Professional Indemnity for Insurance Brokers: The Cover Your Own Firm Is Required to Hold

A broking firm spends its days advising clients on liability exposures, and then holds its own professional indemnity cover at whatever number the renewal invoice happened to carry last year, usually the regulatory minimum, chosen once and never revisited. The firm that would never let a manufacturing client insure a plant at a value fixed years ago carries its own largest liability exposure at a floor set by a formula that has nothing to do with the size of the risk it actually runs.

Professional indemnity, the broker's errors and omissions cover, is the policy that responds when the firm's advice or its execution goes wrong and a client suffers a loss. It is not optional and not a discretionary purchase. Every IRDAI-registered broker is required to hold it as a condition of its licence, and the requirement is specific about the minimum limit and what the policy must cover. That specificity lulls firms into treating the requirement as the answer: hold the mandated cover, tick the box, move on.

The requirement is a floor, not an assessment. The claim that hits a broker is sized by the client's exposure, not by the broker's revenue, and a single placement error on a large risk can dwarf the statutory minimum by an order of magnitude. This post is about the broker's own PI: what the IRDAI regulations require, what the policy actually covers, how claims against brokers really arise, and why the firm that thinks about its own cover the way it thinks about a client's is the firm that will still be standing after its first serious error. The records that defend such a claim are a separate discipline; this is about the insurance that pays when the defence fails.

The obligation is not a general expectation but a specific licence condition under the IRDAI (Insurance Brokers) Regulations, 2018, which require every broker to take out and maintain professional indemnity insurance throughout the validity of its registration. The regulations do more than mandate that the cover exists; they prescribe its size and shape.

The headline requirement is the limit. The indemnity cover must be for a limit not less than three times the remuneration received or receivable in the preceding financial year, subject to a regulatory minimum floor that applies to a new broker or one whose three-times figure would fall below it. The three-times-remuneration basis is the important design feature to understand: it scales the minimum cover with the firm's income, on the reasoning that a larger firm placing more business runs a larger aggregate exposure. It scales with income, not with the size of the individual risks the firm places, which is exactly where the inadequacy discussed later comes from.

The regulations also constrain the policy's terms so that the mandated limit is real rather than nominal:

• The cover must respond to legal liability for breach of professional duty arising from a negligent act, error or omission.
• It must extend to the dishonest or fraudulent acts of the broker's employees, so that a rogue-employee loss is not simply uninsured.
• It must cover loss of documents and the associated costs.
• The policy must not carry a deductible or excess above the proportion the regulations permit, so the firm cannot satisfy the requirement with a policy whose self-insured retention swallows the cover.

A broker PI policy responds to the firm's civil liability to third parties, principally its clients, for failures in the conduct of its broking business. Reading the grant carefully matters, because the difference between a policy that pays and one that argues is in the wording, not the limit.

The core coverage heads a broker should expect and check for:

1. Negligent advice, error or omission. The central grant: liability arising from a breach of the duty of care the firm owes its clients, whether in advice, placement or servicing.
2. Placement errors. Cover for the specific execution failures that generate most broker claims: placing the wrong cover, failing to place or renew, arranging an inadequate limit or sum insured, or an error in the policy documentation.
3. Dishonesty of employees. A fidelity-style extension required by the regulations, covering loss caused by the dishonest or fraudulent acts of the firm's own staff, which is distinct from the firm's own dishonesty and covers the rogue-employee scenario.
4. Loss of documents. The cost of reconstituting client documents lost or damaged in the firm's care.
5. Defence costs. The legal cost of defending a claim, which on a contested professional-negligence allegation can be substantial in its own right, sometimes rivalling the claim.

Some policies add extensions worth having: cover for defamation arising from the firm's work, cover for the acts of appointed sub-agents or POSPs the firm is responsible for, and, importantly for a licensed intermediary, cover that responds to a regulatory investigation's defence costs even where any penalty itself is excluded. The firm should read its own policy against this list rather than assume the mandated minimum grant is the whole of what a good policy offers, because the market wordings differ, and the gaps are where a claim finds its way through.

Understanding the PI policy means understanding the events it answers, and broker claims are strikingly consistent in shape. They almost never come from exotic failures. They come from ordinary execution errors on ordinary accounts, discovered at a claim.

The recurring patterns:

• The missed renewal. A policy lapses because the renewal was not processed, the client suffers a loss in the gap, and the loss is uninsured. This is the cleanest and most indefensible broker claim, because the failure is a fact rather than a judgement, and it is the single most common cause of E&O loss.
• The wrong sum insured. The firm placed the cover at a value too low, the average clause reduced the claim, and the client says the broker should have advised the correct figure. Whether the firm is liable turns on what it advised and recorded, but the exposure is real.
• The placement error. The wrong cover placed, a line the client needed but was never offered, a warranty the client was not told to comply with, an exclusion never flagged that later declined a claim.
• The non-disclosure that voided the cover. The firm failed to pass a material fact to the insurer, the insurer avoided the policy at claim on the ground of non-disclosure, and the client, left with nothing, turns to the broker who handled the proposal.
• The delayed placement. Cover requested and not bound in time, a loss occurring in the interval, and the client uninsured for a period it believed it was covered.

