Manager dispersion in private markets is widening in 2026 because the standard diligence signals have degraded, not because allocators have grown careless. A low headline default rate in private credit now understates stress that has migrated into payment-in-kind conversions, and interim equity marks have detached from realized distributions. The diligence question has changed accordingly.
“Manager dispersion is not widening because allocators have grown careless. It is widening because the metrics that once separated a strong manager from a weak one have quietly stopped carrying the information they used to.”
The observation that manager selection matters more in private markets than in public ones is, by now, the least controversial statement an allocator can make. Nearly every major manager has published some version of it in the past year. The dispersion charts are real, the point is correct, and it has been made often enough that repeating it adds nothing to an investment committee’s thinking.
The more useful question is why the dispersion is widening now, in this particular cycle, rather than treating it as a permanent structural fact of the asset class. The answer running through both private credit and private equity in 2026 is uncomfortable, and it is not primarily about manager skill. The specific instruments allocators have long relied on to identify a deteriorating manager before the deterioration becomes expensive have lost much of their diagnostic power. A clean default rate in credit and a strong interim mark in equity no longer mean what a diligence process built over the last two decades assumes they mean. Dispersion is the visible symptom. The hollowing-out of the standard signal is the cause.
Private credit: the default rate has stopped doing its job
The headline private credit default rate remains low. Depending on the methodology, most 2026 estimates land somewhere between two and three percent, a figure that on its face suggests an asset class absorbing higher rates without meaningful distress. An allocator reading only that number would reasonably conclude the book is healthy.
The default rate itself is intact. What it measures has changed. Two structural shifts have moved credit stress out of the place the default rate is looking.
The first shift is the rise of payment-in-kind used as a pressure valve rather than a structuring choice. As of the fourth quarter of 2025, roughly 6.4% of private credit loans carried bad PIK, meaning interest deferred mid-loan because the borrower could not service it in cash, as distinct from PIK deliberately structured in at origination. That share is close to triple its 2021 level. Lincoln International treats this as a shadow default rate and puts implied distress nearer to 6% against the roughly 2% headline. A loan converting to PIK under strain is not counted as a default. It is a borrower that cannot pay, recorded as a borrower that has chosen to defer. The distinction the default rate draws between those two situations is exactly the distinction that no longer holds.
The second structural shift in private credit is the disappearance of the early-warning architecture itself. Approximately 70% of private credit issuance is no longer covenant-heavy in the way it was a decade ago, according to CAIA Association analysis published in April 2026. The covenants that once tripped and forced a lender and borrower to the table before a missed payment, giving the manager time to act and the allocator time to see it, are largely absent from the current vintage. Stress now surfaces later and more abruptly, because the mechanism designed to surface it early has been negotiated away. The Financial Stability Board’s May 2026 review of the sector reached a similar conclusion, identifying valuation opacity and the absence of harmonized, loan-level data as unresolved vulnerabilities rather than manageable features.
None of this describes a broadly impaired private credit market, and it is important to be precise about that. Underlying credit fundamentals across much of the market remain stable, and a meaningful portion of the anxiety in private credit this year has been sentiment rather than substance. The Blue Owl redemption episode, in which investors sought to withdraw a large share of certain technology-focused vehicles, was driven substantially by software-sector fear rather than by realized deterioration in the loan portfolios, which continued to perform in line with their benchmark. Both things are true at once: much of the fear is overstated, and the instruments allocators use to tell overstated fear from real deterioration are less reliable than they were. That combination is precisely what makes this a manager-selection problem rather than an asset-class problem. When the shared signal degrades, the distance between the manager who was underwriting conservatively all along and the one who was reaching for yield widens, and it widens invisibly until something forces it into view.
Private equity: the mark has detached from the cash
The parallel failure in private equity is not fundraising concentration, though that is real and widely reported. It is the same breakdown of the standard signal, expressed through a different metric.
For most of the last cycle, an allocator could read a manager’s interim IRR and unrealized marks as a reasonable proxy for eventual realized performance. That proxy has broken. Five-year rolling DPI for buyout funds reached its lowest recorded level in 2025. Distributions fell to roughly 6% of AUM in the year to mid-2025, against a ten-year average closer to 14%, per MSCI private capital data. Capital that a mark says exists is not being returned as cash, and the gap between reported value and realized value has grown wide enough that the two can no longer be treated as interchangeable.
This is why DPI, a metric that spent years as a secondary line on the scorecard, has become the dominant one in re-up decisions. DPI is the one number in the private equity toolkit that cannot be held up by a favorable mark, because it records only cash that has actually moved. In a market where marks have detached from realizations, allocators have rationally retreated to the single signal that still carries uncontaminated information, and they are pricing managers accordingly. The exit rebound of 2025 did not resolve this. Deal value recovered, but the large majority of it was concentrated in mega exits, leaving mid-market portfolio inventory largely stagnant and the distribution normalization structurally incomplete.
The private equity value-creation data tells the same story from the manager’s side. In the era of low rates and rising multiples, a mediocre manager and a strong one could post similar interim marks, because multiple expansion and leverage flattered everyone. That tailwind is gone. Among the managers still generating durable returns, an increasing share of those returns now comes from operational improvement rather than from financial engineering, a reversal of the prior decade’s composition. The managers who built the operating capability to grow earnings are separating from the managers who only ever knew how to buy well and wait, and the interim mark, once again, is the last place that separation shows up.
What this changes for the allocation committee
If the widening of dispersion were simply a matter of some managers being better than others, the response would be the familiar one: diligence harder, reference more thoroughly, weight the track record. That response is necessary but no longer sufficient, because it still leans on signals that have partially broken. The track record is a record of a rate environment that no longer exists. The interim mark is a claim about value that the distribution data contradicts. The clean default rate is measuring a category that PIK and the erosion of covenants have quietly emptied.
The more demanding question a committee should be asking of any private-market manager in 2026 is not what its returns have been, but which of its reported signals can still be trusted, and what it takes to see the deterioration the standard metrics now hide. In credit, that means looking past the default rate to the PIK trajectory, the covenant structure, and the manager’s willingness to mark honestly and early rather than to defer. In equity, it means privileging demonstrated distributions over interim marks, and interrogating the source of returns closely enough to distinguish operational value creation from the residue of a leverage-and-multiple era that will not return.
This is also where the distinction between alpha and beta in private markets now actually rests. It is not a question of access to the asset class, which is settled, nor of exposure to a return stream that a rising market once supplied to everyone holding it. It is a question of whether a manager’s results are attributable to capability that survives the withdrawal of favorable conditions, and of whether the allocator can establish that attribution using instruments that still work.
Access to private markets was settled some time ago. Exposure to both credit and equity is now available through more vehicles and more venues than most allocators can evaluate. What is genuinely scarce, and what this cycle has made scarce in a specific and diagnosable way, is the ability to read a manager accurately when the instruments the whole industry once read them by have lost their edge. That is the real content of “manager selection matters” in 2026. It is not an exhortation to choose carefully. It is a warning that the tools most allocators still choose with are not measuring what they think they are measuring, and that the managers who will look best in five years are not necessarily the ones who look best on the current scorecard.