What Deep Tech GPs Need to Prove Before LPs Trust Them with Long-Term Capital

What happened

Deep tech investing creates a different underwriting problem for LPs. In software, growth rates, retention, sales efficiency, and other operating metrics can often help an allocator understand whether a GP is reading the market correctly. The uncertainty is still real, but the signals are usually easier to compare.

In deep tech, the information asymmetry is deeper. Quantum computing, fusion, advanced materials, robotics, bioengineering, space, and other technical categories do not always produce clean commercial metrics early in the life of a company. The scientific premise may be difficult to evaluate. The path to commercialization may be long. The capital intensity may change by stage. Regulatory, manufacturing, and adoption risk can matter as much as the original technical breakthrough.

That makes the question for LPs more complicated. They are not simply asking, "Does this GP understand the technology?" They are asking something harder: "Can this person turn technical uncertainty into investment judgment that we can underwrite over a long period of time?"

That distinction matters for technical founders, researchers, and operators who want to become fund managers. Technical credibility may open doors in venture. It may help a junior investor win trust inside a fund. But when someone wants to raise and lead a fund, the LP question is different. The question is whether long-term capital can be entrusted to that person's judgment.

Why this matters

The common mistake is to assume that deep technical knowledge is the main ticket into deep tech fund management. It is an advantage, but it is not enough.

A technical expert can understand why a technology is elegant. A fund manager needs to understand why it can become an investable opportunity. Those are related, but not the same. The first asks whether the science is real. The second asks whether the science can become a company, whether that company can reach a market, whether the financing path is plausible, and whether the risk-adjusted return can justify the capital.

LPs also need to evaluate a manager across domains. A fund may begin with one area of expertise, but deep tech is not one market. A manager with a background in one field may need to evaluate adjacent or unfamiliar categories. No individual can be the deepest expert in every technical area. The more relevant question is whether the manager can learn quickly, ask the right questions, find the right experts, and identify the risks that matter for investment judgment.

This is why a deep tech GP's credibility is not only a matter of credentials. Credentials can show depth. They do not automatically show translation ability, commercialization judgment, or learning velocity.

How I see it

My view is that LPs are not underwriting technical expertise alone. They are underwriting a translation function.

The strongest deep tech GPs are not simply the people who can explain the science in the most detail. They are the people who can explain why the science matters as an investment problem. They can move between the lab, the customer, the capital stack, the regulatory pathway, the manufacturing constraint, the competitive landscape, and the LP's need for a coherent return story.

That is a different skill from being right about the technical premise. A technology can be real and still be a poor investment at a given stage. It may require more capital than expected. It may take longer to reach adoption. It may be overtaken by a simpler substitute. It may solve a real technical problem but not a painful enough customer problem. It may be too early for the market, or too late relative to an alternative approach.

This is where the phrase "commercialization instinct" becomes useful. It does not mean vague business intuition. It means the ability to read the distance between technical feasibility and investable company formation. How much time is needed? How much capital is needed? What must be proven before the next financing round? Which parts of the risk are scientific, which are manufacturing-related, which are regulatory, and which are market adoption problems?

For a researcher-turned-GP, the danger is becoming trapped inside the identity of the expert. Expertise can make a person confident in what they know. But LPs may be watching what the person does when they do not know. Can they ask basic questions without defensiveness? Can they admit uncertainty? Can they build a learning process around unfamiliar domains? Can they explain what remains unproven?

In deep tech, humility is not the opposite of expertise. It is part of the underwriting case.

Implications

For emerging GPs, the implication is clear: technical depth should be treated as a starting point, not as the whole fundraising argument.

A credible deep tech GP needs to show how technical understanding becomes investment judgment. That means being able to explain a technology to non-specialists without dumbing it down, compare competing technical approaches, identify adoption bottlenecks, and map a financing path that fits the time horizon of the company.

It also means showing a repeatable learning process. AI tools, public technical material, expert networks, customer interviews, scientific advisors, and investment memos can all help. The point is not that the GP must know everything personally. The point is that the GP must be able to reach the depth needed for investment judgment quickly enough, and know what still needs to be verified.

For LPs, this suggests a sharper manager assessment lens. Asking "Does this person have a technical background?" is too shallow. Better questions include: Can the manager explain the technical risk in plain language? Can they separate scientific risk from commercialization risk? Can they identify substitution risk? Can they describe why this approach may win against alternative solutions? Can they update their view when the technical or market evidence changes?

For deep tech founders, there is also an indirect implication. The best deep tech investors may not be the ones who simply admire the science. They may be the ones who can help translate the science into milestones, financing logic, market language, and investor conviction.

For the broader venture ecosystem, the issue is talent formation. More technical experts may want to move into investing, especially as deep tech becomes more strategically important. That is positive. But the path from expert to fund manager is not automatic. The market needs more people who can translate between science and capital, not only more people who can speak the language of science.

Open question

The question I would watch is whether more technical experts can become fund managers who translate science into investable judgment.

Deep tech needs investors who can understand difficult technologies. But LPs need more than technical confidence. They need to believe that a manager can learn across domains, explain uncertainty, judge commercialization risk, and allocate capital over a long time horizon.

If more researchers and technical operators can build that translation muscle, deep tech venture may gain a stronger class of specialist managers. If not, many will remain domain experts with capital ambitions rather than GPs that LPs can underwrite.

References

• Terry Kawanishi, 研究者出身者がディープテックGPとして独立する前に考えるべき3つのこと, June 14, 2026.
• The Japanese note above was inspired by an allocator-side discussion from the podcast Swimming with Allocators. The Ghost version reframes the idea for an international private-markets audience rather than translating the note directly.