Each of these is a routine operation done wrong once, and each can produce a claim many times the annual commission the account earned. The connection to a firm's claims and coverage discipline is direct: most broker E&O claims begin as a client's claim that was rejected or reduced, at which point the client looks for someone to make it whole, and the broker who arranged the cover is the obvious candidate. The PI policy exists precisely for the moment the firm's own error, rather than the insurer's stance, is the reason the client is out of pocket.

A PI limit is only worth what the exclusions leave intact, and the broker should know its policy's exclusions as well as it knows any client's. Several matter specifically for an intermediary.

• The firm's own dishonesty. PI covers the dishonesty of employees but not the deliberate dishonesty of the firm or its principals. A loss caused by the firm's own fraud is uninsured, which is why a separate fidelity or crime cover, and strong internal controls, matter alongside PI.
• Fines and regulatory penalties. A PI policy generally will not pay a regulatory penalty imposed on the firm. An IRDAI inspection that results in a monetary penalty is a cost the firm bears itself, which is a reason the inspection and penalty exposure deserves its own attention; a firm cannot insure its way out of a compliance failure. Some policies will fund the defence costs of the investigation even while excluding the penalty, and that distinction is worth checking.
• Prior known circumstances. A claim or circumstance the firm knew about before the policy incepted is excluded. This makes the honest completion of the proposal and the timely notification of circumstances a coverage issue, not just a formality.
• Insolvency of the insurer placed with. Some wordings exclude, or limit, the firm's liability where the underlying loss is that an insurer the broker placed with failed. A firm that places with lightly capitalised insurers should know how its PI responds if one cannot pay.
• Assumed or contractual liability. Liability the firm took on by contract beyond its common-law duty of care may be excluded, which matters where the firm signs client service agreements with expansive undertakings.

Broker PI is written on a claims-made basis, and the structural features that follow from that are where firms most often misunderstand their own cover.

Claims-made means the policy that responds is the one in force when the claim is made or the circumstance notified, not the one in force when the error occurred. An error made in 2024 and surfacing as a claim in 2026 is answered by the 2026 policy, provided the cover has run continuously and the retroactive date reaches back far enough. This has three consequences a firm must manage:

1. The retroactive date. The policy covers errors going back only to its retroactive date. A firm that switches insurers and accepts a later retroactive date creates a gap: errors made in the covered years but before the new retroactive date fall between policies. Preserve the retroactive date across renewals and insurer changes.
2. Continuity. Because the responding policy is the current one, a lapse in cover is catastrophic, not merely a gap in a period. An error made while insured but claimed during an uninsured interval is uninsured. The perpetual-licence regime removes the renewal prompt that used to prevent this, so continuity now depends on the firm's own diligence.
3. The deductible. The each-and-every-claim excess is the firm's own money on every claim, and the regulations cap how large it can be relative to the limit. A firm should size the deductible to what it can genuinely absorb, because a deductible set high to cut premium is a self-insured layer the firm has quietly taken on.

The feature that the perpetual-licence world makes newly important is run-off cover. When a firm ceases business, sells, or merges, its claims-made policy stops responding to new claims, but its liabilities do not end, because a client can bring a claim years after the firm stopped trading. Run-off cover keeps the policy answering claims made after the firm has wound down, for the tail of the exposure. A firm planning a sale, a succession, or a closure that does not arrange run-off leaves its principals personally exposed to claims on work done while trading. It is the one PI decision that is easy to forget precisely because it arises at the moment the firm has stopped thinking about operations.

The most important thing to understand about the mandated limit is that it measures the wrong thing. Three times remuneration scales with the firm's income. The claim that hits the firm scales with the client's exposure. These are different quantities, and on a firm placing large risks they diverge until the statutory minimum is close to irrelevant.

Consider the arithmetic. A firm earns commission of a few crore and holds the mandated three-times cover. It places a property programme for a client with a sum insured in the hundreds of crore. A placement error on that single account, an under-insurance, a missed cover, a voided policy, produces a client loss driven by the client's asset value, not by the commission the account paid. The claim can be a large multiple of the firm's entire PI limit, and the firm is uninsured above its floor for a liability it created in a single email.

So a firm should set its PI limit the way it would advise a client to set a sum insured: on the largest realistic loss, not on a formula.

1. Anchor to the largest client exposures. Look at the biggest sums insured and liability limits across the book, because those set the size of the worst placement error, and buy a limit that bears a sensible relationship to them rather than to revenue.
2. Consider aggregation. PI limits are usually annual aggregates, so a bad year with more than one claim can exhaust the limit. A firm with concentrated large exposures should think about whether one limit covers a year with two serious claims, not just one.
3. Re-assess as the book changes. Every move up-market, every larger client won, every new class with bigger limits, raises the size of the potential error. The limit set for last year's book underinsures this year's, and the perpetual-licence regime has removed the renewal that used to force the question.

The firm that treats its own PI as a compliance minimum has insured itself for the size of its income against a risk sized by its clients' assets. The firm that sizes it to the worst credible loss and reviews it as the exposure grows, the way it advises a client to, has bought the one policy that keeps a single ordinary error from ending the business.