Pension Consultancy and Scheme Adviser Insurance

The thing to understand about pension consultancy liability is multiplication rather than severity.

An error in this work is rarely a one off. A benefit calculated on the wrong basis is calculated that way for everybody in that category. A misreading of a scheme rule applies to every member it covers. A data migration that corrupted a field affects every record with that field. A set of assumptions used in a funding valuation affects the scheme as a whole. So a single professional error produces a correction exercise across a membership, with recalculated benefits, arrears to some members, overpayments to recover from others, communications to all of them, and a trustee board that has to report it.

That has two consequences for insurance. The limit has to be set against scheme liabilities rather than fee income, because the correction cost scales with the membership and not with what you charged. And the aggregation wording matters more than usual: whether one error affecting two thousand members is treated as one claim or two thousand decides whether a limit is adequate or irrelevant. That is a clause worth reading rather than assuming, and it is where a broker earns the fee on this placement.

Pension administration is built on data that has passed through several hands over several decades, and a new adviser inherits all of it.

A scheme's records may have been kept on paper, migrated between systems more than once, maintained by a previous administrator with different conventions, and affected by a corporate history of mergers and transfers. Service histories are incomplete, salary records are missing, member statuses are wrong, equalisation has been applied inconsistently, and some records simply cannot be reconciled. From the day a consultancy is appointed, calculations are produced from that data and members are paid on the results.

So the position on inherited data has to be explicit rather than assumed. A data review at the point of appointment, with findings reported to the trustees in writing and a clear statement of what has not been verified, is the protection. Where the trustees decline a data cleanse on cost grounds, record it and restate it. And be careful about the scope of any engagement: taking on administration is not the same as warranting the accuracy of records you did not create, and the engagement letter should say so plainly.

The structural awkwardness here is that the person most likely to complain is not the client and has a forum of their own.

A member whose benefit is wrong, whose transfer was delayed, whose calculation they disagree with, or whose scheme made a decision they dislike can complain, and the route runs to the scheme's internal procedure and then to a pensions ombudsman. That body decides on the facts and on maladministration rather than on strict legal negligence, there is no trial, and a determination can direct a scheme to put a member in the position they should have been in plus an award for distress. The consultancy is examined in the process even though the complaint is against the scheme.

So the record and the responsiveness both matter. Calculations and the basis used retained per member rather than only the output, correspondence kept, decisions documented with the reasoning and who made them, and a sensible attitude to early complaints rather than a defensive one. Where a trustee board makes a decision against your advice, record the advice. A consultancy whose file shows what was advised, when, and on what information is in a very different position from one relying on recollection.

There is a cyber exposure in this profession that is specific rather than general, and the sums involved make it worth naming.

A pension administrator holds member identity data, bank details, and the authority to move substantial sums on a transfer or a retirement. That combination is exactly what an impersonation fraud needs: a request appearing to come from a member or a receiving scheme, with changed bank details, diverting a transfer value or a lump sum. The member loses retirement savings, the scheme has to decide whether to make them good, and the administrator's process is examined. Separately, member data at scale is a serious personal data exposure in its own right.

So the controls are ordinary and the discipline has to be absolute. Multi factor authentication on email and systems without exception, verification of any change of bank details by a route independent of the request, a rule that payment details are never changed on a written instruction alone, checks on receiving schemes before transferring, and staff who recognise the pattern. Then check the cyber wording, because funds transfer fraud is commonly sub-limited far below the headline figure and the sums here are large.

Aggregation is the clause that decides whether the limit works. We are an FCA regulated broker and will not tell you we are the best choice. These are the questions that decide it.