Private markets have a peculiar relationship with evidence. Investors want proof before committing capital, but the asset class can take a decade or more to produce it. That tension was easier to manage when the universe of managers was smaller and strategies were more concentrated. It is harder today. The number of active fund managers grew from roughly 3,700 in 2007 to 13,000 by the end of 2022, while private markets simultaneously expanded beyond traditional buyout into growth, venture capital, private credit, infrastructure, secondaries, and other increasingly specialized strategies. The allocator's decision set has multiplied faster than the evidence available to evaluate it.

The conventional answer is track record. Wait for a manager to demonstrate performance, compare it with peers, and allocate once enough evidence exists to separate skill from luck. Private markets complicate that logic because fundraising and realization operate on different clocks. The average time between PE fundraises was 4.7 years in 2025, compared with an average portfolio company holding period of 6.6 years and a typical fund life of 10 years or more. By the time the evidence becomes sufficiently mature to provide comfort, the next allocation decision may already have passed. Wait even longer and the evidence becomes more complete, but potentially less representative of the team, strategy, competitive environment, and market regime being underwritten today.

That leaves investors with an uncomfortable question. How do you identify tomorrow's successful managers when yesterday's performance is incomplete, unavailable, or increasingly stale? The answer is not to abandon track record. It is to understand its limits. The empirical evidence in this report suggests that manager performance can look meaningfully more predictable with hindsight than it did when investors actually had to commit capital. Emerging managers push that problem to its logical conclusion: sometimes the institutional track record barely exists at all. Successful manager selection therefore requires taking calculated risk before certainty arrives, and understanding what the available evidence can actually tell us about the people and strategy responsible for the next fund.

For much of private markets history, access was the constraint. Investors competed to enter a relatively concentrated universe of established partnerships, and manager selection was largely a question of identifying which proven franchises deserved a larger allocation. That universe has changed dramatically. PitchBook counts roughly 3,700 active fund managers globally in 2007 and 13,000 by the end of 2022, equivalent to roughly 12% annual growth. The private markets industry did not simply accumulate more capital. It created thousands of additional organizations asking investors to distinguish genuine investment skill from an increasingly crowded field. 

That growth changes the nature of manager selection. An additional 9,300 active GPs means an additional 9,300 potential combinations of people, processes, incentives, investment philosophies, and institutional capabilities for allocators to evaluate. Yet the evidence available to make those distinctions does not expand at the same speed. A new GP cannot manufacture ten years of realized investment history simply because an LP wants it. Even established firms can change personnel, raise larger pools of capital, or pursue opportunities that differ from those responsible for their historical returns. The manager universe can scale quickly. The evidence required to separate durable skill from favorable circumstances cannot.

That creates the central problem of this report. Private markets ask investors to make decisions under an unusually long feedback loop. Waiting for certainty can mean waiting until the opportunity has changed, while investing early requires accepting that conventional evidence is incomplete. The relevant question is therefore no longer simply which manager has the strongest historical returns. It is what evidence can reasonably identify a team capable of producing future returns before the market has had enough time to prove it? The expansion from 3,700 to 13,000 managers makes that question increasingly difficult to avoid. And as the next section shows, the number of managers is only one dimension of the problem.

What the Data Tells Us

• The selection universe expanded roughly 3.5 times. The number of active managers increased by approximately 9,300 between 2007 and 2022. For LPs, that represents a fundamental expansion in the number of organizations competing for private capital. 

• Manager supply can grow faster than manager evidence. New firms can be formed and new funds can be marketed relatively quickly. A realized investment history cannot. That mismatch creates an information gap precisely where investors would prefer the greatest certainty.

• A longer track record is not automatically a more relevant track record. Historical returns describe investments made under previous financing conditions, valuations, competitive environments, and portfolio circumstances. The longer an investor waits for realized evidence, the greater the possibility that the conditions responsible for that evidence have changed.

• The allocator problem is therefore becoming one of signal extraction. With 13,000 active managers, identifying the strongest historical number is not enough. Investors need to determine which elements of a manager's past performance reflect capabilities that can reasonably survive into the next fund.

• And manager proliferation is only half of the equation. The same private markets industry has also expanded across a broader set of strategies. That means LPs increasingly need to determine not only whether a GP is skilled, but whether its skill is transferable to the specific strategy receiving the next dollar of capital.

