The next great technology winners may be built long before public markets can access them

Phil Cummins’ case for venture capital begins with a simple but powerful observation. Over the last two decades, technology has driven equity market value creation. Much of the most important growth now happens before companies reach public markets.

In that sense, venture capital is no longer a niche sleeve for adventurous allocators. It is becoming a more important way to access the part of the equity market where technological change is actually being incubated and scaled.

Cummins describes venture as the research and development engine of global markets, and that framing is useful. Many of the companies that have gone on to define public equity returns, from Apple and Microsoft to Google and Meta, were venture-backed businesses. But his point is that the structure of value creation has changed.

In earlier eras, a much larger share of the upside from transformational technology accrued in public markets. Today, more of that compounding is happening while companies are still private, and often at a scale that would once have seemed extraordinary.

Private markets are now capturing more of the technology upside

Cummins is keen to push back on the common shorthand that companies are merely “staying private longer”. In his view, that misses the more important point. The issue is not age, but scale.

Companies are now reaching sizes in private markets that would historically have belonged to listed businesses, and they are doing so much faster than previous generations of technology companies. That is where AI becomes central to his thesis.

Cummins argues that the market is in the middle of a genuine technology shift rather than a passing thematic, and one that may ultimately prove larger than previous waves of digital change.

If software “ate the world” in the 2010s, AI may do something broader, diffusing through industries in the way electricity once transformed production and economic activity. It is that scale of transformation that underpins his confidence in venture. The examples he offered are revealing. Companies such as OpenAI, Cursor and Lovable have been able to reach $100 million in revenue in little more than a year, in some cases less.

Compare that with the best SaaS businesses of the previous cycle, where reaching that milestone often took seven to ten years, and the magnitude of the shift becomes much clearer. These are not speculative stories built only on narrative. They are scaling with real revenue at a pace the market has rarely seen before.

AI is broadening venture beyond software

Another important strand in Cummins’ thesis is that venture itself is evolving. He says it is no longer accurate to think of the sector purely as software investing.

In fact, he argues that software is increasingly commoditised at a broad level, while the next layer of venture opportunity is spreading into more capital-intensive and strategically important parts of the economy. That includes energy, defence, semiconductors and space, areas that many investors might once have treated as peripheral to venture capital.

Cummins’ argument is that these sectors are becoming foundational parts of the broader AI stack. If AI is going to diffuse through the economy at scale, it will need infrastructure, chips, power, security and entirely new industrial capabilities. That means the venture opportunity set is becoming wider, deeper and in some ways more economically consequential than during the pure SaaS era.

This reinforces his view that the cycle is far from over, even with several large AI names already well known. On his reading, the market is still early. He likens today’s stage of AI development to the period before the smartphone’s release in the earlier technology arc. Much has happened, but the application layer and wider commercial diffusion are still to come.

“It’s not really about companies staying private longer, it’s about how big companies are scaling in the private markets.”

Access remains the whole game in venture

If the opportunity is large, Cummins is equally clear that venture remains brutally selective. This is still a power-law asset class, where a very small number of companies account for a very large share of returns. Over the last decade, he noted, just 1 per cent of exited venture-backed companies generated 47 per cent of the exit value. That concentration is the defining feature of the asset class.

For advisers and allocators, that means broad enthusiasm for venture is not enough. The challenge is getting exposure to the right companies, and that still comes down to manager access, relationships and information. Cummins says the top outcomes are generally captured by a very small subset of managers, perhaps 50 out of several thousand globally. In other words, venture remains a relationship market first, and an efficient capital market second.

That is why StepStone’s own model matters to his argument. The firm positions itself as a platform, investing through leading venture managers and directly into companies. It also uses secondary markets to build exposure where conviction is highest. Cummins says the edge comes from seeing more of the ecosystem, triangulating information across managers and rounds, and concentrating into companies before they become consensus winners.

Venture should now be part of the equity conversation

Cummins’ broader point is ultimately about portfolio construction. If investors believe technology will remain a dominant driver of equity returns, they must ask if public markets alone provide enough access. Cummins’ answer is no. More value is being created privately, more top companies are reaching scale before listing, and more future winners may be shaped in private markets long before traditional equity investors can buy them.

That does not make venture easy. It remains difficult, access-driven and highly skewed. But Cummins is arguing that it is becoming harder to ignore. Venture capital is no longer just the riskier end of private markets. It is becoming a more central part of understanding where future equity value may actually come from.