All-In’s Best Ideas Pitch Competition: 4 Investors Present Their Top Trades Live

概览

This live episode recreates a Sohn-style best-ideas pitch format: four investors present high-conviction trades, then the All-In hosts and audience challenge the assumptions, compare risk/reward, and vote on the strongest idea.

The four pitches cover very different risk profiles: MGM as a hidden-asset casino and real-estate/value play; Talon Energy as a scarce power-infrastructure bet; Actus Oncology as a binary biotech platform opportunity in radiopharmaceuticals; and GeoNet as a crypto-token exposure to a decentralized high-precision location network for robotics and AI.

The discussion repeatedly returns to sizing, downside protection, liquidity, and catalysts. MGM and Talon are treated as more scalable, institutionally investable ideas, while Actus and GeoNet are framed as potentially high-upside but more binary or liquidity-constrained opportunities.

分段落总结

[00:00] Opening and Competition Setup

[事实] The host explains that the format is inspired by an earlier investment-pitch event connected to the Sone Foundation.

[事实] He recalls past pitches including Amazon as a potential trillion-dollar company, Tesla converts, and a missed opportunity where he pitched AI but chose Box instead of Nvidia.

[事实] The goal of this event is to showcase strong managers who make money for their LPs but do not necessarily get broad media distribution.

[02:29] Aaron’s MGM Pitch

[事实] Aaron introduces himself as running a $4 billion New York firm and says he chose MGM because pitching a tech stock to this audience would be unwise.

[事实] He says MGM owns 13 properties in Las Vegas and is one of the two largest casino asset owners there alongside Caesars.

[事实] He highlights Barry Diller’s aggressive buying, saying Diller owns 26% of MGM and recently bid $48 per share, while MGM has also bought back about half its float over six years.

[事实] Aaron argues MGM has hidden assets: an Osaka casino license in Japan and a possible future Dubai casino opportunity.

[05:03] MGM’s Osaka and Dubai Upside

[事实] Aaron says MGM’s Vegas assets plus China are worth around the low $60s per share.

[事实] He says MGM has a license to open a casino in Osaka, expected to open in 2030, and estimates the project could generate about $2 billion of EBITDA.

[事实] He argues Japan already has a large gambling market through pachinko and horse racing, and Osaka is well positioned for Chinese gamblers compared with Macau and Singapore.

[事实] He says MGM is building a Dubai property with 300,000 square feet of unused space that could become valuable if Dubai legalizes gambling.

[推测] Aaron’s core thesis is that the current bid creates downside support while Japan and Dubai provide underappreciated upside optionality.

[09:36] MGM Q&A

[事实] In response to questions about international customer credit and loyalty, Aaron says MGM has a large customer database and loyalty program.

[事实] He cites his bad investment in the Rio as a lesson in the value of loyalty programs, because losing the Caesars program hurt the asset.

[事实] Asked about entertainment monetization, Aaron says he does not know the answer but reiterates that his pitch is mainly an Asian casino play, not a Vegas play.

[事实] Asked why the market ignored Osaka after approval in 2023, he compares it to Wynn Macau and says markets often start caring about casino openings around three years before launch.

[13:04] Daniel’s Talon Energy Pitch

[事实] Daniel pitches Talon Energy and frames the thesis around a new power-demand cycle driven by technological adoption.

[事实] He says power demand normally grows with GDP, but power-intensive technology cycles can cause demand to spike before efficiencies eventually bring growth back down.

[事实] He argues AI demand is not necessary to keep power markets tight for the next 20 years; AI only turbocharges the shortage.

[事实] He says Talon owns two gigawatts of nuclear power and six gigawatts of natural-gas baseload power.

[事实] He values Talon at a $25 billion enterprise value versus a $45 billion replacement cost, implying more than a double in equity value just to reach replacement cost.

[16:39] Power Scarcity and AI Infrastructure

[事实] Daniel argues the U.S. needs to copy China’s aggressive power buildout, saying China went from having half the U.S. power generation capacity to three times as much.

[事实] He describes data centers as refineries that take electricity as input and produce tokens, photons, or intelligence as output.

