The First 24 Months Now Decide the Deal
With multiple expansion gone and holds stretched to seven years, where the deal proves itself has shifted forward.
PitchBook’s Q1 2026 US PE Breakdown places the present moment in two figures worth holding side by side. The median US buyout multiple has reset from roughly 10.8x EV/EBITDA across 2016–2020 to about 13.0x in the post-2020 environment, and PitchBook’s authors put the consequence in plain language: “multiple expansion can no longer be relied upon to drive returns.” Bain’s 2026 Global Private Equity Report adds the second number — holding periods at exit have stretched to roughly seven years, up from five to six in the prior decade. Read together, those two shifts describe a structural change in where return now has to come from, and when it has to be visible.
The macro pressure traced in Dry Powder and the Pressure to Move — capital aging in a system that cannot recycle it — lands at the deal in this form. Higher entry multiples reduce the margin for error. Longer holds extend the time over which that error must be carried. What used to be absorbed across the life of the deal is now exposed earlier, and the slack a slow first year used to enjoy has narrowed in ways that don’t always announce themselves at the time.
The release valve no longer functions
For much of the prior cycle, multiple expansion functioned as a release valve. It did not eliminate execution risk, but it softened the consequences. A platform that underperformed early could still exit into a stronger pricing environment. A delayed integration could still benefit from valuation lift on the way out. Together those mechanisms created a structural asymmetry — early underperformance was often survivable, and late recovery remained possible — and most operating playbooks of the prior decade quietly assumed that asymmetry would hold.
That asymmetry has weakened. When entry multiples are already elevated and the cost of capital has moved against further expansion, the back end of the deal cannot be counted on to do the work the middle once did. What remains is operational performance — and the timing of when it becomes visible. The shift sounds incremental from the outside; from inside the platform, it changes where the burden of proof sits and when it has to be discharged.
Underwriting has moved to the floor
This shift is now explicit in how deals are described. In Grant Thornton’s sponsor commentary inside PitchBook’s 2025 Annual US PE Middle Market Report, sponsors describe a change of emphasis that maps directly onto the math: underwriting is anchored less on upside cases and more on downside resilience, and increasingly on whether a credible floor can be established early. The same commentary is direct on what this requires of the early hold: “Buyers are highly focused on getting the first two years right.”
This is not a change in ambition; it is a change in what must be proven, and when. Growth theses still matter. They simply no longer carry the deal alone. The floor — what the asset reliably produces under conditions that are not generous — has become the new underwriting frontier, and it has to be visible early enough to recalibrate the rest of the hold around it.
What the first 24 months now carry
The early period after close has always mattered. What has changed is how much of the outcome now depends on it. In the first 24 months, integration assumptions are tested under real conditions, leadership capacity is either sufficient or it isn’t, and sequencing choices begin to compound rather than reset. None of these dynamics is new. What is new is that they are no longer deferrable. The system reveals them earlier because the structure no longer absorbs delay, and the platforms that miss the reveal don’t get a quieter second year to make up the ground.
Cadence tightens — not as a choice, but as a consequence. Proof points are required sooner, not because investors demand them but because the underlying math demands them. What used to emerge gradually now separates quickly, and the separation is rarely loud enough to read as failure when it begins.
The compressed feedback loop
In prior environments, feedback was delayed. Performance issues could take years to surface fully. Integration strain could remain partially hidden inside growth numbers. Narratives could outrun operating reality for extended periods, and a platform’s external story often led its internal one by a year or more. That delay created room for interpretation, and most of the operating playbooks built in that period implicitly relied on it.
Today the feedback loop is shorter. Not because organizations have become more disciplined — but because the system is less forgiving. A weak first year is harder to offset. A mis-sequenced integration compounds earlier. A leadership gap becomes visible under load, not over time. The distinction between a working deal and a strained one forms sooner, and the early signs of strain — the ones that used to be debatable for a year — now resolve into evidence within months.
Early hold versus extended hold
Holding periods have lengthened, but not all extensions mean the same thing. Some reflect timing — waiting for markets to reopen, or for conditions to stabilize before exit. Others reflect structure — a deal that has not established a stable operating base early enough to support exit at all. The two look identical at the aggregate level, and they are routinely conflated in fund-level reporting.
Inside the first 24 months, the difference becomes clear. Whether integration has stabilized or continues to consume capacity, whether performance is compounding or being managed, whether optionality is expanding or narrowing — these distinctions resolve in the early window, not the late one. By the time a deal reaches year five or six, the path is usually set, and what looks like an extended hold for upside is often an extended hold for repair.
Playbooks from a different market
Many operating models in use today were shaped in a different environment. The 2014–2020 vintages operated with lower entry multiples, more reliable expansion, and shorter expected holds, and in that context a slow first year was not ideal but was often recoverable. That experience persists in how platforms are run. Integration pacing, performance expectations, and sequencing decisions still reflect a system that allowed time to compensate for early strain — and the operators who internalized those rhythms most successfully are sometimes the ones least equipped to read the new ones. The environment has changed faster than the playbooks built inside it.
Where the difference shows up
The separation between platforms is not always visible at exit. It shows up earlier, and most of the time it shows up in the same handful of places. The platforms I’ve watched closely tend to declare themselves by month 24 — whether integration has reduced complexity or merely redistributed it, whether leadership is deciding under load or reacting to it, whether the base business is stable or quietly carrying strain that hasn’t yet been named. From the outside, both paths can still look functional well past that point. From the inside, they are already diverging.
The shift in where risk lives
The change is not that deals have become riskier. It is that risk has moved. The risks that mattered most in the prior cycle — exit timing, market conditions, valuation expansion — have been displaced by risks that resolve earlier and inside the platform: early integration, sequencing under constraint, leadership capacity under load. This does not make outcomes more predictable. It makes them visible sooner, and that visibility is the change operators need to design around.
Most operating models I’ve seen carry an implicit assumption that the first year is a setup period — that the real work begins once integration stabilizes and leadership has settled in. That assumption was reasonable in the market the playbooks were built in. It is not reasonable now. The first 24 months are no longer the runway; they are most of the deal.
The question for 2026
The first 24 months do not determine the outcome by themselves. But they now determine whether the outcome remains open. What used to be resolved over the life of the deal is now decided earlier — not by design, but by the structure of the system itself. Multiple expansion was the back-end mechanism that gave operating teams time to learn. With that mechanism gone, learning has to compound forward from close, and the platforms that come out of this period strongest will be the ones whose first 24 months produced visible operating evidence rather than visible operating activity. 2026 is the year that distinction begins to surface in fund-level performance, and the data PitchBook and Bain are publishing this cycle will look obvious in retrospect — once the cohort that absorbed the shift early is visible, and the cohort that didn’t is no longer presenting at LP conferences.