PerpEquitiesIdeias de trade de podcasts
Calls long e short dos maiores podcasts de finanças do mundo - extraídos de cada episódio novo e acompanhados desde a exibição.
Período dos episódios
Variação após
Apenas indicações explícitas de compra/venda a descoberto, uma por episódio, ativo e direção. Sem proxies, indicações de evitar ou referências tardias. Variações fixadas no horizonte escolhido, antes de funding e custos; não são previsões. A variação ao vivo desde o episódio é separada.
All-In Podcast
– Taxa de acerto
Indicações medidas / elegíveis: 8 / 8
Poucas indicações medidas (mínimo 20)
8 de ago. de 2026 – 20 de set. de 2026
Últimos 10 episódios · dados antigos ficam arquivados
All-In Podcast
Naveen Rao: 4D Computing, AI's Energy Wall & Beating Biology
Ideias · 2
- DatabricksLongRao's prior company MosaicML joined Databricks and now accounts for roughly a quarter of its revenue, showing how fast enterprise AI model-building demand scaled. — Speaker 1–
Por que esse trade
“We decided to actually join forces with Databricks. That was in 2023, and and actually, that's a quarter of the total revenue of of Databricks today.” — Speaker 1
- Rao says his GPU-platforming startup joined Databricks in 2023 and that it represents about a quarter of Databricks' total revenue today.
- That implies rapid revenue contribution from AI infrastructure within the broader Databricks data/AI platform.
- Suggests continued enterprise AI demand tailwind for Databricks and comparable AI infrastructure players.
- Unconventional AILongPrivate AI chip startup claims a first physical dynamical computer with ~1,000x power-efficiency target and a data center rack product within two years. — Speaker 1–
Por que esse trade
“we're within two years of of getting it to a full product... we're building a new data center product first. So it's a whole rack.” — Speaker 1
- Rao says Unconventional AI taped out its first dynamical-computer chip in June and already has results.
- He claims this is orders of magnitude more efficient than GPUs and targets a full data center rack product within two years.
- Private, pre-revenue hardware bet; no listed ticker available at this time.
All-In Podcast
Adam Foroughi, Applovin CEO: Surviving a 92% Drawdown, Ads as ML 1.0 & the $50B Game Ad Market
AppLovin CEO Adam Foroughi discussed the company's transformation from a regression model to a deep-learning advertising platform, its massive addressable market in mobile gaming and e-commerce discovery, and its extreme stock volatility (from a $3.8B market cap back to $250B). The most actionable takeaway is his aggressive share buyback as a key driver of value, alongside strong EBITDA margins and growth into new markets like e-commerce.
Posições · 1
- APPLongAppLovin's deep learning ad platform is scaling into a $50B mobile gaming market and e-commerce, with 84% EBITDA margins and aggressive buybacks driving shareholder value. — Adam Foroughi+6,2% desde o episódio
Por que esse trade
“we kicked off a super aggressive buyback program and and over the since then, I think we bought roughly $6,000,000,000 of the company's stock” — Adam Foroughi
- Foroughi claims AppLovin's ML 2.0 ad model creates discovery and economic expansion for advertisers, growing ad spend from $11B to $20B on its platform.
- High margins and strong cash flow allowed the company to buy back ~$6B of stock, retiring 20-25% of shares, which amplified returns.
- As the stock re-rates, continued execution in e-commerce and mobile gaming should drive further upside.
All-In Podcast
Bill Gurley: Searching for Feynman
Sem calls concretos neste episódio
All-In Podcast
Jared Isaacman: A New Era for NASA and American Space Exploration
Ideias · 5
- SPCXLongIsaacman says the US would be 'seriously challenged' in space without SpaceX — it is NASA's most important launch partner for crew, cargo and heavy missions. — Jared Isaacman+1,2% desde o episódio
Por que esse trade
“SpaceX is our our I mean, they're they're incredible. I mean, they're our most important launch partner. We can't send astronauts to and from the space station without them.” — Jared Isaacman
- Isaacman: 'SpaceX is our most important launch partner' and America can't send astronauts to the ISS without them.
- SpaceX is supplying an Artemis lander test vehicle and is repeatedly credited as the enabler of NASA's plans.
- Resulting view: NASA's accelerating lunar/Artemis cadence is a structural demand tailwind for SpaceX, though it is currently private.