[事实] He cites Jensen Huang as saying AI may need vastly more power than exists today.

[事实] He says the system will need every source of power, including nuclear, solar, orbital, and other options, because supply chains and raw materials will be constrained.

[19:18] PJM Grid Crunch and Talon Valuation

[事实] Daniel says PJM forecasts a need for 106 gigawatts of new power over the next 10 years, which he argues is extremely difficult to build in that timeframe.

[事实] He says coal plants are unlikely to retire as planned because the grid will need existing capacity.

[事实] He describes hyperscalers as urgently seeking long-term power-purchase agreements and cites Microsoft supporting the restart of Three Mile Island through a 20-year fixed-price contract.

[事实] He says Talon could generate $50 per share of free cash flow by doing nothing, potentially $70 with more premium data-center contracts, and over $100 if it builds new capacity.

[23:17] Talon Q&A

[事实] Asked about regulatory risk, Daniel says AI is necessary and suggests batteries and peaker plants behind data-center PPAs as a bridge solution for peak-demand hours.

[事实] He says the Talon thesis does not require behind-the-meter colocation because clean firm baseload is scarce either way.

[事实] Asked about competing power sources, he says the market needs all options, including fuel cells, gas turbines, and orbital compute.

[事实] Asked about terminal multiples, he says contracted infrastructure cash flows deserve a higher multiple than merchant power exposure and uses 15 times as a blended rule of thumb.

[27:17] Oleg’s Actus Oncology Pitch

[事实] Oleg introduces Eco R1 Capital as a San Francisco-based, value-oriented biotech fund with about $2.5 billion under management.

[事实] He says biotech investing is usually a horrible idea and compares much of the sector to a casino, but says his fund focuses on margin of safety rather than falling in love with science.

[事实] He explains cancer treatment history through a military analogy, moving from surgery and radiation to chemo, targeted therapy, immunotherapy, and modern radiopharmaceuticals.

[事实] He pitches Actus Oncology, ticker AKTS, as a billion-dollar-market-cap company with a $500 million enterprise value and more than three years of cash.

[31:34] Actus Platform and Catalysts

[事实] Oleg says Actus recently went public through a $300 million IPO that was 18 times oversubscribed and backstopped by a $100 million Eli Lilly order.

[事实] He says the company’s platform can carry radioactive payloads, target multiple tumor markers, and clear the body with minimal side effects.

[事实] He says imaging can verify whether the drug reaches tumors in early trials, which derisks development.

[事实] He says Actus chose known targets including Nectin-4 and B7H3, with initial clinical data expected in 2027 and Nectin-4 data potentially as early as Q1.

[事实] He says Actus could be worth $10 billion, or $200 per share, if even one program makes it to market.

[34:41] Actus Risks and Biotech Q&A

[事实] Oleg warns that Actus has no marketed products, no recurring revenue, possible dilution risk, and an unproven mini-protein radiopharmaceutical delivery approach.

[事实] Asked why radioisotopes are off-limits to China, he says Actus uses Actinium derived from Radium 233 waste from U.S. nuclear programs.

[事实] He says he has not seen China-based competition in this area, though he allows that China could potentially develop something for its own market.

[事实] Asked why the market discounts Actus after the IPO, he says biotech investors are very short-term oriented and may wait until closer to the clinical data.

[事实] Asked about longevity, he says he would take the over on longer lifespans and points to GLP-1 obesity drugs and caloric restriction as relevant.

[40:19] Kyle’s GeoNet Pitch

[事实] Kyle introduces GeoNet as a crypto asset building AI-related infrastructure through a real-time kinematics location network.

[事实] He says RTK can localize position to about two centimeters, compared with roughly two meters for GPS.

[事实] He says GeoNet is the world’s largest and fastest-growing RTK network, live in 150 countries, more than 11,000 cities, and covering about 80% of the global population outside sanctioned countries.

[事实] He says GeoNet scaled quickly through a decentralized model where individuals buy base stations, place them on rooftops, and earn Geo tokens.

