Your audit trail knows what happened. It doesn't know why.

Eighteen months after a portfolio company was marked down, an LP asks why. The audit trail answers in seconds: the valuation was edited on the fourth of March, by a named person, from forty-two million to thirty-eight million. Every field is intact and nothing has been altered since. Then the LP asks the second question — why thirty-eight and not forty, what changed the committee's mind — and the same record that answered instantly a moment ago has nothing to say.

What the audit trail was built to prove

An audit trail records state transitions. It captures who touched a figure, when they touched it, and what it was before and after, and it does this for everything, by default, without anyone deciding a particular change was worth remembering. That completeness is the point. Its job is to defend against a specific accusation — that a number was quietly changed, or that a record was edited after the fact to fit a more convenient story. Against that accusation it is a strong defense, and fund operations has spent years and real money building the infrastructure to produce it.

The trouble is that the accusation has shifted. LPs, auditors, and regulators are less often alleging that a fund tampered with its books and more often asking whether a defensible judgment sat behind a number in the first place. That is a question about reasoning, and reasoning is exactly the thing a log of state transitions was never designed to carry.

The question it can't answer

The record shows the mark moved. It does not hold the comparable company that was debated and set aside, the committee member who argued for holding the number flat, or the distinction between what the deal team knew that week and what only became clear a quarter later. All of that context lived where reasoning usually lives — in a meeting, a thread of replies, a call, or the memory of the two people who were closest to the position.

This is why the reasoning decays faster than the fact it explains. A number can almost always be reconstructed from source documents, because the documents were kept. The argument that produced it has no such backstop, so it fades on the ordinary schedule of human memory and staff turnover, and it is usually gone by the time anyone thinks to ask. When the person who made the call has moved on, the fund still has the number and has quietly lost the only account of why it is the right one.

The missing layer

Other fields have already run into this and named it. Software teams keep a running log of every change to their systems, and years ago many of them concluded the log was not enough — it told you what changed without telling you why anyone thought it was a good idea. The answer that stuck was a short, deliberate note written at the moment of the decision, recording the situation, the options that were weighed, and the reasoning for the one chosen. Some fields describe this as capturing the provenance of a decision rather than only its outcome.

Applied to a fund, that is a second kind of record sitting alongside the audit trail: a decision trail that holds why a mark moved, why an LP was excluded from a deal, why a side letter term was read the way it was. This is the record Orivade is built to keep — the reasoning attached to the number it explains — though the larger point stands on its own. Fund operations has built extensive machinery for proving what happened and almost none for preserving why, and the second record has to be created on purpose, because nothing about normal operations produces it automatically.

Two accusations, one record

The audit trail answers the accusation that something was hidden, and answers it well. It was never meant to answer the harder question that fund operations increasingly has to face, which is whether anyone can still explain the judgment behind a number once the people who formed it are gone. A fund can prove that every figure in its books is untouched and, in the same breath, be unable to defend a single decision that produced them. The record that would close that gap is a different record, and most funds have never kept it.