What now?: The venture secondary market’s biggest names are leaving. This analyst note breaks down what happens when a market this concentrated loses its trophy companies, and which startups are positioned to take their place. Read it here.

Manager performance rankings: Our league tables allow LPs and GPs to cut through the noise of benchmarking returns with a standardized way to identify historically top-performing fund families. Explore them here.

Private markets have become a major component of institutional portfolios, yet most strategies have little performance history outside the relatively benign environment in which the industry matured. Modest inflation, low interest rates, and limited disruption leave allocators with little evidence of how today’s funds might respond to stresses public markets have experienced before.

Our latest analyst note applies the PitchBook Private Capital Return Barometers to three historical episodes: late-1970s stagflation, the Gulf War oil shock, and the dot-com bust. By feeding the conditions from those periods into the models, we estimate how modern private strategies might have performed in environments they never experienced. These historical analogs are not forecasts, but they show how recognizable stresses could transmit through today’s private-market structures.

The dot-com scenario offers a useful test because, unlike the other periods, we can compare the model’s implied VC returns with observed private-market performance. The model’s relationships are estimated using data dating back to 2009, reflecting the broader and more diversified VC universe of the modern private-market era. Dot-com-era portfolios, by contrast, were heavily concentrated in the companies experiencing the sharpest repricing.

The model implies a 29% peak-to-trough decline, compared with the 66% realized drawdown during the dot-com bust. The gap illustrates how modern portfolios might have responded to the same equity repricing differently from the narrower, highly concentrated vintages of the dot-com era.

For allocators, this highlights the question of portfolio concentration. AI accounted for 65% of US VC deal value in 2025, and a small number of megadeals now shape industrywide NAV. In a similar scenario, outcomes would depend on how funds are diversified across companies, sectors, and vintages. More diversified portfolios may behave like the modern universe represented by the model, while portfolios concentrated in the assets driving the current cycle could experience losses closer to those of a bubble vintage.

The full report examines what each historical stress period could imply across private-market strategies, for multiple scenarios.

Download the full report here.

Public enterprise SaaS investors began rewarding AI infrastructure companies again in Q2 2026, but most of the sector’s post-February valuation damage remained, according to our latest Enterprise SaaS Public Comp Sheet. Markets wanted measurable evidence that AI demand was translating into revenue before restoring the rest.

This marked a second phase of the “SaaS-pocalypse.”

Of the 97 companies in our comp set, 60 recorded higher enterprise value/trailing-12-month revenue multiples during the quarter. Even so, the median multiple declined 2.5% to 3.2x, and 74 of 97 companies finished below year-end 2025 levels. The average rose 10.5% to 5.9x, indicating that gains remained concentrated among higher-valued companies and a handful of sharp rebounds.

The rebound, however, was uneven, favoring infrastructure over application software.

The market’s preference was clearest among platforms benefiting directly from AI workload growth. Datadog and Cloudflare reported year-over-year revenue growth of 32% and 34%, respectively, while Snowflake delivered 34% product-revenue growth. Datadog’s multiple rose 115.8% quarter-over-quarter. DevOps, ITOps and developer/automation platforms was the only segment whose median multiple exceeded its year-end 2025 level, reaching 7.1x against 6.7x, and it leads estimated 2026 revenue growth at 21.9%.

Application software remained much less forgiving. Wix and HubSpot multiples fell 75% and 34.4%, respectively, while CRM, sales, marketing, customer experience and collaboration, productivity and creative have settled into single-digit estimated growth for 2026.

Profitability continues to provide support. The median estimated EBITDA margin is projected to reach 23.3% in 2026, up from 20% in 2025. Companies meeting or exceeding the Rule of 40 trade at a median 6.6x trailing-revenue multiple, compared with 2.3x for companies below the benchmark.

In our view, a broader recovery now depends on proof that AI products can drive incremental revenue without accelerating seat erosion. Infrastructure companies have begun to provide that evidence while application vendors remain uncertain.