The manager universe has not simply become larger. It has become more fragmented by strategy. In 2010, buyout represented roughly 75% of private equity AUM. By 2019, that share had fallen to 55%, while growth, venture capital, and other PE strategies expanded from 25% to 45%. At the same time, adjacent private market strategies were developing into substantial pools of capital of their own. Private debt AUM increased from $315 billion in 2010 to $845 billion in 2019, while infrastructure reached $1.1 trillion by 2022. PE secondaries grew from 1.8% of PE AUM in 2000 to 6.5% in 2023. 

That diversification complicates what a track record actually represents. A GP's historical returns are evidence of performance within a particular investment mandate, executed by a particular team, against a particular opportunity set. They are not necessarily evidence that the same organization possesses an advantage everywhere its brand travels. The capabilities required to source and transform a control buyout are different from those required to underwrite downside protection in private credit, manage physical assets in infrastructure, or price liquidity in secondaries. The logo may stay the same while the investment problem changes substantially.

For allocators, this creates a second dimension to the selection problem introduced in the previous section. The universe expanded from 3,700 to 13,000 active managers, but those managers also operate inside a market offering an increasingly broad range of ways to deploy capital. Manager selection can therefore no longer stop at the firm level. Investors need to ask which team generated the historical results, which capabilities produced them, and whether those capabilities actually map onto the strategy being considered. That makes strategy specific evidence increasingly important and makes the industry's long feedback loop considerably more inconvenient.

What the Data Tells Us

• Private equity became materially less concentrated in traditional buyout. Buyout's share of PE AUM declined by 20 percentage points between 2010 and 2019. Growth, VC, and other PE strategies moved in the opposite direction, rising to 45% of the market. 

• The expansion extends beyond equity. Private debt AUM increased by $530 billion between 2010 and 2019, or roughly 168%, while infrastructure moved from $639 billion to $1.1 trillion between 2020 and 2022. The allocator opportunity set is broadening both within PE and across private markets. 

• Specialization creates a transferability problem. A strong historical buyout record can provide evidence about a firm's judgment, culture, or ability to attract talent. It does not, by itself, demonstrate an equivalent edge in credit or infrastructure. Investors need to distinguish firm level quality from strategy specific capability.

• The relevant unit of analysis becomes smaller. Instead of asking only whether the GP is good, diligence increasingly needs to identify the actual people responsible for the prior record, their attributable decisions, their experience in the proposed strategy, and whether the investment process can be reproduced.

• Strategy expansion can make an established GP resemble an emerging manager. A firm may have decades of institutional history while the particular strategy being raised has little or no realized history. In that situation, the investor faces many of the same questions presented by a new firm: which evidence can substitute for a mature fund record?

• This creates an uncomfortable timing problem. Even when the manager and strategy do have a predecessor fund, private markets take years to reveal whether the underwriting was actually correct. The next allocation decision often arrives first.Headline

Track record is supposed to reduce uncertainty. Private markets have a structural problem with that logic: the investment decision often arrives before the evidence does. In 2025, the average time between PE fundraises was 4.7 years, while the average PE portfolio company holding period was 6.6 years and a typical PE fund can run for 10 years or more. The mismatch means LPs evaluating a successor fund may be looking at a predecessor portfolio whose underlying investments have not yet completed the average holding period, much less reached the end of the fund's life. 

The lag is not uniform across the market. Average holding periods in the chart range from 4.33 years in real estate to 6.96 years in energy and utilities, with financial and insurance services at 5.75 years, healthcare at 5.65 years, information technology at 5.15 years, and consumer at 6.28 years. That variation matters because the amount of evidence available to an allocator depends partly on what the manager owns. A portfolio concentrated in longer duration investments may still contain substantial unrealized value when the next fundraising cycle begins. The LP is therefore not always comparing finished investment records. It may be comparing a mixture of realized outcomes, interim marks, operating progress, and assumptions about exits that have yet to occur.

