What a Fund's Operations Look Like From the Outside.

A prospective investor's operational due diligence team sends its questionnaire and books a day on site. The review runs a few weeks and has an allocation behind it.

Its subject is how the fund actually operates: who does the work, how the money moves, where the controls sit. What sets it apart from an internal review is who is asking. The people reading the answers were never in the building, so the fund's material claims and controls have to be supported, not just asserted.

What the review looks at

The operational side of diligence covers the fund's operating environment: who holds the key roles and what happens when one of them leaves, how conflicts are handled, how valuations get struck and by whom, which outside firms touch the money, how compliance and controls and cybersecurity work in practice rather than on paper.

It sits inside a wider review. The same ILPA questionnaire that often frames the exercise runs to twenty sections and asks about strategy, terms, and track record too, and it calls itself a roadmap for further engagement rather than a verdict.

But the operational questions are their own exercise. This team does not stop at the written answers. It interviews the people named in them, calls references, seeks confirmation from the fund's administrator and auditor, and tests the explanations through walkthroughs and follow-up.

Working and showing are different jobs

Here is what catches funds off guard. A fund can run smoothly and still struggle to show how it runs.

A reconciliation performed every month, a wire that always draws a second approval, a valuation everyone understood at the time: each can be entirely real and still hard to lay in front of a stranger, because the proof is scattered across inboxes and spreadsheets, or lives mostly in the routine and the people who carry it out.

When someone asks to see it, what comes back is often incomplete, or spread across too many places. The process worked. It was simply never assembled so that anyone outside could follow it.

What credible corroboration looks like

So the real question a review puts to any given process is what, besides the fund's own account, backs it up. Several kinds of evidence can, and they work best together rather than ranked against each other.

The fund's answer on the questionnaire is where every line starts. On a material claim about a control, it is not where the line can safely end. A contemporaneous record carries more: a dated approval, a signed memo, a log kept as the work happened rather than reconstructed afterward.

Confirmation from an outside party adds something the fund cannot supply itself, though each speaks only to what it actually knows: a bank to the transactions it processed, an administrator to the records it keeps, an auditor to the procedures it performed. And a second person who can walk the process, tested by the reviewer's follow-up, shows the knowledge is not held by one person alone.

None of these is proof on its own. A record can be incomplete, or show the step done once without proving it happens every time; a walkthrough can be rehearsed; a confirmation can be narrow. What a review trusts is the overlap: the more sources that line up on one fact, and the more independent among them, the less the fund is asking to be taken on faith.

Corroboration is not the whole of what a review weighs; the review also forms views on conflicts, valuation, custody, and the design of the controls themselves. But corroboration is how it tests any of those. Whatever it is looking at, the question underneath is how much of it rests on more than the word of the people who run it.

And that confirmation is built long before the questionnaire arrives, in the ordinary choices about what gets recorded, who else is allowed to see it, and which outside parties can attest to it.

A fund that has made those choices can be checked against something other than its own word. A fund that hasn't is left asking to be believed, and on the operational questions, being believed does not always carry the review, or keep an allocation from slipping.