Is Non-Consensus Investing Overrated?

Core Thesis: The Dangers of Non-Consensus and Market Efficiency

• Martin Casado argues that non-consensus investing is inherently dangerous because founders depend on follow-on capital, and a lack of market consensus can lead to funding starvation.
  • Being "blinkered" to how Venture Capitalists (VCs) view a company creates significant risk if the founder cannot attract future funding.
  • Casado clarifies he does not advocate for "consensus investing" per se, but rather warns against being unaware of market consensus.
• Leo Gurevitch agrees that eventual consensus is necessary for company survival but notes his best investments often originated in non-consensus territory.
  • Non-consensus deals often struggle early due to a lack of proof points, but valuations can skyrocket once traction is proven.
  • The potential returns in non-consensus early-stage deals are significantly higher (e.g., 1,000x) compared to later-stage consensus rounds (e.g., 10x–20x).
• Both speakers acknowledge the market is likely more efficient than perceived, with prices generally converging on intrinsic value for good companies.
  • If a company is good, the market will eventually price it high; failing to recognize this suggests an investor is "beating themselves up" rather than the market.
  • There is a risk of "indigestion" where companies fail from raising too much capital too easily, rather than "starvation" from raising too little.

Analysis of Consensus vs. Non-Consensus Outcomes

• The group debates anecdotal lists of "non-consensus" wins (e.g., Airbnb, Uber, Palantir) to determine if these companies truly defied consensus or if the term is ill-defined.
  • Casado argues against conflating a "hard raise" with "non-consensus," noting that many high-profile companies were actually priced well above market median at some point.
  • The discussion highlights that definitions of "consensus" are often too narrow, sometimes labeling highly successful exits (like Palantir or Tesla) as non-consensus simply because they were controversial.
• Peter Thiel's maxim is cited as a counterpoint to the danger of non-consensus: "The faster and higher the up round, the more you should invest because it's like, you know, working."
  • This suggests that rapid, high-valued funding rounds can be a leading indicator of success rather than a bubble.
  • Leo notes that high-priced rounds can still be underpriced relative to the massive potential returns of unicorns.
• A key distinction is made between the "productive asset view" (value comes from the business) and the "human opinion view" (value comes from market sentiment).
  • Casado holds the productive asset view, believing smart investors correctly identify value and price accordingly.
  • Gurevitch suggests that human perception of a company can drive value independent of the underlying asset, noting cases where "super hot" companies with failing business models still succeeded due to momentum.

Founder Implications and Strategic Dilemmas

• Founders face a dual reality: they must be non-consensus regarding product-market fit to generate alpha, yet must appear consensus-compliant when raising capital.
  • Being perceived as non-consensus makes securing 18–24 month follow-on funding significantly harder.
  • Founders report feeling intense pressure to "cater to VCs," often mimicking market trends to avoid being passed over.
• There is a strategic trade-off between "hot" consensus rounds and "frugal" non-consensus rounds.
  • Consensus rounds can lead to over-spending and fragility if growth assumptions are not met immediately.
  • Non-consensus rounds often force founders to be more cash-efficient and frugal, potentially creating a more resilient business.
  • Consensus rounds can suffer from "soft diligence" where investors copy others (e.g., following Sequoia or a16z) without rigorous vetting.

Market Evolution and Future Trends

• The speakers debate whether the market is becoming more efficient over time.
  • Gurevitch suggests non-consensus investing is becoming more efficient (easier to find niche investors) while consensus investing is becoming less efficient (hyper-inflated prices for hot deals).
  • Casado argues that for the "mean" investment, prices are converging on fair value as more capital and smarter investors enter the asset class.
• Current market conditions show a dichotomy: speculative AI companies raising money with no business models vs. traditional infrastructure companies unable to raise despite strong fundamentals.
  • The AI wave is characterized by faster growth cycles (months vs. years) but potentially weaker moats compared to historical trends.
  • Deep tech sectors like defense and robotics have experienced valuation distortions where hype outpaced fundamentals (e.g., humanoid robotics).
• Fund mechanics and capital availability are identified as primary drivers of valuation inflation.
  • The size of VC funds is expanding (e.g., SoftBank, Tiger, Thrive) because the number of potential "100 billion dollar" outcomes is increasing.
  • If outcomes are 10x larger, fund sizes must be 10x larger to capture sufficient returns, even at higher entry prices.

Specific Sector and Strategy Observations

• Deep Tech & Humanoids: Gurevitch notes that humanoids are currently over-hyped with unknown unit economics, making standalone investment difficult without vertical integration.
  • Many deep tech companies rely on M&A rather than independent success, which limits the appeal for pure venture models requiring standalone scale.
• Multistage Investing: The conversation questions the thesis (e.g., Ron Cohen/Sequoia) that multistage firms dominate seed rounds.
  • Casado's data suggests that while multistage firms have advantages for repeat founders, seed funds still dominate the majority of top-tier outcomes.
  • Multistage firms often have a leg up on "known" founders (e.g., ex-unicorn founders) who command higher prices at seed.
• AI Investment Strategy: Gurevitch notes he has invested little in pure AI due to the deep tech angle, preferring companies where AI is an enabler rather than the core product.
  • Successful AI models (e.g., OpenAI, Anthropic) show great unit economics, but applying this model to other spaces without proof points is risky.
  • The "market tam sloppiness" of assuming infinite market size justifies any price is a common fallacy in venture.

Future Research and Data Analysis

• Casado, Gurevitch, and the host plan to conduct a quantitative analysis to test these theories.
  • The analysis will cohort companies into "winners" vs. "losers" to determine if winners raised at prices above or below the median for their stage.
  • A secondary metric will measure whether the bulk of venture returns come from high-priced or low-priced deals.
• The goal is to determine if price arbitrage is possible or if the market is efficient enough that one should not look for price advantages.
  • Anecdotal evidence suggests Casado's best investments were non-consensus (hard to raise), while his biggest misses were sometimes high-priced consensus deals.
• The conversation concludes that while individual VC identity is tied to being non-consensus, the aggregate market acts efficiently to allocate capital to the most promising opportunities over time.