PerpEquities

Trade ideas

Podcast trade ideas

Long and short calls from the world's biggest finance podcasts - extracted from every new episode and tracked from the moment it aired.

Episode period

Return after

Explicit long/short calls only, once per episode, asset and direction. Excludes proxies, avoid calls and late price captures. Returns are fixed after the selected horizon, before funding and trading costs; not a forecast. The live move since an episode is separate.

All-In Podcast
– Hit rate
Measured / eligible calls: 14 / 22

Too few measured calls (minimum 20)

Aug 29, 2026 – Nov 14, 2026

Latest 10 episodes · older records remain archived

All-In Podcast

Jensen Ousted as Nvidia CEO: "Spy" Scandal, Export Restrictions, Market Impact

Nov 14, 202636h

Positions · 1

  • ANTHROPICLongI think Anthropic is building the Microsoft killer and every SaaS company killer by bundling — Speaker 00.0% since episode
    Why this trade
    “I think Anthropic is building the Microsoft killer and every SaaS company killer by bundling” — Speaker 0
    1. Jason argues Anthropic is rapidly bundling enterprise/SaaS products onto a $20-50/month subscription, making the offering stickier and stealing share from incumbents.
    2. The IPO raises a huge war chest (~$90B) and is 6x oversubscribed, with analysts modeling a bull case far above the $1.4T target.
    3. If the winner-take-all recursive-self-improvement story is even partly true, allocating to the IPO is a once-in-a-generation opportunity.

Ideas · 5

  • NVDAShortWith Jensen gone, Nvidia loses its 'velocity' — the annual chip-tapeout tempo that is its real moat — and the 'Bank of Nvidia' capital engine that funds the AI economy may slow or stop. — Speaker 1-1.0% since episode
    Why this trade
    “if you take him out of the picture, there's just nobody comparable” — Speaker 1
    1. Sacks argues Nvidia's advantage is rapid annual iteration (Ampere→Hopper→Blackwell→Vera Rubin), and Jensen was fundamental to that pace.
    2. Chamath/Friedberg note Jensen personally directed capital/guarantees to startups and data centers ('the beating heart of the AI economy').
    3. Without him, growth and the AI capex flywheel slow, pressuring the stock and broader AI trade.
  • PANWAvoidChamath names CEO Nikesh Arora as a top external Nvidia CEO candidate ('best capital allocator of the last five or ten years'), so Palo Alto faces key-man/leadership risk if he's poached. — Speaker 3-0.4% since episode
    Why this trade
    “the most obvious external candidate is Rene Haas from Arm and Nikesh Arora from Palo Alto Networks” — Speaker 3
    1. Chamath says the obvious external CEO candidate for Nvidia is Nikesh Arora from Palo Alto Networks.
    2. Arora is described as the standout public-market capital allocator of the past decade, central to Palo Alto's execution.
    3. His potential departure is a negative headline/leadership risk for PANW even without a direct sell call.
  • ARMAvoidChamath flags Arm CEO Rene Haas as an obvious external candidate to run Nvidia, creating poaching/key-man risk for Arm at a moment it can ill-afford leadership churn. — Speaker 3-0.7% since episode
    Why this trade
    “Rene, because he worked at NVIDIA for many, many years, worked directly for Jensen and has been running ARM successfully” — Speaker 3
    1. Chamath names Rene Haas as a leading external Nvidia CEO candidate given his long tenure at Nvidia under Jensen.
    2. Haas has been 'running ARM successfully' and is steeped in chips, making him attractive.
    3. Losing him would be a negative for Arm's momentum, so avoid until the succession resolves.
  • MSFTLongJason argues Microsoft, the historic bundler, now must build a competing product and go head-to-head with Anthropic or risk losing the entire Office franchise. — Speaker 0-0.6% since episode
    Why this trade
    “Microsoft has to create a competing product and go head to head against them or they're gonna lose the Office franchise” — Speaker 0
    1. Jason says Anthropic is bundling meetings, sheets, legal and finance — the core Office/SaaS surface — into one subscription.
    2. He explicitly says this means Microsoft 'has to create a competing product and go head to head against them or they're gonna lose the Office franchise.'
    3. Microsoft's huge Office franchise is at risk, forcing a competitive response that frames the stock's outlook.
  • Chinese AI chips / HuaweiLongSacks argues that if Nvidia is slowed for 2-3 years, Huawei catches up far faster, especially as China's government forces its market toward Huawei chips (latest DeepSeek trained exclusively on Huawei). — Speaker 1–
    Why this trade
    “If Nvidia is massively slowed down ... it's mostly will help Huawei” — Speaker 1
    1. Sacks describes Huawei closing the frontier gap via CloudMatrix rack-scale networking (384 chips) and energy advantage.
    2. He says if Nvidia is 'crippled for two or three years,' Huawei will catch up much faster; the CCP wants to indigenize chips regardless.
    3. Watch Huawei-relevant suppliers/foundries rather than a single US-listed ticker — a theme-level long.