- GOOGLLongSpaceX → As SpaceX's largest outside shareholder, Alphabet is the most direct listed proxy for the NASA-demand tailwind Isaacman describes for SpaceX. — Jared Isaacman+2,0% desde o episódio
Por que esse trade
“SpaceX is our our I mean, they're they're incredible. I mean, they're our most important launch partner. We can't send astronauts to and from the space station without them.” — Jared Isaacman
- Isaacman says the US would be 'seriously challenged' without SpaceX, making SpaceX central to NASA's lunar and launch plans.
- Alphabet holds a well-known minority equity stake in SpaceX (with Google also a SpaceX customer).
- Resulting trade: own GOOGL as a liquid listed way to hold SpaceX exposure; note the stake is small relative to Alphabet's core business.
- Blue OriginLongBlue Origin is one of the two lander companies Isaacman names for Artemis 3, positioning it as a direct beneficiary of NASA's multi-launch lunar campaign. — Jared Isaacman–
Por que esse trade
“Artemis three will launch on SLS into low Earth orbit and rendezvous with lander test vehicles from Blue Origin and SpaceX” — Jared Isaacman
- Isaacman: Artemis 3 will 'rendezvous with lander test vehicles from Blue Origin and SpaceX.'
- He frames the next four years as 'dozens of landers, dozens of rovers' — a broad demand signal to industry.
- Resulting view: a listed spacecraft/defense supplier building Blue Origin's lander should benefit; Blue Origin itself is private.
- AMZNLongBlue Origin → Amazon is the listed parent whose founder funds Blue Origin, offering indirect exposure to NASA's Artemis lander contracts at a trivial market impact. — Jared Isaacman+2,3% desde o episódio
Por que esse trade
“Artemis three will launch on SLS into low Earth orbit and rendezvous with lander test vehicles from Blue Origin and SpaceX” — Jared Isaacman
- Isaacman names Blue Origin as an Artemis 3 lander test-vehicle provider and a key partner in NASA's lunar base plans.
- Blue Origin is wholly owned by Jeff Bezos and funded by Amazon share sales.
- Resulting trade: own AMZN as a listed adjacency to Blue Origin, while noting the exposure is immaterial to Amazon's P&L.
- RKLBLongIsaacman explicitly names Rocket Lab among NASA's key commercial partners and calls for a 'demand signal' of dozens of landers and rovers — a structural tailwind for launch providers. — Jared Isaacman+11,2% desde o episódio
Por que esse trade
“If you're doing exactly what SpaceX, Blue, Rocket Lab, Stoke, ULA, and others are doing in the industry except you're doing it off, you know, 50 year old shuttle hardware” — Jared Isaacman
- Isaacman: if NASA does what SpaceX, 'Blue, Rocket Lab, Stoke, ULA, and others' do with old shuttle hardware, 'you're gonna lose that workforce.'
- He argues NASA should be 'one customer of many' and hand mature services back to industry.
- Resulting trade: a NASA lunar/Artemis cadence with 'dozens of landers, dozens of rovers' broadens the addressable launch market for Rocket Lab.
All-In Podcast
Nick Shirley: Exposing Government Fraud, Suing California & Taking on the Media
This episode centers on journalist Nick Shirley's investigations into government fraud, the California high-speed rail boondoggle (costs ballooning from $33B to an estimated $236B with no track laid), and his lawsuit against California over a law limiting his reporting. The hosts praise his work and compare it to legacy investigative journalism, but no actionable stock, ETF, or tradable asset calls are made by any participant.
Sem calls concretos neste episódio
All-In Podcast
Meta's Dina Powell McCormick: The Case for Data Centers, Backlash, AI Job Boom & Meta's Future
Ideias · 2
- METALongPowell McCormick argues Meta's data-center investment is a net positive that communities want, and its 3.6B daily-user platform plus open-source AI give it a durable distribution advantage. — Dina Powell McCormick+9,2% desde o episódio
Por que esse trade
“we know that 3,600,000,000 people are on our platform every single day. 200,000,000 small business owners operate their business every single day.” — Dina Powell McCormick
- She says Meta pays for its own generation, grid resilience and upgrades, taking burdens off local consumers, and that most states want these investments.
- That dynamic supports continued AI/data-center capex funded by Meta's core ad business on 3.6B daily users.
- Net bullish on Meta as both an AI infra builder and distribution owner.
- EssilorLuxottica (Luxottica / Ray-Ban)LongMeta's glasses 'are so excited about this as a product' with features like conversation focus and translation, and Luxottica is the exclusive partner making the stylish frames. — Dina Powell McCormick–
Por que esse trade
“Well, they're very stylish... we are so excited about this as a product. The engineers have been working on this for ten years.” — Dina Powell McCormick
- She stresses style matters ('if people don't look good in them... they're not going to put them on') and that billions already wear glasses.
