Some of the most coveted AI companies may not reach public markets for years. Companies are staying private far longer than they once did, and secondary markets have changed who can participate and when. Early investors can now sell before a listing, and new investors can buy in without waiting for one. AIR Asset Management and Rainmaker Securities examine what that structural shift demands of managers and the advisors allocating on behalf of clients.
The shift is structural, and the data reflects it. Companies are averaging 12–14 years before going public today, against roughly four years in the 1990s, per the Jay Ritter/University of Florida IPO database and Apollo. By the time a private AI company draws institutional interest, much of its early success is already known. The investor arriving through the secondary market may be buying from someone who got in years earlier. That compressed upside is one reason AIR Asset Management concentrates on companies closer to the end of their financing arc, where more is known about the business and another liquidity event may be nearer. The tradeoff is less room for error on valuation.
AIR focuses on businesses that have already demonstrated product traction, attracted substantial capital, and built strategic relationships that newer competitors may find difficult to replicate. The fund holds 10 positions across foundational models, infrastructure and data, and applications. Selection favors proven operators. "The companies we identify as allocations for our fund are leaders in their verticals," says Stephen Luongo of AIR Asset Management. "Generally speaking, getting behind leaders tends to result in better outcomes over time." Within the same vertical, valuation divergence has made the case for multiple positions. "It does make sense to hold two different names, even if they're in that same vertical," Luongo says.
Secondary markets have made private shares tradable long before an IPO. Estimated global secondary-market volume reached $220–$240 billion in 2025, a gain of more than 40 percent year over year, per William Blair, Evercore, and Jefferies, with volume projected to exceed $250 billion in 2026. AIR sources positions through Rainmaker Securities. Glen Anderson, co-founder, CEO, and president of Rainmaker Securities, says the firm carries between $2 billion and $3 billion of private share liquidity at any given moment, with comparable demand on the other side. When Anderson founded Rainmaker, only a handful of private names had any tradable liquidity at all. "Now there are 1,000 unicorns that are traded," he says. Advisors can verify Rainmaker's FINRA registration at www.adviserinfo.sec.gov. "We have visibility into two-way price flow that we think is better than others," Luongo says.
A position can go back into the same private secondary market it came from once AIR believes the investment case has played out. Anderson says the fund is not built around a listing or an acquisition. "If there's an M&A event, fantastic. If there's an IPO, fantastic. But that's not what we're waiting for. We're waiting for the right time, where we've built the value to exit that particular position," he says. Secondary liquidity can widen the exit options, but a buyer still has to be found at a price AIR will accept.
The economics of the three layers have already started to separate. Infrastructure valuations have held firm against sustained demand while equity values elsewhere have moved differently. "You have a real-life experience where some of these verticals hold up versus another," Luongo says. Holding positions across all three layers is less a diversification thesis than a recognition that the sector's internal dynamics already require separate positions. The fund's 10 holdings span each layer, giving AIR direct exposure to how the splits develop rather than concentrating on a single part of the AI stack.
Leadership can be fleeting in AI, and private positions cannot be repositioned as easily as listed stocks. Beleutz places weight on what the largest companies have already accumulated: early-mover status, deep capital pools, and strategic relationships with Meta, Google, Amazon, and Nvidia. "In many cases, just implementing cross-business opportunities with their cornerstone investors can create tens if not hundreds of billions of valuation," Beleutz says. Scale can also change the competitive equation in a direction that favors incumbents. "Today's market leaders are likely to become tomorrow's consolidators," Beleutz says. "Rather than being displaced, these competitors typically acquire, integrate, and scale the up-and-comers," he says. He points to Facebook's acquisition of Instagram as an illustration of what a well-capitalized incumbent can do when a challenger emerges.
Public AI plays are crowded. The hyperscalers have been bid up. The chip names have been bid up. "It's the private markets. That's where the money hasn't caught up with the opportunities," Anderson says. PitchBook's Q1 2026 data shows AI companies make up approximately 37 percent of unicorns by count but nearly 48 percent of total unicorn value, roughly $4 trillion of the sector's $8.5 trillion valuation. AIR Nova Fund LP is AIR Asset Management's vehicle for this allocation. It holds 10 positions across foundational models, infrastructure and data, and applications, sourced through Rainmaker's secondary market. For advisors, entry price, the durability of competitive advantages, and the eventual route to liquidity are bound together from the beginning. Further analysis on private market strategies is available through InvestmentNews premium reports.