Behind the video
Ask Me Anything: July 2026 Edition
Overall Thesis
Hoium sees excessive optimism in AI infrastructure spending, commodity-like neocloud economics, and expensive speculative businesses, while finding attractive valuations in selected consumer platforms and long-term asymmetric holdings. He favors businesses with durable customer relationships and identifiable cash-flow potential, while acknowledging uncertain timing, execution risk, and weak recent portfolio performance.
Narratives
GOOGLAlphabetAlphabet remains a major holding and his preferred way to invest in AI because its profitable core business offers financial flexibility. However, elevated valuation, slower core-business growth, and negative free cash flow from aggressive capital spending explain his earlier trim and current caution.
Key Arguments
- AI capital spending is approaching $200 billion and could rise further.
- Consensus assumes a rapid future cash-flow payoff that he considers optimistic.
- Alphabet could substantially improve free cash flow by reducing capital expenditures.
- Search and YouTube growth are solid but insufficiently exceptional to dismiss valuation concerns.
Risks acknowledged
- Alphabet has greater funding capacity than smaller AI infrastructure competitors.
- He still prefers Alphabet over other AI investments despite its current valuation.
Predictions (1)
ORCLOracleOracle illustrates his concern that AI infrastructure expansion depends on continually accessible debt and equity financing. Rising borrowing costs and falling share prices could undermine the economics of its spending commitments.
Key Arguments
- Higher financing costs can eliminate the spread between project returns and capital costs.
- Large AI commitments require funding that must come from operations, debt, or equity.
- He questions whether all participants in the AI value chain can sustain high margins.
Risks acknowledged
- Some AI companies may build substantial profitable businesses.
CRWVCoreWeaveHe avoids neocloud investments because GPU rental capacity looks like a commodity rather than a durable platform. CoreWeave is explicitly included among the companies whose spending and financing sustainability he questions.
Key Arguments
- Hyperscalers can use outside capacity temporarily and subsequently bring workloads in-house.
- Compute capacity, chip performance, and model efficiency are all expanding rapidly.
- Revenue commitments alone do not establish attractive bottom-line returns.
- Debt-funded capacity is exposed to falling rental prices.
Risks acknowledged
- GPU rentals can generate exceptional profits while supply remains constrained.
Predictions (1)
NBISNebius GroupNebius is part of the neocloud group he excludes from his portfolio. He also chooses cash when asked to select investments from a group containing Nebius.
Key Arguments
- Short-duration GPU rentals do not create the platform moat associated with early cloud infrastructure.
- Future clearing prices and returns on enormous capacity additions are difficult to forecast.
- Debt funding increases vulnerability if customer demand or pricing weakens.
Risks acknowledged
- Current AI demand can support strong near-term economics.
Predictions (1)
IRENIRENHe includes IREN in the AI infrastructure rotation and rejects neoclouds as portfolio candidates. His objection centers on uncertain investment returns and a lack of durable differentiation.
Key Arguments
- AI infrastructure expansion can attract excess competing capital.
- Financing commitments precede certainty about future GPU rental profitability.
- Hyperscalers may internalize capacity when economically advantageous.
Risks acknowledged
- The AI buildout has generated substantial near-term investor interest.
Predictions (1)
He continues to endorse Mobileye as a holding because its existing business provides cash generation and a strong balance sheet while more advanced autonomy offers upside. Commercialization delays and leadership uncertainty have hurt the stock, but he has not concluded that the investment thesis is broken.
Key Arguments
- Approximately $1.4 billion of cash supports ongoing research and development.
- Higher-value driver-assistance systems could lift revenue and margins.
- Established relationships with major automakers support future adoption.
- Advanced autonomous systems and possible recurring revenue provide optionality.
Risks acknowledged
- Autonomous-driving deployments are slower than investors expected.
- The outgoing founder's related-party acquisition raised governance concerns.
- Leadership transition creates uncertainty.
- Recent stock performance has been poor.
Predictions (1)
UBERUber TechnologiesHe recommends owning Uber as a durable aggregator of transportation demand rather than assuming autonomous vehicles will displace it. He also expects improved stock sentiment if alternative autonomous suppliers allow Uber to demonstrate a competitive fleet and faster scaling.
Key Arguments
- Consumers are unlikely to adopt dozens of separate autonomous-ride apps.
- Uber's customer relationship and distribution can capture value across vehicle suppliers.
- A valuation below roughly 20 times earnings appears attractive relative to its growth.
- New autonomous supply could challenge the perceived dominance of vertically integrated rivals.
Risks acknowledged
- Waymo and Tesla remain a perceived competitive overhang.
- The timing of stock repricing is unknown.
- Alternative autonomous fleets are not yet operating at the required scale.
