Skip to content
KELTON PARTNERSINTELLIGENCE

Building control without duplicating your fund administrator

How to build selective, evidence-based administrator oversight without creating a second operating factory inside the manager.

Outsourcing fund administration does not remove the investment manager’s need for operational control. It changes where that control should sit.

The wrong response is to reproduce every administrator process internally. That creates two operational factories, two sets of records and a large reconciliation burden. The better response is an oversight model that gives the manager enough evidence, visibility and exception ownership to challenge outputs without rebuilding the administrator.

Start with accountability, not duplication

A fund administrator may prepare books and records, calculate NAV, maintain investor records and produce recurring reports. The manager still needs to understand whether agreed deliverables arrived, whether material movements are reasonable, whether exceptions were resolved and whether information provided to investors or management is supported.

The oversight model should therefore begin with four questions:

  1. What is the administrator contractually responsible for?
  2. What does the manager need to review or approve?
  3. Which evidence demonstrates that the review occurred?
  4. Which exceptions require escalation, and to whom?

This prevents the internal team from performing controls that do not change a decision or reduce a meaningful risk.

Define the control surface

An effective oversight process usually concentrates on a limited set of recurring outputs and high-impact events.

Data completeness

Confirm that expected files, reports and source data were received for the correct fund, period and version. Missing inputs should be visible before a downstream deadline is affected.

Reconciliation and breaks

Understand which cash, position, transaction and investor-record reconciliations the administrator performs, what tolerance applies and which aged breaks remain open. The manager does not necessarily need to repeat the reconciliation; it needs visibility into the result and the unresolved exceptions.

NAV and accounting review

Review material movements, unusual accruals, fee calculations, capital activity and other items that warrant management attention. The depth of review should reflect fund structure, frequency, asset complexity and the agreed responsibility model.

Deliverable control

Track the production, review and approval of NAV packs, financial information, investor communications and regulatory data. The purpose is to make deadlines, versions and authority explicit.

Service and issue management

Record service incidents, recurring causes, owner, ageing and remediation. A pattern of small operational failures can matter more than one isolated issue.

Use exceptions to focus attention

Oversight becomes expensive when every line item receives the same level of attention. A more scalable model uses thresholds and exceptions.

Examples include:

  • movement outside an agreed percentage or value tolerance;
  • a new or changed manual journal;
  • a reconciliation break older than a defined period;
  • missing pricing evidence for an agreed asset type;
  • a deliverable received after the service deadline;
  • a change in source file, methodology or sign-off owner;
  • an investor transaction that does not match the expected status.

The threshold is not a substitute for judgement. It is a way to direct judgement to the areas where it is most valuable.

Build one evidence trail

Email chains and uncontrolled spreadsheets make oversight difficult to demonstrate. A workable evidence trail should connect:

  • the period or event under review;
  • the source deliverables;
  • the checks performed;
  • the exceptions identified;
  • comments and supporting evidence;
  • reviewer and approval status;
  • final resolution and date.

This does not require a large platform replacement. A controlled workflow can often be introduced around the existing administrator portal, email and reporting files.

Separate provider performance from fund judgement

Two different questions are often mixed together:

  1. Did the administrator perform the agreed process correctly and on time?
  2. Does the manager agree with the resulting judgement, estimate or treatment?

The first belongs to service oversight. The second may require the investment manager, valuation specialist, auditor, legal adviser or another responsible party. Keeping the two questions separate makes escalation clearer.

A practical first implementation

A focused administrator-oversight engagement can begin with one recurring cycle, such as monthly NAV review:

  1. Inventory the administrator deliverables and deadlines.
  2. Define required checks, materiality thresholds and owners.
  3. Create one exception taxonomy and ageing logic.
  4. Link each check to the underlying evidence.
  5. Pilot the process for two or three cycles.
  6. Remove checks that add little value and strengthen those that expose real risk.
  7. Automate file intake, comparisons or report preparation only after the control logic is stable.

The objective is not to shadow every accounting entry. It is to know what happened, what requires attention and who is accountable.

The operating principle

The strongest oversight model is selective, evidence-based and compatible with the administrator’s process. It gives the investment manager control without creating a second administrator inside the firm.

Sources and further reading

  • SS&C on shadow administration and oversight: Source
  • Citco on common fund-services control areas: Source
  • Juniper Square on administrator oversight, unified data and exception-oriented operations: Source

This article is general information and is not legal, accounting, regulatory or investment advice.

CONTINUE READING

From fragmented files to a controlled operating data layer

Read next