All-In's Best Ideas Pitch Competition: 4 Investors Present Their Top Trades Live
The All-In Best Ideas Pitch Competition features four investors presenting their top trade ideas: Aaron Cowen pitching MGM Resorts, Dan Dreifus 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 ranked MGM first. The event was modeled after a similar investor pitch competition started in honor of Ira Sone.
Summary
The event opens with context about its inspiration from the Sone Foundation's pitch competition, where the host previously pitched Amazon as a trillion-dollar company in 2015 and was laughed out of the room by David Einhorn. The goal is to give great fund managers distribution beyond CNBC.
Aaron Cowen of Serreta Capital (formerly CIO for Steve Cohen, ran equities for Soros) pitches MGM Resorts. His core thesis is not about Vegas operations but two hidden assets: (1) MGM's license to open the only casino in Osaka, Japan in 2030 — a market he estimates is worth ~$50/share given Japan's $40B gambling market and Osaka's geographic proximity to Chinese gamblers relative to Macau and Singapore; and (2) an MGM/Bellagio/Arya complex being built in Dubai with 300,000 sq ft of 'flex space' that could become a casino if Dubai legalizes gambling. He values Vegas+China at ~$60, Japan at ~$50, and Dubai optionality at ~$40-50, suggesting the stock could triple from ~$37-48. Barry Diller's 26% stake and $48 bid (which Cowen says he would not sell to) provides downside protection.
Dan Dreifus pitches Talon Energy, a power producer with 2 GW of nuclear and 6 GW of natural gas baseload. His thesis centers on a Sam Zell-style framework: buy hard assets below replacement cost that will be needed in the future. Talon trades at $25B enterprise value vs. $45B replacement cost, implying more than a double just to reach replacement cost. He argues AI demand turbocharges an already tight power market, citing PJM's forecast for 106 GW of new power needed in 10 years — equivalent to Japan's entire consumption. He outlines three scenarios: $50/share FCF (7x at current ~$380 price doing nothing), $70/share if they sign more data center contracts, and $100+ if they build new capacity. He also highlights Microsoft's deal to restart Three Mile Island at $100/MWh vs. market $50 as evidence of hyperscaler desperation for power.
Oleg Nodelman of Eco R1 Capital pitches Actis Oncology (AKTS), a $1B market cap / $500M enterprise value radiopharmaceutical company. Using military analogies to describe cancer treatment evolution (surgery as medieval siege, chemo as mustard gas, immunotherapy as recruiting local T-cell allies), he frames modern radiopharmaceuticals as 'micro-drone assassinations' with 100-micron blast radius. Actis uses mini-proteins to deliver radioactive payloads to tumors, with a key de-risking feature being imaging to confirm target engagement. Their lead programs target Nectin-4 (bladder cancer) and B7H3 (expressed on all major solid tumors including prostate, colorectal, lung), with data expected in Q1 2027. He values the company at $10B/$200/share if one program succeeds. He notes China cannot replicate due to actinium sourcing from U.S. nuclear waste programs.
Kyle Samani of Multicoin Capital (who led all three Solana pre-launch investment rounds) pitches Geonet (GEOD), a decentralized RTK (Real-Time Kinematics) network providing 2cm precision geolocation vs. GPS's ~2m. Geonet was founded in 2021, has deployed 22,000 base stations globally (2x the next three competitors combined: Trimble, Hexagon, Topcon), covers 150 countries and 80% of global population. The network is built by hobbyists/small business owners who buy ~$300 base stations and earn GEOD tokens. Customers include DJI (world's largest drone maker), TomTom (AV map supplier), John Deere (autonomous spraying), and robotic lawnmower manufacturers. Revenue just crossed $1M annualized run rate, growing 3x YoY, with 80% of revenue used to buy back tokens on the open market. Fully diluted market cap is ~$150M. Samani argues it's a natural telecom monopoly with network effects.
In the panel discussion, the besties debate sizing (Talon and MGM can absorb tens of millions; Geonet is too illiquid), risk profiles, and rankings. Gavin Baker flags Talon's regulatory risk and interest rate sensitivity on PPAs. David Friedberg notes the sphere entertainment effect at Venetian ($1M/day incremental EBITDA when a show runs) as support for MGM's entertainment upside thesis. The audience voted Talon first (50%), MGM second (24%), Actis third (21%), Geonet fourth (5%). The besties panel inversely ranked MGM first, Talon second, Geonet third, Actis fourth.
Key Insights
- Aaron Cowen argues that MGM's Osaka casino license — opening in 2030 in the only Japanese prefecture that approved gambling — is a hidden asset worth ~$50/share because Osaka is geographically closer to Shanghai and Beijing than both Macau and Singapore, positioning it to capture Chinese gambling demand in a first-world country at a higher multiple than Macau.
- Dan Dreifus argues that AI demand alone is not necessary to keep power markets tight for 20 years — it merely turbocharges an already constrained system — and that Talon Energy trades at $25B enterprise value versus $45B replacement cost, meaning equity more than doubles just to reach replacement cost before any upside from new data center contracts or power plant construction.
- Dan Dreifus cites PJM's forecast of 106 GW of new power needed in 10 years for just one U.S. region — equivalent to Japan's entire national consumption — and argues this is physically impossible to build given supply chain constraints on critical minerals shared with rockets and solar panels, meaning existing capacity like Talon's will remain structurally scarce.
- Oleg Nodelman argues that Actis Oncology is protected from Chinese biosimilar competition — which has depressed much of biotech since the Amgen Supreme Court ruling — because actinium, their radioactive payload, is manufactured from radium-233 sourced exclusively from U.S. Cold War nuclear program waste, making the supply chain unavailable to Chinese manufacturers.
- Kyle Samani argues Geonet is an extremely cheap asset at $150M fully diluted market cap because 80% of its ~$11M annualized revenue is contractually directed to open-market token buybacks via a foundation agreement, making GEOD effectively a revenue-share token growing 3x year-over-year in a crypto bear market where no one is paying attention.
Topics
Full transcript available for MurmurCast members
Sign Up to Access