LYFTLyftHe plans to hold Lyft long term and considers its valuation unusually compelling compared with its operating trends. He rejects the market's implication that the company faces serious financial or strategic distress.
Key Arguments
- Forward free-cash-flow and earnings multiples are low.
- Gross bookings, rides, and active riders continue to increase.
- Autonomous suppliers could use Lyft's demand aggregation and distribution.
- Share repurchases could create value at the current valuation.
Risks acknowledged
- Revenue and insurance reserves can be volatile.
- He does not know when investor sentiment will improve.
- He explicitly declines a year-end price target.
Predictions (1)
He sees Hims as a potentially large global healthcare platform built around direct consumer relationships, subscriptions, and personalized treatment. He expects a higher valuation if second-half growth accelerates, while describing the long-term investment as a high-upside, potentially binary outcome.
Key Arguments
- Subscriber growth could reaccelerate after difficult comparisons.
- Weight-loss customers can increase average revenue per subscriber.
- A 50% or greater second-half growth rate could warrant a higher sales multiple.
- Global brand consolidation, lab services, and consumer data could deepen relationships.
- Specialized healthcare relationships may be more durable than general retail offerings.
Risks acknowledged
- The growth and guidance catalysts remain unconfirmed.
- He allows that his view of specialized healthcare platforms could be wrong.
- The company could ultimately fail rather than realize its upside.
He questions Zoom's differentiation because competing meeting tools are bundled into broader products. Although existing users may be sticky, he views video meetings as potentially a feature rather than a sufficiently differentiated standalone business.
Key Arguments
- Many competing products provide similar functionality.
- Customers may already receive meeting software through other subscriptions.
- The moat and disruptive opportunity are unclear.
Risks acknowledged
- Existing enterprise users may be reluctant to migrate.
- The product may be difficult to replace through simple AI-generated code.
SOFISoFi TechnologiesHe is actively adding to SoFi and intends to continue buying during declines while the business remains healthy. His thesis emphasizes compounding financial-services profits and the long-term advantages of digital banking.
Key Arguments
- Financial-services revenue and contribution profit are growing rapidly.
- The loan platform creates fee income without requiring every loan to remain on the balance sheet.
- Digital products appear more compelling than traditional bank branches.
- SoFi can gradually expand its customer relationships and move upmarket.
Risks acknowledged
- The stock is likely to remain volatile.
- Moving upmarket could take 10 to 20 years.
- The eventual size of its traditional lending opportunity is uncertain.
ONONOn HoldingHe remains happy to own On because growth, pricing power, and margins look strong relative to the valuation. He cannot predict the next earnings reaction and accepts that the market may take time to recognize the business.
Key Arguments
- Revenue continues to compound.
- Gross margin has exceeded the company's 60% ambition.
- A forward earnings multiple around 19 appears reasonable.
- Higher-income consumers continue to purchase premium footwear.
Risks acknowledged
- Apparel stocks are out of favor.
- Currency translation complicates reported growth.
- A still-strong but lower margin could receive a negative market reaction.
Predictions (1)
He endorses Joby as a speculative portfolio holding with potentially large long-term rewards if commercialization succeeds. He describes a possible interim fall to $3 and a binary long-term outcome rather than a dependable near-term return.
Key Arguments
- The aircraft and infrastructure strategy define a recognizable future business.
- Blade provides demand relationships and access to landing infrastructure.
- Hydrogen could expand range using a related powertrain.
- He believes execution has been strong despite relatively few promotional announcements.
Risks acknowledged
- The company is pre-revenue and very high risk.
- Commercial operations by year-end may be a stretch.
- Battery range is limited.
- The stock could decline substantially before any long-term success.
The satellite discussion identifies AST SpaceMobile as the company he is short, despite garbled captions around the initial name. He believes its valuation overlooks weak satellite-connectivity economics and ongoing replacement capital expenditure.
Key Arguments
- Telecom partners capture revenue without funding the satellite infrastructure.
- The number of customers willing to pay extra for satellite access may be small.
- Multiple competing satellite networks constrain differentiation.
- Satellites require continual replacement and reinvestment.
Risks acknowledged
- The company has an appealing growth narrative.
- He is willing to be proven wrong.
Predictions (1)
COINCoinbase GlobalHis Coinbase thesis centers on blockchain financial infrastructure rather than cryptocurrency trading or Bitcoin prices. He sees Base, stablecoins, and new financial applications as potential sources of asymmetric business value, without making a distinct stock-return call here.
Key Arguments
- Regulatory clarity could enable broader experimentation and adoption.
- Base is attracting applications beyond speculative trading.
- Stablecoins may reduce merchant payment costs.
- Coinbase's balance sheet provides acquisition flexibility.
Risks acknowledged
- He has not completed his analysis of the Clarity Act.