- Meta has an exclusive partnership with Luxottica/Ray-Ban to build these wearable AI glasses.
- Success of the glasses category would directly benefit Luxottica as the manufacturing/brand partner.
All-In Podcast
Brad Gerstner: No AI Bubble, Semis Eat the Nasdaq & AI's Take Off Problem
Posições · 1
- Long duration equities broadly (rates-sensitive)EvitarRising rates (hike odds >90%) raise the hurdle rate as AI capex is debt-funded; 5.5% on the 10-year would be 'a big burden' on equities, so he stays 'medium' rather than large. — Brad Gerstner–
Por que esse trade
“If you can earn five and a half or 6% on your money without taking equity risk, then, you know, it's going to be a challenge for stocks.” — Brad Gerstner
- Gerstner claims rate hikes are coming with over 90% probability, and AI data centers are financed with borrowed money, lifting hurdle rates.
- Citing Buffett — rates are to stocks what gravity is to matter — 5.5-6% risk-free is a challenge for equities.
- Result: stay 'medium' positioned, be mentally flexible, and add or cut based on rates and AI lab revenue.
Ideias · 11
- NVDALongNot a bubble: Nvidia trades at only ~14x next year's fully-taxed GAAP earnings while AI capex super cycle continues. — Brad Gerstner+5,4% desde o episódio
Por que esse trade
“Nvidia, trading at 14 times next year's fully taxed GAAP earnings. This is no bubble like it was in 2000.” — Brad Gerstner
- Gerstner claims Nasdaq/SOX/Nvidia all trade well below average multiples and earnings are up 26% driven by AI infrastructure.
- Nvidia is the core GPU supplier of the token makers whose capex-to-FCF nearly matches hyperscaler capex.
- Result: stay invested in the AI trade, with semis as the engine of returns.
- Nasdaq Composite / Semiconductor Index (SOX)LongSemis are 70% of the Nasdaq's return and the index trades below average multiples in an earnings-driven expansion, not multiple expansion. — Brad Gerstner–
Por que esse trade
“Semiconductors are 70% of the Nasdaq's return. 70% of the return.” — Brad Gerstner
- Gerstner claims this is an earnings-driven market expansion with multiple contraction this year.
- Semiconductors are 70% of the Nasdaq's return, tying index performance to the AI capex super cycle.
- Result: the Nasdaq/SOX remains the expression of the AI infrastructure cycle, though concentration is 'both good and bad'.
- DELLLongMakers of the tokens are making the money — infrastructure suppliers like Dell are posting venture-like returns (up 5x in 18 months). — Brad Gerstner+1,4% desde o episódio
Por que esse trade
“we have massive public companies that look like venture capital returns. Dell up five x, Hynix up nine x in just eighteen months.” — Brad Gerstner
- Gerstner claims the token makers capture the profits while token buyers 'go along for the ride'.
- Infrastructure tightness has driven Dell up 5x in just 18 months on AI server demand.
- Result: own the infrastructure makers, not just the model buyers.
- SK HynixLongMemory supplier Hynix is up 9x in 18 months as tight AI infrastructure demand flows nearly dollar-for-dollar from hyperscaler capex into semis' free cash flow. — Brad Gerstner–
Por que esse trade
“Dell up five x, Hynix up nine x in just eighteen months.” — Brad Gerstner
- Gerstner claims hyperscaler capex is almost dollar-for-dollar free cash flow to infrastructure companies.
- Hynix is cited as a massive public company that looks like a venture return, up nine-fold in 18 months.
- Result: memory/infrastructure suppliers remain key AI trade beneficiaries.
- ANTHROPICLongAnthropic's monthly revenue surging from $2B to $11B (year-end run-rate seen at $100-110B) is the fuse that lit the AI trade and could drive liftoff and an IPO. — Brad Gerstner+1,6% desde o episódio
Por que esse trade
“Anthropic's monthly revenue. ... And in March it was 11,000,000,000. ... Anthropic IPO ... I think we could go higher” — Brad Gerstner
- Gerstner claims the March jump in Anthropic revenue to $11B lit the historic April-May run in AI stocks.
- Anthropic reportedly reaches ~$100-110B this year on only ~1.5GW of compute, and adding 4-5GW supports another $100B of revenue.
- Result: watch Anthropic's monthly revenue as the single most important data point; an Anthropic IPO this year would signal liftoff.
