All-In's Best Ideas Pitch Competition: 4 Investors Present Their Top Trades Live
The All-In Best Ideas Pitch Competition featured four investors presenting their top investment ideas: Aaron Cowan pitching MGM Resorts, Dan Dreyfus pitching Talon Energy, Oleg Nodelman pitching Actis Oncology (AKTS), and Kyle Samani pitching GeoNet (GEOD token). The audience voted Talon Energy as the winner, while the besties panel voted MGM as their top pick.
Summary
The All-In Best Ideas Pitch Competition was modeled after the Ira Sohn Foundation investment conference, where top fund managers pitch their best investment ideas. The format was designed to showcase lesser-known managers who generate strong returns for their LPs but lack mainstream media distribution.
Aaron Cowan of Soretta Capital, a $4 billion New York-based generalist fund, pitched MGM Resorts. His thesis centered not on Vegas operations but on two hidden assets: a licensed casino in Osaka, Japan (opening 2030, estimated at $50/share in value) and a potential casino in Dubai where MGM has pre-built 300,000 sq ft of space. He noted Barry Diller's aggressive accumulation of 26% of the company and a subsequent $48/share bid, arguing shareholders should not tender. His sum-of-parts valuation suggested the stock could be worth $100-$150, calling it a potential triple.
Dan Dreyfus pitched Talon Energy, a power producer with 2 gigawatts of nuclear and 6 gigawatts of natural gas baseload power. His thesis was built around the AI-driven power demand supercycle, arguing that Talon trades at a $25 billion enterprise value against a $45 billion replacement cost. He outlined three scenarios: at $50/share of free cash flow (stock in high $300s), just running the business delivers a double; more data center contracts gets to $70/share FCF at 15x = ~$1,050; and building new capacity could push FCF above $100/share. He cited PJM's forecast of needing 106 gigawatts of new power in 10 years as evidence of structural scarcity.
Oleg Nodelman of EcoR1 Capital, a $2.5 billion San Francisco biotech fund with a 20% annualized return since inception, pitched Actis Oncology (AKTS). He framed modern radiopharmaceuticals as 'precision drone strikes' against cancer — mini proteins that navigate the bloodstream, find tumors by molecular recognition, and detonate with a 100-micron blast radius. Actis has a $1 billion market cap, $500 million enterprise value, 3+ years of cash runway, and initial clinical data expected in Q1 2027. Key de-risking factors include Eli Lilly's $100M IPO backstop, known valid targets (Nectin-4 and B7H3), and the fact that radioisotopes are off-limits to Chinese generic manufacturers due to supply chain constraints. His triangulated valuation suggested $10 billion or $200/share if one program succeeds.
Kyle Samani, founder of Multicoin Capital, pitched GeoNet (GEOD), a decentralized RTK (real-time kinematics) network providing 2cm precision geolocation versus GPS's 2-meter accuracy. GeoNet has grown to 22,000 base stations across 150 countries in just four years — twice the size of the next three legacy competitors combined — using a crypto incentive model where anyone can earn GEO tokens by hosting a base station. Customers include DJI, TomTom, John Deere, and robotic lawn mower manufacturers. The business generates $11M in annualized revenue, growing 3x YoY, and returns 80% of revenue via open-market token buybacks. Fully diluted market cap is approximately $150 million.
The bestie panel (Chamath, Gavin, Freeburg, and J.Cal) provided feedback emphasizing position sizing, liquidity, and risk-reward. MGM and Talon were seen as large, liquid positions with capped downside; Actis and GeoNet were framed as lottery tickets with potential zeros but massive upside. The audience voted Talon Energy first (50%), MGM second (24%), Actis third (21%), and GeoNet fourth (5%). The bestie panel flipped this, voting MGM first and Dan Dreyfus second, with Aaron Cowan taking the top prize in the panel vote.
Key Insights
- Aaron Cowan argued that MGM's real value lies not in its Vegas properties but in a licensed Osaka casino (est. $50/share value) opening in 2030 and a pre-built 300,000 sq ft space in Dubai that could become a casino if legalization occurs, together making the stock worth potentially $100-$150 vs. Barry Diller's $48 bid.
- Dan Dreyfus argued that Talon Energy trades at a $25 billion enterprise value against a $45 billion replacement cost, meaning the equity more than doubles just to reach replacement cost — applying Sam Zell's framework of buying hard assets below replacement cost during a build-out cycle.
- Dreyfus contended that AI doesn't create the power supercycle on its own — it merely 'turbocharges' an already tight power market, and that the PJM alone needs 106 gigawatts of new power in 10 years, a build that is physically impossible given current supply chains.
- Dreyfus framed data centers as the new refineries: electricity is the crude oil input, and tokens/intelligence are the refined output, making power producers structurally analogous to upstream hydrocarbon companies during a demand spike.
- Oleg Nodelman argued that radiopharmaceuticals represent a 'precision drone strike' modality in oncology — molecularly targeted, with a 100-micron blast radius — and that Actis's use of known valid targets (Nectin-4, B7H3) significantly de-risks early clinical development since imaging confirms target engagement.
- Nodelman argued that radioisotope-based drugs have a genuine moat because Actinium sourced from U.S. Cold War nuclear waste is unavailable in China, making this one of the few biotech modalities structurally protected from Chinese generic replication.
- Nodelman stated that GLP-1 obesity drugs are already among the best longevity drugs ever developed because caloric restriction is the only intervention proven in data to extend lifespan, and GLP-1s mechanistically replicate that effect.
- Kyle Samani argued that GeoNet is likely a natural monopoly in RTK positioning because telecoms historically monopolize due to network effects, and GeoNet has already achieved 2x the base station coverage of the next three legacy competitors combined in just four years.
- Samani argued that GeoNet's decentralized model — paying hobbyists in GEO tokens to host base stations — creates an unbeatable cost structure versus traditional telecom infrastructure rollouts, with 80% of $11M annualized revenue going to open-market token buybacks at a $150M fully diluted market cap.
- Chamath argued that while all four pitches had asymmetric alpha, the key differentiator was sizing capacity — Talon and MGM can absorb tens of millions without moving the market, whereas GeoNet is so illiquid that even getting $1M deployed would move the price significantly.
- Freeburg argued that Talon is more interest-rate sensitive than MGM because power purchase agreements are long-duration bond-like cash flows, meaning a rate spike would compress the 15x multiple framework used in the Talon valuation.
- Gavin argued that MGM had the best risk-reward of the four pitches because Barry Diller's $48 bid effectively caps the downside, while Japan and Dubai represent unpriced call options — and separately noted that entertainment monetization improvements at Vegas properties could drive gambling revenue significantly higher.
Topics
Full transcript available for MurmurCast members
Sign Up to Access