The four limitations of private-capital data — and what closes them
Private-capital data has four structural limitations that make it hard to analyze: no standardized terms, no comparability across funds, sparse and lagging reporting, and opaque attribution of returns. They are named by academia, recognized by ILPA, and priced in by the secondaries market — and they share a single root cause.
TL;DR. Every one of these limitations traces back to the same thing: nobody models the documents. The terms that define a fund’s economics sit in prose, unmodeled and uncompared, so they can’t be standardized, benchmarked, reconstructed, or attributed. Model the documents and the four limitations close in turn. That is the layer DealAnalysis built.
The root cause
A private-capital document is written to be read by a lawyer, not modeled by an analyst. The fee, the waterfall, the commitments, and the conditions that change all three live in language. Until that language becomes structured, provable variables, none of the downstream problems — comparability, benchmarking, attribution — can be solved, because there is nothing consistent to compare, benchmark, or attribute against.
The four limitations, and what closes each
| The limitation | Why it persists | What closes it |
|---|---|---|
| No standardized terms | Every PPM uses different language for the same economics | Structured variables — every term mapped to one schema |
| No comparability | No benchmark exists across funds, vintages, strategies | A growing memory — percentile positioning on every term |
| Sparse, lagging data | Private companies report quarterly, partially, and late | Modeled reconstruction — partial inputs completed to a working model, each estimate labeled with its confidence |
| Opaque attribution | GPs claim alpha; LPs can’t verify skill versus market | PME attribution — skill, timing and market decomposed |
Why “model the documents” is the unlock
The four answers above are not four products. They are four views of one capability: terms turned into provable variables, benchmarked against a growing memory, with every figure labeled as observed, estimated, or computed. The standardization makes comparison possible; the comparison makes benchmarking possible; the modeling makes reconstruction and attribution possible. Solve the first and the rest follow.
“Virtually everything sold as a fact was not quite so.” — Ludovic Phalippou, the most-cited academic in private-equity performance and fee structures.
The fix is not louder claims — it is provable ones.
FAQ
Who says private-capital data has these limitations?
They are well established across academic research on PE performance, ILPA’s standardization work, and the pricing behavior of the secondaries market.
Does DealAnalysis invent data it doesn’t have?
No. Where inputs are partial, it reconstructs a working model and labels every estimate with its confidence — it never presents an estimate as a provable figure.
What is PME?
Public market equivalent — a method that decomposes a return into skill, timing and market, so an allocator can see what a manager actually contributed.