- MSFTLongHyperscalers like Microsoft build AI capacity 'to rent it', so their fortunes are tied to the off-take revenue that must show up to pay for ~$1.5T annual capex. — Brad Gerstner+1,6% desde o episódio
Por que esse trade
“Microsoft's not paying for it. They're building it to rent it. ... We have to have the off take revenues in order to pay that rent.” — Brad Gerstner
- Gerstner claims Microsoft, Google and Amazon aren't paying for capex — they build to rent it — so renters' off-take revenue must materialize.
- Microsoft is named as a hyperscaler building capacity to rent, and is OpenAI's key partner (MAIA/Azure).
- Result: hyperscalers are levered to AI lab revenue growth; a lab-revenue surge or oil/rate retreat justifies adding chips.
- UBERLongAI-driven margin expansion: Uber says it will grow 20% without growing headcount, converting AI into operating leverage. — Brad Gerstner-0,6% desde o episódio
Por que esse trade
“Uber says we're gonna grow 20%, we're not gonna grow headcount.” — Brad Gerstner
- Gerstner claims AI can lift margin expansion from ~38bps to ~100bps for Nasdaq companies.
- Uber is cited saying it will grow 20% without adding headcount, the largest cost input (humans/engineers) staying flat.
- Result: own companies showing AI-enabled productivity dividends.
- SNOWLongAI margin expansion: Snowflake expects 30% growth without headcount growth, a direct example of the productivity dividend. — Brad Gerstner+1,7% desde o episódio
Por que esse trade
“Snowflake says we're gonna grow 30%, we're not gonna grow headcount.” — Brad Gerstner
- Gerstner claims the AI productivity dividend can turn 38bps of annual margin expansion into 100bps.
- Snowflake is cited saying it will grow 30% without growing headcount.
- Result: companies monetizing AI internally without adding engineers offer margin-expansion upside.
- OPENAILongOpenAI's monthly revenue trajectory (is it $4B or $8B?) is the second key datum alongside Anthropic's; strong numbers mean 'takeoff' for the AI trade. — Brad Gerstner+10,8% desde o episódio
Por que esse trade
“Anthropix monthly revenue is OpenAI's monthly revenue gonna be 4,000,000,000 or 8,000,000,000? ... I think it's takeoff.” — Brad Gerstner
- Gerstner claims the top-three labs (Anthropic, OpenAI, SpaceX) collectively run ~$100B and need to reach ~$180B by year-end to keep the AI trade intact.
- If OpenAI's monthly revenue comes in near $8B, Gerstner calls it 'takeoff' and expects an IPO this year.
- Result: watch OpenAI monthly revenue as a key trigger; add chips if it's strong, reduce if not.
- SPCXLongAmong the top-three AI labs by run-rate revenue, SpaceX's valuation is up two-and-a-half-fold in a nasty backdrop — evidence of the AI-driven private-market boom. — Brad Gerstner+0,4% desde o episódio
Por que esse trade
“OpenAI and Anthropix valuation up two x, SpaceX up two and a half x in a pretty nasty backdrop.” — Brad Gerstner
- Gerstner groups SpaceX with Anthropic and OpenAI as the top-three labs with ~$100B collective run-rate revenue.
- He cites SpaceX valuation up 2.5x this year despite tariffs, geopolitics and regulation worries.
- Result: private AI-adjacent leaders continue to compound; watch their revenue and IPO activity.
- BitcoinEvitarDespite many Besties predicting gold and Bitcoin would soar, Bitcoin is down 10% this year while AI-driven earnings market works — the AI trade is where the money is. — Brad Gerstner–
Por que esse trade
“We have a lot of people in the Bestie group who said gold was gonna be off the charts this year. It's flat. Bitcoin's down 10%.” — Brad Gerstner
- Gerstner claims the scoreboard shows gold flat and Bitcoin down 10% in a year the market rose 39% since Jan.
- The AI super cycle, not crypto, has generated the returns (Nvidia revenue 2x, hyperscaler capex 2x).
- Result: don't chase the gold/Bitcoin narrative; stay in the earnings-driven AI trade.
All-In Podcast
JD Vance on AI, Entitlement Fraud, Iran War, Israel, H-1B Abuse & the Midterms
Vice President JD Vance appeared on the All-In Podcast to discuss the administration's first two years, defending the conflict with Iran as necessary to protect global energy markets, pushing back on AI data-center backlash by advocating massive power buildout, and highlighting H-1B reform and entitlement fraud savings ($250B). Market-relevant topics include energy supply security, AI infrastructure and power demand, and fiscal deficit concerns, but no specific stocks, ETFs, or tradable assets were recommended or given directional views by any speaker.