- The eventual business model and applications remain uncertain.
- He has not read the cited short report.
Comfort Systems represents his concern that AI-related industrial stocks capitalize unusually high margins as if they were permanent. He questions its valuation and profitability durability but does not issue an explicit stock-price forecast or trade instruction.
Key Arguments
- Data-center construction demand has accelerated growth.
- Margin expansion has amplified earnings and share-price gains.
- Commodity-like construction services may not sustain current pricing power.
- Trailing and forward earnings multiples appear demanding.
Risks acknowledged
- The business has genuinely benefited from stronger demand and margins.
Nintendo's stock decline has attracted his interest, but weak industry growth and uncertain Switch 2 demand keep him unconvinced. He values the brand and ancillary opportunities without endorsing a purchase or forecasting an absolute share-price decline.
Key Arguments
- Console demand and upgrade incentives may be insufficient.
- Memory costs could pressure margins.
- Console cycles make product mistakes costly for many years.
- Theme parks and films provide incremental benefits rather than the primary earnings engine.
Risks acknowledged
- Nintendo has a trusted family-oriented brand.
- Memory-cost pressure could reverse.
- The lower stock price and roughly 20-times earnings valuation are more interesting.
ADBEAdobeAdobe appears unusually inexpensive to him relative to the stickiness of its products. Leadership and business-model uncertainty help explain market weakness, and his discussion stops short of an explicit purchase or return forecast.
Key Arguments
- An approximately eight-times earnings valuation appears strikingly low.
- He considers Adobe potentially stickier than ServiceNow.
- AI may strengthen established software companies rather than displace them.
Risks acknowledged
- Uncertainty about the next CEO and future business model remains material.
AMZNAmazonHe considers Amazon expensive with slowing growth and exposure to the same capital-spending questions facing Alphabet. He sees useful healthcare infrastructure capabilities but doubts that its general retail relationship replicates specialist healthcare platforms.
Key Arguments
- Heavy AI capital expenditure can pressure cash flow and future returns.
- A general-purpose retailer may struggle to build specialized medical relationships.
- Amazon can still support delivery and healthcare infrastructure.
Risks acknowledged
- His family has successfully used Amazon for a prescription.
- AI could act as a sustaining innovation for established companies.
- He is unsure why the stock is down that day.
He identifies Owlet as an existing holding whose specialized pediatric relationship could support telehealth services. His discussion concerns product opportunities rather than a new purchase, hold recommendation, or stock-return forecast.
Key Arguments
- Child profiles could support ongoing healthcare relationships.
- Telehealth could connect families with prescriptions and local pharmacies.
- Specialization may create a more natural customer relationship than a general retailer.
Risks acknowledged
- The service is still in an early or beta stage.
- The specialist-platform thesis could be wrong.
SPCXSpaceXHe explicitly avoids SpaceX shares because the valuation and future business mix do not meet his fundamental investing standards. He questions satellite economics, recurring capital requirements, and the expansion into AI infrastructure.
Key Arguments
- The sales multiple is extremely high.
- Satellite constellations require repeated replacement spending.
- Terrestrial fiber and 5G limit the addressable market for satellite service.
- The future business mix and competitive moat are unclear.
- He does not see positive free cash flow supporting the valuation.
Risks acknowledged
- Satellite connectivity has applications on aircraft, boats, and connected devices.
- Market enthusiasm for Elon Musk's ideas can differ from fundamental valuation.
Predictions (1)
TSLATeslaHe maintains a long-term bearish Tesla position through puts because the valuation is inconsistent with weak growth and ordinary automotive economics. Slowing robotaxi deployment and competitive alternatives reinforce his view, although he acknowledges the trade's risk and poor initial timing.
Key Arguments
- Tesla is not currently delivering the growth implied by its valuation.
- Its earnings and sales multiples are far above those of General Motors.
- Robotaxi mileage growth appears to have flattened.
- Competing autonomy systems offer redundancy that Tesla lacks.
Risks acknowledged
- Puts are a risky way to express the short thesis.
- His current position has lost money over its longer holding period.
- He had not reviewed the full latest earnings report.
Predictions (1)
He explicitly stays out of Dutch Bros because he lacks conviction in its differentiation and future. His brief response offers little valuation analysis or evidence of expected operational deterioration.
Key Arguments
- He views the product category as ordinary coffee.
- He does not have a compelling investment thesis for the company.
Risks acknowledged
- A stronger consumer endorsement could change his interest.
Predictions (1)
CELHCelsius HoldingsHe previously sold Celsius after concluding that its core growth and competitive position no longer offered the upside he once expected. This describes a completed sale and reduced enthusiasm rather than a new forward trade.
Key Arguments
- Core growth has disappointed his expectations.
- The energy-drink category is becoming more competitive.