This creates a Catch 22 for manager selection. Waiting longer gives investors better evidence about what the GP actually delivered, but waiting also means making fewer decisions when access is available and evaluating performance generated further in the past. Investing earlier preserves access and allows an LP to identify talent before it becomes obvious, but requires greater reliance on incomplete evidence. Neither option eliminates uncertainty. It simply changes its form. And that raises the question that matters most for this report: how much predictive value does the track record visible at the moment of allocation actually contain?

What the Data Tells Us

• The fundraising clock runs faster than the investment clock. The average 4.7 year interval between fundraises is nearly two years shorter than the 6.6 year average portfolio company holding period. An LP can therefore face the next commitment decision before the average underlying investment has reached an exit. 

• A successor fund decision is not necessarily based on a realized predecessor fund. A typical fund life exceeds 10 years, more than twice the average fundraising interval shown here. That means the conventional sequence of invest, observe, evaluate, then recommit does not neatly describe how private markets actually operate.

• The evidence gap varies by strategy and sector. Energy and utilities show an average holding period of 6.96 years, compared with 4.33 years for real estate. Investors should therefore care about the maturity and realization profile of the actual portfolio rather than treating fund vintage alone as a measure of track record quality. 

• Unrealized performance is evidence, but it is not the same evidence. Interim valuations can contain useful information about portfolio development. They simply carry a different degree of certainty from cash that has actually been returned. A headline IRR without understanding the realized component can therefore create more precision than the underlying evidence warrants.

• Waiting for three mature funds does not completely solve the problem. By the time an investor observes multiple realized vehicles, years may have passed since the investments responsible for those results were originated. Team composition, fund size, financing conditions, competition, entry valuations, and even the strategy itself may have changed.

• The tradeoff is therefore between completeness and relevance. Early track records are more contemporaneous but less realized. Mature track records are more complete but potentially less representative of what the GP will encounter next.

• This is why performance persistence needs to be measured using information available at the time of the decision. Looking backward after a fund has fully matured can tell us whether good managers eventually revealed themselves. It cannot tell us whether an LP could have identified those managers when the successor fund was actually asking for capital.

That distinction is not theoretical. When we compare final predecessor performance with the performance investors could actually observe at fundraising, the predictive signal changes dramatically.

The previous section established the timing problem: LPs are often asked to commit to a successor fund before the predecessor portfolio has fully matured. That distinction matters because private equity performance looks considerably more informative once investors know how the story ends. For post 2000 buyout funds, managers whose predecessor funds ultimately finished in the first quartile generated successor funds with an average 1.27x PME, compared with just 0.96x for managers whose predecessors ultimately finished in the fourth quartile. Viewed from the end of the fund's life, performance persistence looks real and economically meaningful.

The problem is that an LP considering the successor fund did not have those final quartile rankings. Harris, Jenkinson, Kaplan, and Stucke address this by repeating the exercise using predecessor performance measured at the time the successor fund was being raised. The staircase nearly disappears. Successor fund PME averages 1.17x, 1.16x, 1.13x, and 1.18x for managers whose predecessor funds appeared to be in Q1 through Q4 respectively. The information available when capital actually had to be committed did little to separate the eventual winners from the losers.

This is not evidence that track records are useless or that manager skill does not exist. It makes a narrower and more consequential point: the predictive power of track record depends on when it is observed. Mature results can reveal meaningful differences between managers, but those differences may become visible only after the allocation decision has passed. For LPs, the challenge is therefore not simply finding historical winners. It is identifying the characteristics behind those winners before realized performance makes them obvious. That moves manager underwriting away from a pure ranking exercise and toward a search for earlier evidence of repeatable skill.

What the Data Tells Us

• With hindsight, prior performance separates managers remarkably well. When predecessor funds are classified using their eventual fund end results, successor PME falls from 1.27x for Q1 predecessors to 0.96x for Q4 predecessors. That is a 0.31x PME spread between the two groups.

• At the actual allocation point, that spread disappears. Using predecessor performance known at fundraising, the difference between the highest and lowest average successor PME across the four quartiles is only 0.05x. More importantly, there is no orderly relationship between predecessor quartile and subsequent performance.

• The Q1 result is particularly revealing. Among post 2000 managers whose predecessor fund appeared Q1 when the next fund was being raised, 24.0% of successor funds ultimately landed in Q1. Exactly 24.0% landed in Q4. A top quartile interim track record did not provide the clean signal that the label might suggest.

