Interview, Fireside Chat
Larry Aschebrook, Founder & MP @GSquared: How We Lost Money on Uber and Made Millions on Lyft
- Larry Ashbrook's venture capital firm, G Squared, generated approximately $800 million in returns for Limited Partners (LPs) from its largest holding, Coursera.
- The firm deployed roughly 40% of its third fund ($380 million total) into Spotify, eventually becoming a top-10 global shareholder.
- In contrast to the Lyft success (approx. 3x return), the firm lost approximately $50 million on its Uber investment, selling at a 20-cent discount per dollar due to IPO pricing dynamics.
- Ashbrook pivoted his strategy in 2021 after overpaying for assets like Toast and Getir; he admitted to LPs that the firm "fucked up" and needed an additional $300 million to protect capital.
- The firm shifted its 2020 vintage to a 70% primary investment strategy with 40% of that capital structured with IRR hurdles and multiple ratchets to mitigate valuation risks.
- Ashbrook personally absorbed a $2–3 million loss to settle a Theranos transaction, refusing to let LPs bear the cost of his decision to proceed despite "gut feeling" red flags.
- The firm lost approximately $70 million on 23andMe by chasing valuation multiples rather than adhering to disciplined sell-guardrails.
- A significant loss occurred with Getir, where the firm deployed a second $100 million tranche to restructure equity; Ashbrook estimates this decision, combined with fundraising impact, cost the partnership roughly $500 million in total value.
- The firm's portfolio concentration strategy aims to keep 80–90% of risk concentrated in only 10 companies, contrasting with traditional indexing approaches.
- Ashbrook attributes early success to being a "secondary direct buyer" in a fragmented market, securing assets at roughly 35 cents on the dollar when few other buyers existed.
- The firm's investment thesis is built on four megatrends: SaaS, Fintech, Consumer Internet, and Mobility, seeking to balance high-growth momentum plays with established cash-flow businesses.
- Ashbrook believes the current LLM space will consolidate around a few winners (OpenAI, Anthropic), viewing application-layer entrants as high-risk "zombies" or "vampires."
- The firm has adopted a "DPI (Distributions to Paid-In) as the North Star" metric, explicitly advising LPs to ignore TVPI and MOIC as they do not represent realized cash returns.
- Ashbrook's early investment approach involved "mosaic theory" and sending analytical work to low-cost research teams in India to validate theses before deploying capital.
- He describes his leadership style as "productively paranoid," a trait that helped him exit the Theranos deal early but also caused friction when selling winners like Instacart to preserve liquidity.
- The firm's strategy relies on high-frequency, micro-transactions (sub-$2 million) to build large positions over time, creating a "Trojan horse" mechanism for deep company access.
- Ashbrook admitted to building the firm on a "brand and lifestyle" trap in 2020-2021, believing they were "smarter than the market," which led to a loss of focus on core value creation.
- The firm is currently navigating a difficult fundraising environment for its 2022 vintage ($1.2 billion), contrasting with the easy capital availability seen during the 2020-2021 boom.
- Ashbrook emphasizes that "liquidity is harder than entry," noting that selling winners at scale in private markets is a critical, often underestimated skill.
- He cites Cal Ripken Jr.'s 16-year streak of playing every game as his model for the "20-mile march" and long-term operational discipline required for G Squared.
- The firm plans to continue focusing on "picks and shovels" in the AI sector, investing in infrastructure companies like Lambda and CoreWeave alongside foundation models.
- Ashbrook revealed he sold Instacart shares at 20x revenue partly due to a personal negative customer experience during the pandemic, highlighting the role of non-financial data in decision-making.
- The firm's early success was fueled by a "land and expand" strategy, starting with small checks to gain access and data before concentrating capital in winners like Alibaba and Spotify.
- Ashbrook warns that the industry's shift toward "evergreen" and "continuation funds" is a reaction to a broken fund structure, though he notes continuation funds often only acquire high-quality assets at fair value.
- He advises endowments to stop chasing "fake numbers" like TVPI and MOIC and to focus exclusively on DPI as the only metric that can fund current obligations.
- The firm's "G Squared Eight" is a risk-management framework modeled after Mitchell Green's approach, focusing on quantitative guardrails to counter qualitative bias.
- Ashbrook's "worst three years" (post-2021) were defined by the necessity to restructure over-inflated assets and battle with founders of companies like Getir that had over-expanded.
- He maintains that the firm's identity is built on the logo and team, not individual partners, to ensure longevity across leadership transitions, akin to Renaissance Technologies.
- Ashbrook notes that the firm's "spider sense" (intuition) is a critical risk management tool, often leading to withdrawals from deals that lack transparency, such as the Theranos incident.
- The firm's strategy involves "chasing the next win" rather than resting on past laurels, a mindset rooted in Ashbrook's upbringing in poverty and his fear of returning to financial instability.