Outsourcing decisions in fund administration get evaluated the way most vendor decisions do. Is it cheaper than doing it in-house. Does it free up a team that is already stretched. Can the provider absorb our volume at year-end without falling over. What do the service levels commit to, and what happens if they are missed.
These are reasonable questions. They are also incomplete ones, and the gap between them and the question that actually matters is where the risk sits.
A provider can lose the client. The fund administrator answers for the failure itself.
Cost and capacity are the wrong first questions
NAV production, investor servicing, fund accounting, transfer agency: each of these gets put out to a third party at some point, and the business case is almost always built the same way. Cost per unit of work, headcount released, coverage across time zones, an SLA with credits attached to it.
None of that is wrong. It is just a commercial frame applied to something that is only partly a commercial decision. The part it does not touch is what happens on the day the process fails, and who has to stand behind it when it does.
Follow a failure through to the end and the answer is always the same
A capital call notice goes out two days late. A KYC file is approved on documentation that should have been escalated. A NAV figure is published with an input that was never reconciled. In each case, walk the consequence forward and see where it lands.
Contacts the regulated entity on the filing. Delegation of an activity is not delegation of responsibility for it, and no supervisor has ever accepted an outsourcing contract as an explanation.
Holds the firm whose name is on the notice, the statement and the investor communication. They did not choose your provider and have no relationship with them.
Asks the administrator to evidence the control, the approval and the timeline - and expects that evidence to be produced now, not requested from a third party.
The contract and the SLA govern the commercial relationship between two firms. They allocate cost, remedy and, occasionally, blame. What they do not do is change who is accountable to the investor or the regulator. The provider's worst case is losing an account. Yours is a finding, a remediation programme, and an LP who now reads every statement twice.
Diligence asks what the provider will do, not what you will still have to prove
Outsourcing conversations are built around the provider's capability. Their process maps, their controls, their certifications, their onboarding plan. Almost none of it is framed around the question that arrives eighteen months later: when someone asks this firm exactly what happened on a specific case, on a specific date, who approved it and on what basis, can this firm answer without asking permission first?
If the only record of the work lives inside the provider's own systems, the honest answer is no. The evidence exists, but it belongs to someone else, and it arrives on their timetable, in their format, assembled by people whose incentive at that moment is not identical to yours. That is a firm which has taken on the accountability while handing away the means to discharge it.
If a regulator asked this afternoon what happened on one outsourced case last quarter, could you answer from your own records - or would you have to raise a ticket with your provider and wait?
Keep the governed record, wherever the work happens
The protection is not to stop outsourcing. For plenty of firms it remains the right operational call, and doing the work in-house badly is worse than having it done well elsewhere. The protection is to stop treating the provider's record-keeping as your record-keeping.
That means one governed view of the process that sits above the outsourcing relationship rather than inside it: the steps, the approvals, the exceptions and the timestamps held independently of whichever party performed the underlying task. The provider still does the work. The provider's people still complete steps, through a guest interface, against the same process definition your own team uses. But the evidence of what happened, when, and who signed off accrues to you as a natural by-product of the work, not as a report you have to request.
This is also what makes the relationship easier to manage in the ordinary case, not just the failure case. Exceptions surface where you can see them rather than in a monthly service review. Turnaround times are measured from your own data. A change of provider becomes a change of who completes a step, not a migration of institutional memory.
More than 300 fund specialists at Ocorian run global fund and investor operations on Next Matter, with maker-checker approvals and a timestamped record applied by default and held independently of whichever team or third party completes a step. Trade Republic runs bank-grade client operations at consumer scale, b2venture runs venture operations to roughly 800M EUR AUM, and Swan runs embedded-finance operations on the same pattern. Read the Ocorian case study
One of these transfers. The other never does.
Outsourcing the work is a legitimate choice, made every day by serious firms for good reasons. Outsourcing the accountability is not a choice at all, because it is not available. It stays with the name on the fund whatever the contract says, and the only real question is whether the firm holding it kept the means to answer for it.
Decide the outsourcing question on cost and capacity if you like. Just do not mistake that for having decided the governance one.
For the practical governance detail, see how to keep governance and audit control when outsourcing or insourcing fund operations. If you are weighing platforms alongside the provider decision, compare Next Matter and Caruso and Next Matter and Juniper Square.