Why LPs are looking at DPI, not IRR

Why LPs are looking at DPI, not IRR

A 25% net IRR used to open doors. In 2026, it's just as likely to close them.

Here's the number that matters more to the LP reading your deck right now: DPI — distributions to paid-in capital. A fund showing a 25% unrealized net IRR with a 0.4x DPI reads, to most institutional allocators, as unrealized paper gains and not much else. A fund with an 18% net IRR and a 1.4x DPI reads as credibility. Same market, opposite reception.

Why the shift happened

LPs are liquidity-constrained. DPI across the industry has sat in the single digits as a percentage of NAV for several consecutive quarters. Allocators aren't just evaluating your fund anymore — they're evaluating whether private markets as an asset class are giving them cash back on any reasonable timeline. IRR can be inflated by subscription lines and early distributions, and LPs know it. DPI can't be engineered the same way. It's the plainest evidence a GP has that the strategy actually returns capital, not just marks it up.

This isn't a minor shift in emphasis. It changes what belongs on page one of your pitch.

What this means for your data room

If your deck leads with IRR and buries DPI in an appendix, you're answering a question LPs stopped asking as their primary one. The fix isn't complicated:

Lead with DPI and realized proceeds, fund by fund, not blended. Show the distribution timeline, not just the current multiple. If DPI is low because the fund is young, say so directly and pair it with realistic, deal-by-deal PME benchmarking — LPs want to see you understand the gap, not hope they won't ask.

Operational due diligence has become the silent disqualifier sitting underneath all of this. 87% of LPs have rejected a manager on ODD concerns alone, without ever telling the GP why. A fund with strong DPI but a data room that raises valuation-independence or key-person questions still won't close. The two workstreams — investment diligence and operational diligence — are evaluated in parallel, and either one can end the conversation quietly.

The compressed timeline makes this less forgiving

The ILPA DDQ 2.0 now runs 21 modules and 250-plus questions, and response windows have shrunk from roughly 14 days to 5. A GP who hasn't pre-built DPI-forward materials and a DDQ response library is negotiating against a clock that didn't used to exist. Fixing your data room narrative after an LP has already requested it is too late — the request itself is the highest-intent signal in the entire funnel, and a slow or defensive response reads as unpreparedness at the exact moment an LP is paying closest attention.

The takeaway

If you're heading into a raise, or already in one, pull up your current data room and ask a blunt question: does it lead with what we've returned, or what we hope to return? LPs already know the difference. The GPs who close faster in this market are the ones who stopped making them dig for it.

This is the first in a series on what LPs are actually evaluating in 2026 — drawn from PCD's work introducing fund managers to institutional and private wealth allocators across our network. If you want a second read on how your data room and positioning will land with LPs, get in touch.