Sem calls concretos neste episódio
All-In Podcast
Satya Nadella on the AI Doomer Slowdown, Microsoft's Master Plan & Who Wins AI
Ideias · 4
- MSFTLongNadella argues AI's value shifts to the app and middleware layer, where Microsoft's 30M+ paid Copilot seats and enterprise harness position it to capture margin as model prices collapse. — Satya Nadella+0,2% desde o episódio
Por que esse trade
“The apps are gonna become much more viable economically, which is great for the ecosystem.” — Satya Nadella
- Token-price competition (open-weight models at a fraction of frontier pricing) compresses the model layer's take rate.
- He says that dynamic makes app and middleware layers 'much more viable economically,' and he cites real enterprise Copilot workflows with 30M+ seats.
- MSFT owns the M365 knowledge-work base (450M users) and the harness/middleware layer, so it is the listed beneficiary of this shift.
- MSFTLongNadella frames Microsoft's OpenAI investment and IP access as a core asset while Azure builds a $80B+ multi-customer AI cloud for the long tail, not just two model labs. — Satya Nadella+0,2% desde o episódio
Por que esse trade
“I'm going to spend $80,000,000,000 building out Azure.” — Satya Nadella
- He says he is 'good for' $80B building out Azure and that the company is renting capacity because it is short on supply.
- He stresses Azure must serve many third parties and its own first-party estate, with OpenAI as one of its largest customers.
- That demand-driven hyperscale model implies durable Azure revenue tied to the AI buildout.
- OPENAILongNadella repeatedly calls the OpenAI investment a source of pride and says Microsoft will keep using its IP, while noting frontier model companies can manage token pricing and 'will do fine.' — Satya Nadella+12,1% desde o episódio
Por que esse trade
“We're thrilled about, obviously, our investment in OpenAI, the access we have to their IP, which we have for a long time.” — Satya Nadella
- He says Microsoft is 'thrilled about, obviously, our investment in OpenAI' and its long-term IP access.
- He argues model companies retain pricing power via model families even as open-source checks emerge.
- The implication is a constructive stance on the private company itself, separate from Microsoft's equity stake.
- NVDALongNadella says Nvidia remains Microsoft's primary AI compute platform while adding heterogeneous silicon, implying continued core demand for Nvidia's accelerators even as the mix diversifies. — Satya Nadella+7,4% desde o episódio
Por que esse trade
“We have Jensen stuff, which is I think our primary thing.” — Satya Nadella
- He describes Microsoft's kit as short-term, demand-driven assets (racks, chips) representing ~60% of cost.
- He says 'We have Jensen stuff, which is our primary thing' alongside AMD and custom chips.
- Microsoft's continued hyperscale inference/training build therefore supports Nvidia demand even amid diversification.
All-In Podcast
Elon Musk & Gwynne Shotwell on AI Risks and Peer Review, Starship, Terafab, SpaceX/Tesla Merger
Ideias · 2
- ANTHROPICLongMusk says Anthropic puts more care into safety than OpenAI and that peer-review testing will make unsafe releases legally and reputationally costly, favoring the safer lab. — Speaker 0 (Elon Musk)+0,9% desde o episódio
Por que esse trade
“on balance, I think Anthropic is... puts more care into their safety than than OpenAI” — Speaker 0 (Elon Musk)
- Musk states Anthropic is better on safety than OpenAI and that both sets of models are 'quite close in capability.'
- If peer review exposes dangerous releases, liability and 'egg on face' risk shifts to the labs that ignore warnings.
- Anthropic, perceived as the more careful lab, benefits reputationally and potentially in enterprise adoption.
- GOOGLLongMusk lists Google among the AI companies whose test harnesses would add unique heterogeneous coverage in peer review, implying its model stack is competitive and integral to the emerging AI safety ecosystem. — Speaker 0 (Elon Musk)+2,5% desde o episódio
Por que esse trade
“SpaceX is running its test harness and Google and Meta were doing that” — Speaker 0 (Elon Musk)
- Musk argues multiple heterogeneous test harnesses (SpaceX, Google, Meta, Chinese firms) would be needed for effective peer review.
- That framing implicitly includes Google as a leading frontier player expected to participate in the standard.
- If peer review becomes the de facto standard, leading labs like Google are positioned to shape and benefit from it.
Extração automática de falas de apresentadores e convidados - não é recomendação de investimento, precisão não garantida.