Deep Tech, the Asset Class
A framework for LP allocation in the new deep tech cycle
Before getting into the framework, a bit of context on how I think about this.
I spent time at J.P. Morgan’s Private Bank on the Innovation Economy team, working closely with founders, venture firms, and the investors behind them.
That vantage point gave a clear view across the full capital stack — not just what was being built, but who was funding it, and where the biggest gaps in understanding were.
One theme kept coming up: deep tech.
There was real interest from investors — they wanted exposure and could see something important was happening. But there wasn’t always a clear framework for how to evaluate it or where it fit within a portfolio. There was constant discussion around how to invest and own it — whether through public markets, private venture, or somewhere in between.
Part of the challenge is that deep tech doesn’t fit neatly into a single box. It spans multiple industries — semiconductors, AI, quantum computing, energy, infrastructure, and robotics — often overlapping in ways that don’t map cleanly to traditional venture categories.
We’re still early in what will become one of the most important allocation shifts of the decade.
The core shift: necessity, not trend
The last wave of technology — social, mobile, SaaS — was built on a simple assumption: that the physical world just worked. Infrastructure was stable. Supply chains held. Energy was available. Chips were abundant. You could build software on top of a functioning system and not worry about what was underneath.
That assumption broke — and it broke fast.
AI demand ran into hard physical constraints: not enough chips, not enough power, not enough bandwidth. Geopolitical shifts made clear that dependence on adversarial nations for critical technology isn’t just a supply chain issue — it’s a national security one. And the energy transition stopped being an ESG conversation and became a resilience problem.
What makes this moment different isn’t that one of these things happened — it’s that all of them happened at the same time. The cracks didn’t appear one by one — the whole foundation shifted at once.
And the capital response has had to match that reality. Governments, militaries, sovereign wealth funds, and strategic corporates are all investing alongside venture now. That doesn’t happen for incremental products. It happens when the need is structural.
Deep tech is what rebuilds the foundation. That’s why this is necessity capital — not trend capital.
The numbers are hard to ignore
According to BCG, deep tech now accounts for 20% of all global VC funding — double its share from a decade ago — and McKinsey puts average net IRR for deep tech funds at roughly 17%, compared to around 10% for traditional tech. Deep tech companies now represent 25% of all new unicorn startups globally, and the sector is projected to grow from $150 billion in 2025 to over $425 billion by 2034. The category has already created significant value. The next wave is still in front of us.
On liquidity — that picture is changing too. SpaceX is preparing what could be the largest VC-backed IPO of all time, and dozens of deep tech companies are eyeing listings over the next 12 to 18 months across AI, infrastructure, quantum, and beyond. M&A in 2025 came roaring back. The exit market continues to show promising signs of reopening.
How the best LPs are thinking about it
The commercialization window is closer than most LPs think. The best GPs today are investing in companies that have already cleared key technical milestones, are generating early revenue, and have credible paths to scale — and the risk profile reflects that.
Diligence is the edge. Deep tech rewards LPs who ask better questions. Understanding whether the science is real, whether the IP holds, and whether the team can operate across both technical and commercial dimensions — that’s where sophisticated allocators separate themselves. It also makes manager selection more important than in any other category. The GPs who can actually underwrite this have a real and durable advantage.
Early movers are still early. The companies being built right now are likely to become the Intels, GEs, and Boeings of the next 50 years. LPs who are building relationships and exposure now — before the category feels obvious — are positioning themselves for the vintages that will define the next decade.
The bottom line
At Massive, we talk about this constantly — how these verticals are colliding and overlapping in real time. AI running into energy constraints. Energy intersecting with infrastructure. Robotics and space converging with compute.
It’s not one big wave. It’s multiple waves hitting at the same time — and the companies being built at those intersections are the ones we find most compelling.
We believe this will be one of the defining venture vintages of our generation. Not because deep tech is suddenly popular — but because the underlying forces driving it aren’t going away.
This is a multi-decade rebuild of the systems the world actually runs on.
For LPs, the question is no longer whether to have exposure.
It’s whether you’ve built the relationships, the diligence capability, and the portfolio construction to access the right opportunities.
There’s still a window to do that — alongside GPs who have been in this space well before it became consensus. The opportunity is real, the timing is right, and the companies being built today will look obvious in hindsight.
If you're an LP, allocator, or fellow emerging manager thinking through how venture is shifting — particularly across deep tech and portfolio construction — I'll be here writing about all of it. Follow along, and feel free to reach out directly if you want to compare notes.
Matthew Klein is a Partner at Massive, focused on capital formation and investor relationships. Previously he co-founded Backbone PLM, co-founded Sweater Ventures, and worked in J.P. Morgan's Innovation Economy group advising founders, venture firms, and family offices.