• Poor interim performance was not an obvious exclusion signal either. Managers whose predecessor funds appeared Q4 at fundraising subsequently produced an average 1.18x PME, slightly above the 1.17x generated by apparent Q1 managers. That does not mean investors should deliberately select Q4 managers. It means interim quartile ranking alone was a weak sorting mechanism in this sample.

• The distinction between realized and unrealized evidence becomes critical. A predecessor fund can migrate materially between quartiles as investments mature. An LP should therefore look beneath headline IRR or quartile rankings and ask how much performance has actually been realized, which assets are driving remaining NAV, and what assumptions are embedded in those valuations.

• Track record should be decomposed rather than discarded. Attribution becomes more valuable when aggregate fund performance is noisy. Which partners sourced the investments? Which decisions generated value? Were returns driven by operating improvement, leverage, multiple expansion, sector exposure, or a handful of exceptional outcomes? Those questions begin to isolate the mechanism behind the number.

• There is an important implication for access. If LPs wait until superior performance becomes unambiguous at fund end, they may be identifying successful managers only after other investors have reached the same conclusion. The manager may also have raised more capital, changed economics, expanded the team, or moved into a different opportunity set.

• The objective is therefore not certainty. It is better uncertainty. Private market investors cannot eliminate the information gap created by long fund lives. They can decide which forms of incomplete evidence deserve greater weight.

That brings the report to the group for whom this problem is impossible to avoid: emerging managers. With limited institutional history, there may be no mature fund record to lean on at all. Yet LPs continue to allocate meaningful capital to them across a surprisingly broad range of strategies.

If mature track records were the only credible basis for manager selection, emerging managers would struggle to raise institutional capital at all. Yet the market tells a different story. PitchBook data show emerging managers attracting capital across private equity, venture capital, real estate, real assets, private debt, funds of funds, and secondaries. Private equity represents 46.5% of emerging manager capital raised in the chart, followed by venture capital at 28.2%, with the remaining 25.3% distributed across other private market strategies. 

That matters because an emerging manager forces investors to confront the information problem directly. There may be no Fund III to analyze and no decade of institutional returns to rank. But "no fund track record" does not mean "no evidence." A new GP may consist of investors who spent years originating, underwriting, operating, or exiting assets elsewhere. The diligence task therefore shifts from evaluating a finished institutional record toward establishing attribution. Which investments did these individuals actually lead? Which decisions were theirs? Did their advantage come from sourcing, asset selection, operations, structuring, or simply participation in a successful platform? The less complete the fund level history, the more important it becomes to understand the evidence beneath it.

The breadth of emerging manager fundraising makes that exercise more difficult. The evidence that matters for a new buyout team is not necessarily the evidence that matters for a private credit or real estate strategy. And as Segment 2 established, private markets themselves have become more specialized. Emerging manager underwriting therefore sits at the intersection of the report's two central problems: limited historical evidence and increasingly strategy specific sources of return. The question is not whether investors should ignore track record and take more risk. It is whether they can replace false precision with a more deliberate assessment of the capabilities required to produce the proposed strategy.

What the Data Tells Us

• Emerging manager risk is not confined to venture capital. Nearly half of the capital in the chart, 46.5%, is associated with private equity managers. Venture represents another 28.2%, while the balance stretches across real estate, real assets, credit, funds of funds, and secondaries. 

• The chart demonstrates breadth, not outperformance. It would be a mistake to interpret these fundraising shares as evidence that emerging managers generate superior returns. The data establish that investors are allocating to emerging managers across strategies. Whether they are compensated for that risk requires separate performance evidence.

• A first fund is not necessarily a first investment. Fund sequence and investor experience are different variables. A team launching Fund I may bring years of attributable investment experience from another platform. Conversely, an established brand can launch a new strategy in which the relevant team has a much thinner record. The diligence process needs to distinguish institutional age from actual strategy experience.

• Attribution becomes one of the most valuable substitutes for fund history. If an emerging manager presents deals completed at a previous employer, LPs need to determine what role the individual actually played. Sourcing a transaction, sitting on the investment committee, leading operational work, and inheriting a successful asset are not equivalent evidence.

