Short research notes on the recurring decisions institutional limited partners actually face — written for investment committees, not for the conference circuit.
A named research column from PrivateMetrics. Each entry takes one recurring LP decision — hold-or-sell, what attribution actually measures, when private equity is prudent in a retirement plan, what fiduciary-defensible GP research looks like — and works it through with the discipline an allocator already lives with. Short by design; the full papers sit behind them.
Hold everything about a direct-lending fund fixed except the line that says what the management fee is charged on. The headline fee does not change. The LP's net IRR moves by 120 basis points and the multiple by five points, because in the early years a fee on committed capital is levied on money that is not yet in a loan, exactly where the J-curve is deepest.
Hold the number of defaults and the recovery rate fixed, so the dollar loss is identical, and move only where in the loan book the defaults fall. The final multiple does not change. IRR moves roughly 40 to 60 basis points. Distributions received by the end of the investment period fall by a fifth. A fund whose defaults come late is indistinguishable from a fund with no defaults for four years. And inside the waterfall's catch-up band, the GP captures the entire timing benefit.
The question an auditor, a board member, or a plaintiff's counsel eventually asks is not whether the NAV is right. It is how the NAV got into the report. An LP whose figures travel one controlled path from GP document to committed record can answer that question in one click. An LP whose figures were re-keyed cannot.
Evergreen redemption gates were designed for stress, and in the second quarter of 2026 they met it. The limits functioned. The lesson is not that semi-liquid vehicles failed; it is that an LP holding one owns a redemption schedule, and should model it the way it models a closed-end fund's drawdown.
Most software-comparison advice lists features. The better question for an institutional LP is which decisions the tool actually supports across the investment lifecycle — and where the seams fall between one tool and the next. Here is a framework for running that evaluation.
When competent researchers using overlapping data disagree about the sign of buyout alpha by six points, no single manager's track record can be reliably risk-adjusted. The decision is then made under belief — so the honest move is to make the belief explicit, anchor it in peer evidence, and document it as a falsifiable claim.
The quarter-end board report is where an LP operations team's month goes: re-keying GP figures, rebuilding the deck, and answering 'which NAV is right?' The fix isn't a faster spreadsheet — it's reporting that reads from one reconciled source and drafts itself.
In a single private-credit fund, roughly a third of committed capital risks sitting idle. Run the same capital as a long-horizon evergreen program and model it loan by loan, and the finding sharpens: the return given up to idle capital — and recovered by managing it actively — grows as the investment horizon lengthens.
The evergreen structure quietly rewrites the LP's problem: no fixed life, no J-curve to pace around, liquidity by policy rather than by wind-down. That removes real friction — and introduces new questions about NAV, redemption mechanics, and what 'diversified by vintage' even means when there are no vintages.
Best-practice guidance for LPs usually arrives as a list of tips. This guide organizes the discipline around the actual lifecycle an LP program runs — source, evaluate, decide, monitor, report, govern — and states, at each stage, the practice that separates a defensible program from a fragile one.
A point forecast of a private-market cash flow is not identifiable from the available data — careful analysts reach materially different answers, and the data don't adjudicate. A more honest frame evaluates the decision across the full range of beliefs an LP could reasonably hold, and surfaces when the choice is genuinely close.
The 401(k) rule is the news; the discipline underneath it is the point — and it is not unique to defined-contribution plans. A process-based safe harbor only protects the fiduciary who can actually run the process.
Before asking how to attribute private-markets performance, ask what the attribution is of. An LP commitment fixes a strategy, a geography, a vintage, a manager, a size, and a pacing position all at once — and an attribution is only meaningful along the dimensions the process actually decided on.
Pacing is not a scheduling detail — it is a first-order policy choice on a profit–liquidity frontier, with a material, compounding cost of getting it wrong that an IRR comparison cannot see.
Commit exactly at target and you underinvest by design. A look at the four commitment policies LPs actually use — and the one that wins without needing any forecast at all.
The ILPA Performance Template lets LPs recompute performance under their own assumptions. Once you can compute anything, the discipline is matching each metric to the question — and knowing what each one cannot see.
Why GP research is process evidence, what fiduciary-grade source taxonomy actually requires, and how to evaluate whether an AI-augmented research platform meets the standard a regulator or plaintiff's counsel will apply two years after the decision is made.
One finding, one chart and one question for LPs, every three weeks, the day the article goes live. Confirm by e-mail; unsubscribe from any edition.
Double opt-in. Unsubscribe from any edition. Nothing else is sent to this address.