Private Markets: Beneath the Headline Performance
How AI can help investors separate signal from noise
Oct 3, 2026 · Jessica Elengical
Private markets risk often builds beneath the surface before it is reflected in headline performance metrics. A portfolio can appear stable even as the underlying risk profile is shifting across companies, funds, and strategies.
Consider a simple example. A fund ends Q1 with $100 million of cumulative contributions, $40 million of cumulative distributions and $100 million of NAV, producing a TVPI of 1.40x. By the end of Q2, cumulative distributions have increased to $45 million and NAV to $105 million, so TVPI has increased to 1.50x. On the surface, performance looks stronger. But break down the numbers and a different signal emerges.
The fund distributed only $5 million in Q2, down from $20 million in Q1. At the same time, a $25 million software investment that had been expected to generate liquidity this year remains marked at the same value. In the quarterly letter, however, the manager says the exit has been delayed as buyer appetite has weakened and software valuations remain uncertain.
The reported performance has improved, but the outlook for realizing that value has not. The questions now become: Why have distributions slowed? Which expected exits are being pushed out? Are unchanged marks still supported by company performance and market conditions? And is similar pressure appearing elsewhere in the portfolio?
This pattern is increasingly relevant in the current market, where distributions remain constrained, assets are being held for longer and exit conditions remain uneven across sectors. Yet those changes may not show up immediately in NAV or TVPI.
TVPI is not wrong. It is simply the starting point. Getting to the fuller picture requires investors to deconstruct performance, evaluate what is happening at the underlying companies, scrutinize the manager’s explanation and connect those findings across the portfolio. The challenge is not doing that for one fund, but doing it consistently across many funds and hundreds of portfolio companies, using thousands of manager reported documents, then bringing those signals together into a coherent view of the portfolio.
Most technology today solves only parts of the problem. Traditional platforms organize and report data. Newer AI tools can search across large document sets, answer portfolio questions and automate portions of portfolio review and monitoring. The harder challenge is connecting those capabilities into an analytical process that can trace reported performance back to its underlying drivers, test the explanation across multiple sources and determine whether the same pattern is emerging elsewhere.
Take the fund above. A summary tells you the manager delayed an exit. An investigative process connects that delay to the drop in distributions, checks whether the unchanged mark is still supported by the company's performance and market conditions, and flags whether other funds in the portfolio hold similar assets facing the same pressure.
AI can make portfolio-wide investigation possible at scale. As allocators bring more of their portfolio data together, AI can help them use it to investigate their holdings more deeply. Meaningful insights require validated, connected data that is traceable to source; private markets context to interpret what the data means; and the judgment to know where to look next. It also requires guardrails to recognize when the evidence is insufficient. When it comes to investment decisions, no answer is better than the wrong one.
Investors today face a challenging market, but much of the information they need to navigate these challenges already exists across cash flows and performance data, manager communications and broader market information. The opportunity is to connect those signals, interpret them in context, and surface changes in risk beneath the headline numbers.