• Team continuity matters alongside individual experience. A collection of accomplished investors does not automatically constitute an accomplished investment team. LPs should ask how long the principals have worked together, how decisions were made historically, whether economics support retention, and who owns responsibility when an investment deteriorates.

• The appropriate evidence should change with the strategy. A buyout GP might need to demonstrate proprietary sourcing and repeatable operational value creation. A private credit manager needs evidence of underwriting discipline, documentation, monitoring, and workout capability. Real estate demands asset level operating and local market expertise. Energy introduces commodity exposure, technical knowledge, and operating risk. A universal manager score can obscure the capabilities that actually matter.

• Emerging managers also expose the limits of brand as a diligence shortcut. The earlier sections showed that observable interim fund performance can be a weak predictor. Emerging managers remove even that comfort. Investors are forced to examine the underlying investment engine rather than rely on a familiar franchise or quartile label.

• That may be the broader lesson for all manager selection. The diligence applied to emerging managers should not disappear once a GP becomes established. If the objective is to identify repeatable skill, LPs should ask many of the same questions of Fund I and Fund VIII: who generated the results, how were they generated, and why should the mechanism continue to work?

The report therefore arrives at a different conclusion from "wait for more track record." More history can improve the evidence, but it cannot remove the timing problem, strategy drift, attribution problem, or changing market environment. Investors still need a way to make decisions before certainty arrives.

The final question is the practical one: what exactly should an LP underwrite when the rearview mirror is not enough?

The Price of Waiting for Proof

The easiest manager to underwrite is the one whose success is already obvious. It may also be the manager whose opportunity has changed the most. By the time multiple funds are realized, the team may have expanded, fund size may have grown, competition may have intensified, and the market conditions responsible for earlier returns may no longer exist. The evidence becomes stronger precisely as its relevance can begin to decay. That is the central paradox running through this report.

The data do not make the case for ignoring track records. They make the case for being much more precise about what a track record actually proves. For post 2000 buyout funds, predecessor performance measured at fund end produced a clear relationship with successor performance: average successor PME ranged from 1.27x for Q1 predecessors to 0.96x for Q4 predecessors. But using only the predecessor performance investors could observe when the successor fund was actually being raised, that relationship largely disappeared, with successor PME ranging only from 1.13x to 1.18x across quartiles. Historical performance contained information. The problem was when that information became available.

That distinction becomes increasingly important in a market with 13,000 active managers, a broader strategy menu, and meaningful capital already flowing to emerging managers across PE, venture, real estate, real assets, private debt, and other strategies. The investment question therefore cannot end with what did this fund return? It has to move underneath the number: who produced the result, what decisions produced it, which capabilities were responsible, how much has actually been realized, and why should those capabilities remain relevant to the strategy being raised today?

Private market investing has never offered the luxury of perfect information. The more useful objective is not to eliminate uncertainty, but to decide which uncertainty is worth underwriting. Waiting for three realized funds may reduce one kind of risk, but it can introduce another: paying for a manager whose advantage has already been recognized, scaled, or altered. Investing earlier carries greater uncertainty, but it can also provide access before the evidence becomes consensus.

The best allocators will not be those who predict the future with certainty. They will be those who become better at identifying repeatable skill before the track record makes it obvious.

McKinsey. Global Private Markets Annual Review. https://www.mckinsey.com/~/media/mckinsey/industries/private%20equity%20and%20principal%20investors/our%20insights/mckinseys%20private%20markets%20annual%20review/2020/mckinsey-global-private-markets-review-2020-v4.pdf?utm_source=chatgpt.com

Science Direct. The persistence of buyouts in private equity. (LINK)

McKinsey. Private Capital Inisghts. https://www.mckinsey.com/industries/private-capital/our-insights/global-private-markets-report/private-equity?utm_source=chatgpt.com

Pitchbook. Private Equity Strategies. https://files.pitchbook.com/website/files/pdf/Q3_2023_Allocator_Solutions_Timing_Is_Everything.pdf?utm_source=chatgpt.com

S&P Global. PE buyouts record longer holding periods. https://www.spglobal.com/market-intelligence/en/news-insights/articles/2025/12/private-equity-buyouts-record-longer-holding-periods-in-2025-96348743