Interview, Fireside Chat
Is Non-Consensus Investing Overrated?
- Non-consensus investors face difficulty securing follow-on capital if they lack consensus support, potentially leading to company failure within 18 to 24 months if capital markets remain inaccessible.
- Market efficiency varies by sector, with early markets remaining efficient for good companies while the consensus market for established firms risks becoming inefficient with valuations reaching 5x fair value.
- The current AI landscape is characterized by faster growth trajectories, such as reaching 100 million ARR in one to two years, yet potentially featuring weaker competitive moats than previous cycles.
- Traditional infrastructure companies not aligned with the current AI "sweet spot" are predicted to struggle raising capital, while specific entities like open AI, Anthropic, and Cursor are expected to grow significantly.
- Risks include market distortions where multi-trillion dollar market potential justifies high prices despite a one percent chance of success, and "indigestion" where companies raise excessive capital too easily.
- Deep tech Series A rounds require specific milestones to justify checks of $5 million to $20 million, as raising $100 million for early-stage milestones is often deemed not feasible.
- The number of $100 billion companies is expected to persist, estimated at 10 to 20 over the last 20 years, driving a need for larger fund sizes to capture outsized outcomes.
- Decacorn valuations are projected to be an order of magnitude higher than a decade ago, with potential for seed-like returns even at Series A or B pricing.
- Multi-stage funds hold a competitive advantage for founders with prior exits, potentially securing deal terms at valuations double those available to other investors.
- Humanoid robotics is identified as a highly hyped area where valuations may become disconnected from revenue before commercialization, making investment handicapping difficult.
- Autonomous vehicle unit economics remain comparable to Uber, presenting significant challenges for building standalone businesses without superior financial models.
- Historical data indicates dot-com bubble years produced terrible median fund returns due to overpayment, contrasting with the 2010 era where market pessimism enabled top quartile fund success.
- In a purely consensus world, competition shifts to the lowest cost of capital, potentially reducing investment diversity and "fun" while increasing the likelihood of asset prices reaching 5x fair value.
- The venture market has expanded to 100 times its size two decades ago, with capital inflows continuing as companies remain private longer, though SoftBank and Tiger-era mixed success is weighed against macro cycles.
- Most VC earnings are expected to originate from companies raised at high prices, driven by the outsized returns of the few winners in a batch of mostly non-hot companies.
- Defense company valuations experienced two- to four-fold increases following geopolitical conflicts in Ukraine and Israel despite unchanged fundamentals.
- Venture capital efficiency is viewed as a net positive for humanity by focusing on growth rather than preserving incumbents, though this efficiency may marginalize non-consensus ideas.
- Access to LP capital is cited as the primary constraint preventing further price increases, suggesting that increased capital availability could lead to higher valuations to secure top-tier positions.
- Investors may accept half ownership for half risk if they possess more capital, allowing them to pay higher valuations today for the best companies in a competitive landscape.
- Deep tech investors may face challenges convincing consensus investors for large checks without proven milestones, whereas multi-stage funds can leverage prior founder exits to accelerate deal closure.