. If your regulator requires a limit “in the aggregate” and your schedule says each and every claim, you exceed the floor. If it is the other way round, you may not comply at all.

Step 2: Collect the contractual minimums

In most sectors the binding constraint is contractual rather than regulatory. Appointments, framework agreements, panel terms, procurement portals and public sector contracts all specify insurance requirements, and they are rarely consistent with one another.

Build a simple register: client, required limit, required basis, whether defence costs must be in addition, the period cover must be maintained after completion, and any requirement about the insurer’s rating or the territory. The highest figure in that register is a floor. The longest maintenance period is a run-off commitment you have already made, which feeds directly into the run-off decision later. Where a contract defines your liability cap by reference to your insurance, note that too — our page on breach of contract cover explains why that drafting can leave a gap.

Step 3: Model the worst realistic single loss

Now ignore the floors and ask an engineering question: what is the largest amount of harm this firm could cause with one piece of work?

The right proxy depends on the profession. For construction and engineering consultancies it is usually the value of the largest project you advise on, or the cost of remedying the element you are responsible for — which can exceed your fee by orders of magnitude and, in remediation cases, exceed the original construction cost. For advisers handling transactions it is the value of the transaction. For firms giving tax, valuation or actuarial advice it is the sum at risk in the advice, not the fee. For technology and outsourced service providers it is the client’s business interruption and data exposure.

Then add the parts people forget: the client’s costs, your own defence costs, interest over a long claim, and the cost of consequential losses your contract does not exclude.

Step 4: Test the aggregation exposure

Step 3 assumes one claim. The scenario that most often defeats a limit is the systemic one: a template, a calculation method, a standard clause, a design detail or a piece of advice repeated across dozens of clients, all of which turns out to be wrong.

Ask three questions. How many clients could be affected by one repeated error? Would the wording’s aggregation clause treat those as one claim or many? And is your limit written each and every claim or in the aggregate? A firm with an aggregate limit and a systemic exposure has the most fragile combination available; a firm with an each-and-every-claim limit whose wording aggregates broadly may be no better off. Our page on the limit basis works through the mechanics.

Step 5: Adjust for how defence costs are treated

If defence costs are inclusive within the limit, the limit is not what it appears to be. On a technically contested professional negligence claim, expert evidence, disclosure and a multi-day trial can absorb a serious share of a modest limit before a penny reaches the claimant.

The adjustment is simple in principle: on a costs-inclusive wording, the limit needed to fund a given settlement is the settlement plus a realistic estimate of defending it. If you cannot obtain costs in addition, buy the difference in limit rather than pretending the difference does not exist.

Step 6: Read your own claims history properly

Look at frequency and severity separately. Frequency tells you how likely a second claim is in one policy year, which is the argument for an each-and-every-claim basis. Severity tells you whether your largest matter to date came anywhere near the limit — and more usefully, how far the initial reserve moved between notification and resolution. Reserves on professional indemnity claims commonly develop upwards as the picture clarifies, so a claim that closed comfortably inside the limit is weaker evidence than it looks.

Also look at the tail. Professional indemnity is written on a claims-made basis, so the relevant question is not only what you do now but what you have done in the last decade that has not yet surfaced. See claims-made for why that matters.

Step 7: Decide, document, and revisit

Take the highest of the floors from steps 1 and 2, compare it with the modelled figure from steps 3 to 5, and choose. Where the two diverge sharply, the options are a higher primary limit, an excess layer sitting above it, or a project-specific policy for the one appointment that drives the number.

Then write down the reasoning in a paragraph and put it in the board or partnership file. That record does three things: it demonstrates the decision was taken deliberately, it gives next year’s renewal a baseline to argue from, and it protects the individuals who made the call. Revisit it whenever the firm takes on a materially larger project, enters a new sector, acquires a book of work, or signs a contract with a higher requirement than anything in the register.

Frequently asked questions

How do we know if our PI limit is too low?

Compare three numbers: the highest limit any of your contracts requires, your regulator’s minimum, and the cost of putting right the largest single piece of work you are responsible for — including the other side’s costs and your own defence costs. If the third number is above the limit you carry, the limit is doing less than the firm assumes.

Should the limit be based on our fees or on project value?

On exposure, not income. Fees describe what you earn; the claim describes what it costs to correct or compensate for the work. In construction, engineering and transactional work the gap between the two is routinely very large, which is why fee-based rules of thumb tend to understate the limit required.

Does an each-and-every-claim limit remove the need for a bigger limit?

No. It changes how the limit behaves across multiple claims in a year, but it does not enlarge the amount available to any single claim, and broad aggregation wording can pull several related matters into one claim sharing one limit. Basis and size are separate decisions that need making together.

How often should we review the limit?

At every renewal as a matter of routine, and immediately on any material change — a much larger project, a new service line, an acquisition, a move into a new sector, or a client contract demanding more than the register currently shows. A short documented review each year is far easier than reconstructing the reasoning during a claim.

This page is general insurance information about how UK professional indemnity policies are commonly structured. It is not legal advice, and it is not a statement of what any particular policy covers. If a claim, a circumstance or a contract term is 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.