In public equity, the performance gap between an average manager and a strong one is real but bounded, because continuous price discovery corrects mistakes in real time. Private markets have no such mechanism. A manager who overpays or misjudges leverage can carry that mistake at book value for years, by which point the capital is committed, and the window to react has closed.
That claim is not novel, and presenting it as though it were would itself be an error. Nearly every major manager has published a version of it in the past year, which is a fair reason for an investment committee to discount yet another assertion of it. What follows is the sourced version: named institutions, dated figures, the contested academic literature, and the evidence that cuts against the claim. Our companion piece, on why manager dispersion is widening in 2026, takes the next step and examines why the spread is opening in this particular cycle.
How wide is the spread between managers?
Two funds can raise capital in the same year, buy similar companies into the same valuations, and sell into the same exit window. One returns more than 35 percent. The other returns less than 10. That is the 2016 US private equity vintage measured top decile against bottom decile, a dispersion of more than 25 percentage points that holds across most vintages (1). Neither outcome was determined by the asset class, because the asset class was identical. The vintage was identical. What differed was the manager.
That spread has a name: “manager dispersion”.
Manager dispersion is the difference in net return between the best- and worst-performing funds raised in the same asset class and vintage year. In public equity, it is measured in fractions of a percentage point. In private equity, it is measured in tens of percentage points, and that difference in scale drives the argument that follows.
Narrowing the comparison from deciles to quartiles compresses the spread without eliminating it. On average, top- and bottom-quartile private equity managers show a 21-percentage-point performance differential (1.1). Another measure places the top-to-bottom-quartile spread at approximately 14.3 percentage points in private equity, compared with 5.1 points in private credit and 6.7 in infrastructure (1.2). At the vintage level, the spread widened to nearly 14 percentage points for the 2021 buyout vintage, the largest in a decade (1.3). The figures differ because their scopes differ, not because one is wrong: the honest summary is a range of roughly 14 to 25 percentage points of net IRR, depending on how the question is framed.
For scale, consider the public equity equivalent. Over the 10-year period, the threshold for top-quartile active large-cap managers trailed the S&P 500 by 0.5 percent, and 79 percent of active large-cap US equity funds underperformed the index in 2025, the fourth-worst year for those managers in the scorecard’s 25-year history (1.4). Public managers cluster tightly, and they cluster near the benchmark. An advisory firm that selects a poor public equity manager gives up a portion of an index return. That is a real cost, and it is a bounded one. One disclosure belongs alongside the comparison: private equity dispersion is measured on money-weighted IRR and public dispersion on time-weighted returns, so the contrast is directionally valid rather than identical-metric.
Private markets are not bounded in the same way, and venture capital sits furthest out. Venture showed the widest dispersion among the alternative asset classes reviewed: a median fund IRR of 8.2 percent, a top quartile between 17.0 percent and 79.0 percent, and a bottom quartile between 0.3 percent and negative 22.4 percent (1.5). The bottom-quartile figure warrants attention. It represents underperformance against a benchmark but ss of capital.
That is the practical difference. In public equity, manager selection determines how much of an index return a client captures. In private markets, it can determine whether the client captures a return at all, which is why the two questions that follow are: why the spread has not been competed away, and what an advisory firm can do about it.
Why does the market not correct it?
The absence of continuous price discovery is the structural reason dispersion persists rather than compresses. A public manager who overpays learns the market’s verdict within days. A private manager who overpays holds the position at a carrying value set by a valuation policy, not by a bid, and the error may not surface until an exit is attempted years later, when the capital is drawn, and the options are a discounted secondary sale or waiting.
Two explanations operate at different timescales, and both hold. The absence of price discovery is the permanent condition that allows dispersion to exist. What varies is how visible a deteriorating manager is in the current cycle. Several signals diligence was built around have lost diagnostic power, as our companion piece on manager dispersion in 2026 sets out. The structural argument explains why the spread can exist; the cyclical one explains why it is opening now.
“Diversification across vintage years and strategies reduces timing risk, but it does not protect against consistent selection of mediocre managers. In private markets, manager selection is not a secondary consideration. It is the primary one.”