[43:33] GeoNet Customers and Token Economics

[事实] Kyle lists agriculture, robotic mules, John Deere spraying systems, TomTom autonomous-vehicle maps, robotic lawn mowers, and DJI drones as use cases or customers.

[事实] He argues GeoNet has telecom-like network effects and a low cost structure because it does not deploy all base stations itself.

[事实] He says GeoNet recently crossed $11 million in annualized run-rate revenue and is growing more than 3x year over year.

[事实] He says about 80% of revenue is used to buy Geo tokens on the open market, while the remaining 20% funds R&D and business development.

[事实] He says the token trades on Solana under the ticker GeoD.

[48:36] GeoNet Q&A

[事实] Kyle says GeoNet’s fully diluted market cap is about $150 million, while crypto websites may show $60 million to $70 million because not all tokens are floating.

[事实] He says GeoNet is a U.S.-based corporation with a San Francisco team and a contractual relationship with the GeoNet foundation to use 80% of revenue for token purchases.

[事实] He says token holders, not equity holders, are the intended value-accrual path in his pitch.

[事实] Asked about securities-law risk, he says the Clarity Act would be positive but notes he is not a lawyer.

[事实] Asked why John Deere would use GeoNet rather than its own RTK system, he cites lower cost, OpEx instead of CapEx, and global availability.

[事实] Asked about satellite-based alternatives, Kyle argues satellites cannot compete on cost and that ground-based stations are better for battery-sensitive applications.

[54:48] Judges’ Feedback on Risk and Sizing

[事实] Chamath says he likes all four ideas but differentiates them by alpha, downside risk, liquidity, and position sizing.

[事实] He says GeoNet is hard to size because buying meaningful amounts could move the market, while Talon could absorb tens of millions of dollars.

[事实] He frames Actus as having discontinuous downside risk but possible 10x upside, while MGM and Talon can support larger position sizes.

[事实] Gavin says MGM has the best pure risk/reward because Barry Diller’s bid caps downside while Japan and Dubai create future value.

[事实] Gavin also finds Talon compelling but flags regulatory risk around AI and electricity prices.

[60:58] Host Rankings and Counterarguments

[事实] One host says MGM may have additional upside from better entertainment monetization at Vegas properties, citing the Sphere’s impact on Venetian EBITDA.

[事实] He says Talon may have 3x to 5x upside but over a longer timeline and with interest-rate sensitivity through PPA discount rates.

[事实] He says Actus faces China risk, competing therapeutic-modality risk, and potentially large downside on bad clinical data.

[事实] He argues GeoNet may face long-term risk from satellite systems that could replace RTK and GPS.

[事实] Jason groups AKTS and GeoD as lottery-ticket-like ideas, while MGM and Talon feel more protected because people will keep gambling and using electricity.

[65:58] Voting Results

[事实] The audience vote is based on 150 votes.

[事实] Kyle finishes fourth with 5% of the audience vote, Oleg third with 21%, Aaron second with 24%, and Dan first with 50%.

[事实] The Bestie vote ranks Kyle fourth, Oleg third, Dan second, and Aaron first.

[事实] The audience selects Talon as the winner, while the Bestie vote selects MGM.

播客点评/总结

[推测] The episode’s main value is that it compresses four different investment styles into one live format: event-driven value, infrastructure scarcity, clinical biotech optionality, and crypto-network economics. The Q&A is useful because it forces each pitch to confront catalysts, legal risk, market structure, competition, and liquidity.

[推测] The strongest part is the contrast between good storytelling and investable sizing. The hosts repeatedly separate “interesting idea” from “position you can actually put capital into,” which makes the discussion more practical than a simple stock-pick showcase.

[推测] The main limitation is that all four pitches are highly promotional and cannot substitute for due diligence. Several key claims, including valuation assumptions, clinical timelines, token-law treatment, and power-market forecasts, are asserted by the speakers rather than independently verified in the transcript.

[推测] This episode is best suited for investors or market-oriented listeners who already understand asymmetric risk and want a menu of ideas to research further, not for listeners looking for conservative financial advice.