From GP Reports to Portfolio Insights: The Path Every Number Should Travel

Every quarter, an LP's data arrives from outside. Capital account statements, capital call and distribution notices, quarterly reports, and, increasingly, ILPA-standardized templates come in from each general partner in that GP's own format. The operations team turns them into the figures the institution runs on: NAV, called and distributed capital, unfunded commitment, fees, and the multiples and rates built from them.

In most LP organizations that conversion is manual. Someone reads the statement and keys the figures into a spreadsheet or a system. The figures are usually right. The problem is not accuracy on the day; it is that a re-keyed figure has no path behind it. When a board member asks where a number came from, or an auditor asks to see the control that produced it, the answer is a person and a memory. This article is about the alternative: one path that every number travels, with the source attached at every step.

Why the path matters more than the number

A number can be correct and still indefensible. The fiduciary standard that governs institutional investing is a process standard: the question two years after a decision is not whether the outcome was good but whether the process was prudent and documented (Fiduciary Governance in Private Markets). The same logic applies to data. A NAV that entered the book through a controlled path, with the GP document linked and a reviewer's approval recorded, is evidence of a process. A NAV that was typed in is an assertion.

The practical consequences are familiar to any operations team. Two screens show two NAVs and no one can say which was keyed last. A distribution is recorded once in the cash ledger and once in the performance file, with different signs. A board pack goes out with a figure that was corrected in the source system after the deck was built. Each of these is a path problem, not an arithmetic problem, and no amount of checking after the fact fixes a path.

The four stages

A datum entering an LP's book should pass through four stages, in order, and should not be able to skip one.

1. Upload. The GP document, whatever its format, is stored as received and becomes the anchor for everything derived from it. Nothing is transcribed by hand. The document is the source of record, and every figure extracted later points back to it.

2. Process. The document is turned into structured data and validated before anyone looks at the figures. For ILPA-standardized templates, validation means the template's own arithmetic: the platform checks the four signed-convention reconciliation formulas, NAV roll-forward, commitment reconciliation, fee netting, and carried-interest accrual, to a USD 1 tolerance per fund, so a statement that does not compose to the penny is flagged here rather than in a board meeting. The reasoning behind that tolerance, and why it is a meaningful auditability property rather than a marketing figure, is worked through in ILPA Reconciliation: Signed-Convention Formulas for LP Auditability. For non-template documents, processing extracts the figures and applies the same reconciliation logic against what the book already holds.

3. Approve. A person reviews what processing produced, sees the exceptions, and decides. Nothing enters the book unreviewed. The approval is recorded with who, when, and what was overridden, so the control is visible later. This is the stage most spreadsheets skip entirely: the keying and the approving are the same act by the same person, which is the definition of an uncontrolled path.

4. Commit. The approved figures become the canonical record. Every screen, every report, every analytic reads that record; none recomputes its own version. The dashboard NAV and the board-pack NAV are the same number because they are the same row. Provenance runs the other way as well: from any figure in any report, one click reaches the GP document it came from.

What the path unlocks downstream

Once every number travels the same path, capabilities that are otherwise expensive become nearly free, because they are reading a clean, sourced, reconciled record rather than reconstructing one.

• Reconciliation becomes a by-product, not a project. The GP-reported figures and the LP's own calculated figures are compared at process time, at the same USD 1 tolerance, so discrepancies surface before commitment rather than before a board meeting.
• Anomalies are visible across the whole book. A fee that moved outside its pattern, a distribution that arrived without a matching notice, a NAV that moved against every peer in the same strategy: these are detectable only when every fund's data has the same shape and the same provenance.
• Reporting reads from one source. The six standard report types, and the composable templates behind them, draw on the committed record; the narrative drafted around the figures is grounded in those figures and reviewed before it leaves the building (Board-Ready Reporting for LPs).
• Forward views start from the truth. Commitment pacing, liquidity planning, and hold-or-sell analysis are only as good as the record of what has actually been called, distributed, and marked. A forward view built on a re-keyed book inherits every path error the book contains.

Why now: the ILPA 2026 transition

The Institutional Limited Partners Association's updated Reporting Template applies to funds still in their investment period in the first quarter of 2026 and to funds commencing operations from January 1, 2026, replacing the 2016 template on a go-forward basis. The new Performance Template applies on the same go-forward basis to funds commencing operations on or after January 1, 2026, with the first performance templates expected after four full fiscal quarters, in the first quarter of 2027. For funds no longer in their investment period as of the first quarter of 2026, GPs may continue to provide the 2016 template, so an LP's book will carry both formats for some time.

For an LP operations team this means more standardized inputs arriving, in more than one template version, across the same book. That is an opportunity and a hazard. Standardized inputs are far easier to process automatically and to reconcile arithmetically. But an LP that receives them into a manual process gains nothing; the template's own arithmetic is only a control if something checks it. The path described above is what turns the template from a better-formatted PDF into a validated record.

If you are evaluating

Ask any vendor to take one of your own GP statements, not a sample, and walk it through the four stages in front of you: upload it, show the validation exceptions it raises, approve it, and then click from a figure in a generated report back to the page of the statement it came from. A platform that can do that on your data has a path. One that shows the end result without the path is asking for the same trust a spreadsheet asks for. The buyer's guide sets out the wider framework for that evaluation.

The Design Partner Program is a selective deployment for institutions battle-testing the platform ahead of general availability. If your quarter still begins with re-keying, bring one quarter's statements and we'll walk them end to end.

Template timing per the Institutional Limited Partners Association's Reporting Template and Performance Template guidance, and BDO's implementation guidance of October 2025. The platform's ILPA reconciliation is aligned to the 2025 template, and formula coverage continues to expand.