— BIP Capital, Private Markets Primer, Advisor Edition 2026, Section IV
Track record is a weaker guide to the next fund than conventional wisdom assumes, at least in buyout. Venture capital performance remains persistent across funds raised by the same general partner, whereas buyout performance persistence has become noticeably weaker over time. Measured on only the information available to investors at the time of fundraising, there is little or no evidence of persistence for buyouts overall or post-2000. For post-2000 buyouts, the conventional wisdom to invest in previously top-quartile funds does not hold (2). Where modest buyout persistence remains, it is driven by bottom-quartile rather than top-quartile performance, meaning the reliable signal is which managers keep underperforming.
The older and more optimistic finding is superseded rather than contradictory. Returns were once found to persist strongly across a partnership’s successive funds (2.1), but that dataset was later shown to be flawed, which is why the 2023 study revisited the question.
The practical implication of that finding is close to even odds. Among North American private equity funds of funds raised between 2009 and 2018, a fund following a top-quartile predecessor landed above the median 49 percent of the time, while a fund following a bottom-quartile predecessor landed above the median 48 percent of the time (2.3). Committing to a manager whose prior fund was bottom quartile would have carried roughly the same odds of beating the median as committing to one whose prior fund was top quartile.
A fair reading of the evidence includes the skeptical case. Private equity dispersion of roughly 20.7 percent is statistically indistinguishable from that of a simulated portfolio of levered, cheap micro-caps at roughly 20.4 percent, implying that much of the observed spread reflects portfolio construction, concentration and leverage rather than manager skill alone (2.4). That argument does not remove the dispersion. It reframes what an advisor underwrites when selecting a manager, which is a combination of process, structure, and risk posture rather than a stable personal talent.
Does diversification solve manager risk?
Diversification narrows the range of outcomes without eliminating the selection problem, and it costs upside. Private equity funds of funds have historically shown lower return dispersion than the individual buyout, growth equity, and venture funds they typically hold, so a fund of funds may reduce dispersion risk relative to a single manager, but it also appears to offer less potential upside (3). That is the trade: fewer catastrophic outcomes, fewer exceptional ones, and an additional layer of fees between the client and the underlying assets.
The number of managers required is higher than most advisory practices expect. For programs with a three-year commitment period, the optimal ranges are 25 to 30 funds for pure buyout and 40 to 45 for venture (3.1), while three or more positions are enough to smooth cash flows (3.2). Those answer different questions: three positions smooth capital calls, and twenty-five to thirty is the count required to approach the asset-class return rather than one manager’s.
Selection also matters more than vintage timing, though not because vintage is unimportant. Skipping commitments to the three lowest-performing vintage years in buyout and growth equity across 2000 to 2020 would not have significantly improved performance; improving outcomes requires identifying top-quartile managers (3.3). Manager selection in private equity matters far more than vintage selection, and it has the added virtue of being a feasible exercise (3.4). Feasibility is the distinction. Vintage conditions demonstrably affect outcomes; the difficulty is that no allocator has reliably identified good vintages in advance, while manager diligence is work an advisor can perform.
What does an advisory firm evaluate before committing?
Manager evaluation in private markets is a process assessment more than a returns assessment, because the returns data available at commitment is the least reliable part of the record. A recent fund’s reported performance is dominated by unrealized marks, and the J-curve means early numbers understate outcomes for structural rather than diagnostic reasons. Most funds take at least six years to settle into a final quartile ranking (4), so the fund an advisor is most focused on is the one whose quartile is least knowable.
Four risks govern that assessment: manager selection, liquidity, vintage year, and complexity or transparency. This article focuses on manager selection, the only factor an advisor can materially influence after the allocation decision is made. BIP Capital’s Private Markets Primer sets out all four in a full table with an advisor-considerations column, alongside the four evaluation questions that follow from them.
Two diligence items deserve specific attention given the dispersion evidence above. The first is loss ratio, the proportion of investments or invested capital carrying a total-value-to-paid-in multiple below 1.0x, calculable on either a count or a capital-weighted basis (4.1). A manager's loss ratio, and the basis on which it is calculated, is a more revealing figure than a top-line IRR. The second is realization: distributions to paid-in capital tell an advisor what has been returned, while total value to paid-in capital includes marks the manager itself set.
Those limitations are permanent features of how private funds report, and they’re why evaluation should weigh process over reported returns. They have been compounded in the current cycle by another problem: several of the metrics allocators have long read as diagnostics have quietly stopped carrying the information they once did. The two belong together in any diligence checklist, since the questions in our companion piece on 2026 manager dispersion are the ones that expose deterioration the standard metrics now hide.