- The Alani Nu acquisition created a favorable opportunity to realize gains.
Risks acknowledged
- He considers the acquisition a strong deal.
- His investment produced profits.
He cites GE Vernova as another industrial beneficiary of AI infrastructure spending whose long-term sustainability warrants scrutiny. The comment concerns durability of operating tailwinds and does not establish a separate price prediction.
Key Arguments
- Current AI infrastructure demand supplies strong near-term tailwinds.
- He questions whether those conditions can persist long term.
Risks acknowledged
- Near-term demand benefits are real.
NOWServiceNowHe considers ServiceNow increasingly interesting after its valuation decline and thinks the disruption narrative may be overstated. He has not committed to buying and notes that comparable software multiples can fall further.
Key Arguments
- A roughly 20-times earnings multiple is becoming compelling.
- AI disruption may be less severe than the market assumes.
- Other established software businesses trade at considerably lower multiples.
Risks acknowledged
- He does not know whether the multiple could fall to single digits.
- He still needs to review software candidates before adding a position.
Intuit intrigues him as a low-valued incumbent that could use AI to improve tax preparation rather than lose its customer relationship. He explicitly has not bought shares and offers a business thesis without a distinct investment-direction recommendation.
Key Arguments
- The earnings multiple is around 10.
- Existing data relationships could support automated tax preparation.
- Customers may prefer a specialist to a general AI subscription.
Risks acknowledged
- He has not initiated a position.
- The proposed AI service is illustrative.
He considers Lilly expensive but deserving of a premium because of strong execution. Intensifying peptide competition and eventual patent expirations prevent him from seeing a clear asymmetric investment opportunity.
Key Arguments
- GLP-1 and broader peptide markets are becoming highly competitive.
- Future breakthrough therapies may come from competitors.
- Acquisitions can diversify the business ahead of patent expirations.
Risks acknowledged
- Management is executing well.
- Premium valuation can be justified by business quality.
- He does not know whether the shares offer asymmetric upside.
He has remained outside Lemonade because he does not understand its differentiation or see the operating inflection he wanted. He acknowledges the stock's strong performance and that his assessment may be mistaken.
Key Arguments
- The durable competitive moat is unclear to him.
- Operating metrics have not shown the inflection he sought.
Risks acknowledged
- The stock has performed well.
- He may have missed important strengths.
He regards exchange infrastructure as a potentially solid business and finds ICE's valuation reasonable relative to its history. Its moderate growth does not fit his asymmetric portfolio objective, and he raises questions about future blockchain and prediction-market competition.
Key Arguments
- Infrastructure businesses can generate durable economics.
- Forward and trailing earnings multiples appear reasonable.
- Long-term growth has been solid rather than exceptional.
Risks acknowledged
- Blockchain and prediction markets could alter industry economics.
- He does not identify sufficient asymmetric upside.
He lacks a current Nokia thesis and views its weak historical growth as requiring a turnaround case. He asks what the compelling argument is rather than making a recommendation.
Key Arguments
- Historical growth across several periods appears negative.
- He does not identify an obvious catalyst for a turnaround.
Risks acknowledged
- He has not studied the current structure and may be missing relevant developments.
He questions whether pest control is directly threatened by AI and treats capital rotation as a more plausible framing. He supplies no directional stock call.
Key Arguments
- Pest and wildlife control is not an obvious AI-disruption business.
- AI enthusiasm may redirect capital from more stable companies.
Risks acknowledged
- He has not researched the company sufficiently for a firm view.
He sees Cellebrite as an interesting and potentially solid investigative-security business. Its growth and valuation do not offer the particular upside profile he seeks for his portfolio.
Key Arguments
- Its investigative tools address a distinctive customer need.
- A roughly 24-times forward earnings multiple accompanies approximately 15% growth.
Risks acknowledged
- A solid business need not fit his asymmetric investment mandate.
He does not consider Copart an obvious direct AI-disruption story. He suggests capital rotation may explain the viewer's concern but does not provide a stock recommendation.
Key Arguments
- Auto auctions do not have an obvious direct connection to the AI narrative.
- Investor capital may rotate away from stable businesses toward AI infrastructure.
Risks acknowledged
- He has not studied the company closely.
Hedges & Caveats
- He does not make short-term price projections and cannot predict earnings reactions or when market sentiment will change.
- Several holdings could remain dead money for years before their theses are recognized.
- Autonomous-driving commercialization is developing more slowly than anticipated.
- Hims and Joby have substantial execution risk and potentially binary long-term outcomes.
- His Tesla put position is risky and its initial timing was poor.
- He has not fully reviewed several earnings calls and declines firm opinions on businesses outside his expertise.
- AI infrastructure profitability depends on uncertain future demand, pricing, financing costs, and returns on capital.