The "Heavy" Market: Navigating AI, IPOs and the Venture Reality Check

The great divergence

Speaking to the Investing News Network after the Web Summit, Zheng characterized the current landscape not necessarily as healthy, but as heavy. She traced today’s split market back to the US Federal Reserve's tightening cycle in 2022, but pointed to the 2025 cohort of listings as an early signal of a potential reset.

“A lot of these companies that did go public went public at a lower valuation than their highest private valuation,” she said, adding that this effectively turned them into down-round IPOs. She described this as a healthier step than resisting markdowns, which resulted in a logjam of deal-making and exit activity.

That reluctance to reprice, she said, “was trapping them in golden handcuffs.”

At the same time, Zheng stressed that today’s venture landscape is far from normal — a handful of highly hyped, marquee names are capturing a disproportionate share of dollars and attention, whereas in a typical market activity would be far less concentrated. For example, data from PitchBook shows that SpaceX’s acquisition of xAI for US$250 billion accounted for 72 percent of US$347.30 billion in exit value in Q1.

Valuations in Series A and B stages are also highly elevated, largely driven by a so-called “AI premium,” but also the high costs of building and running AI infrastructure that force companies to raise larger rounds.

The critical question for 2026 is whether this dispersion in value is genuine or merely price inflation.

“In terms of whether genuine value is being created,” Zheng explained, “(that) will ultimately come down to how quickly AI can become profitable.” While some of the shift reflects ordinary price inflation, she believes the majority of the dispersion can be attributed to AI hype right now.

The IPO landscape: Mega names vs. the next cohort

Investors remain focused on SpaceX, as well as Anthropic and OpenAI’s upcoming “mega IPOs." With that in mind, the market’s performance upon their public listings will be key to watch.

Zheng outlined two potential paths forward: successful debuts could validate the high AI valuations currently dominating the sector and trigger a wave of new exits and dealmaking. Conversely, poor performances could chill secondary trading and force a major re-evaluation of AI pricing across the board.

With so much late-stage liquidity now running through the secondary market, that fork in the road matters for retail investors. Many are accessing private companies through vehicles on platforms like Robinhood Markets (NASDAQ:HOOD) or Fundrise, which Zheng stressed are diversified baskets rather than single-name bets. She pointed to a “next cohort” of high‑conviction companies in secondary trading, such as Databricks, Stripe and Anduril.

“They’re not necessarily quite as big or quite as dominant in the secondary market right now, but they certainly are capturing a lot of activity. Time will tell where things will play out, but I think there’s just going to be a lot of enthusiasm for the next cohort … once we see these companies go public," the said.

At the same time, Zheng warned retail is still coming in “much later in the life cycle,” often when firms are over a decade old. These positions are less liquid and less transparent than traditional public assets, keeping risk elevated.

Zheng views the “democratization” of venture as promising, but still highly experimental, and the race to standardize this infrastructure is not without conflict. A recent patent infringement lawsuit filed by Nasdaq Private Market against Vancouver-based Hiive underscores the high stakes in controlling the plumbing of secondary trading.

Nasdaq Private Market alleges that Hiive’s process for converting non-standardized information, such as emails, into standardized exchange data infringes on its newly granted patent for share transfer settlement.

While Hiive has dismissed the claim as meritless, Zheng said the outcome of this legal battle may determine which entities and automated technologies will define the legal standards for venture secondaries moving forward.

For retail investors who rely on these platforms for liquidity, the case serves as a reminder that the systems facilitating their access are themselves subject to significant legal and regulatory headwinds.

For all the record-breaking headlines, today’s venture capital market appears to be less of a triumphant new regime and more of a fragile, AI‑heavy late cycle built on the promise that profitability will catch up — and it won't be long before that promise is tested by the coming wave of mega AI IPOs.

In that sense, 2026 is less about “democratization” as an accomplished fact and more as an open experiment. Whether this heavy, AI‑driven cycle hardens into a sustainable new normal or gives way to a painful reset will depend on how quickly AI businesses can turn cost into cash, and how the market chooses to price that transition in real time.