All-In Podcast

Freedom vs. Safety: Axon CEO Rick Smith on the Future of Policing

Nov 10, 2026112h

Axon CEO Rick Smith recounts building TASER into a public-safety technology platform, with body cameras and cloud software driving the bulk of the company's growth. The most actionable market view is Axon's near-50% investment in drone maker Skydio alongside exclusive distribution rights, which Smith frames as one of the biggest future parts of policing; he also notes a Ring (Amazon) partnership for video sharing and Skydio's growing role. No outright buy/sell recommendation on a stock is made, so these are discussion-level views.

Positions · 1

  • SkydioLongSmith says Axon bought almost half of Skydio because he believes millions of drones will be deployed in the US and that drones will arrive on scene far faster than people. — Speaker 1–
    Why this trade
    “We we bought almost half the company. We we wanted to buy about as much as we could because we're big believers that there's gonna be millions of drones” — Speaker 1
    1. Smith claims drones will be one of the biggest parts of the future of policing, arriving before officers.
    2. Axon invested almost half the company and secured exclusive rights as the channel into its markets.
    3. Result: a bullish view on Skydio as the leading US drone maker, held privately with Axon as largest investor.

Ideas · 2

  • AXONLongSmith argues Axon's mission-driven device-plus-cloud platform (taser, body cams, counter-drone, drone/robot response) makes it the operating system for public safety and the durable winner as policing technology scales. — Speaker 1+3.7% since episode
    Why this trade
    “virtually, I think every major police department now wears body cameras and we have... The vast majority are on our system” — Speaker 1
    1. Smith says virtually every major police department wears body cameras and most are on Axon's system, with over a million officers uploading video daily.
    2. That recurring cloud/evidence business plus new drone and robotics offerings extend Axon's platform and revenue per agency.
    3. Result: a long view on Axon as the scaled public-safety technology platform.
  • AMZNLongRing → Smith notes Ring's cameras remain popular for home protection and crime video sharing despite privacy backlash, and Axon partners with Ring on safeguards for sharing footage with police. — Speaker 1-3.2% since episode
    Why this trade
    “they go buy Ring cameras specifically because Ring has some of the best technology for being able to protect your home” — Speaker 1
    1. Smith says people still buy Ring cameras because of the technology and ability to share video with police after a crime.
    2. Ring is owned by Amazon, which gains from continued consumer camera adoption and the Axon partnership.
    3. Result: a mild positive read-through to Amazon's Ring business.

All-In Podcast

Midterm Blood Bath, $4T AI Sell Off, Data Center Freeze, What's Next for Trump

Nov 7, 2026188h

The hosts dissect the Democrats' midterm sweep, attributing it to the Iran war's inflation impact and a broken economy for wage earners, and debate how Trump can recover. The most actionable market view is that data-center permitting uncertainty after state-level regime change benefits hyperscalers (which own power and capacity) at the expense of neoclouds and frontier labs, while Jason Calacanis says he has been buying SpaceX and Tesla shares. Sacks expects meaningful capex in 2027 but a greatly reduced pipeline by 2028, and Chamath flags SpaceX's compute deal as a potential kingmaking lever.

Positions · 2

  • SPCXLongCalacanis says he has kept buying SpaceX (and Tesla) through the selloff, arguing space-based data centers make SpaceX strategically undervalued as compute migrates offshore. — Jason Calacanis+2.4% since episode
    Why this trade
    “SpaceX, which is severely undervalued as a stock and as SpaceX, Tesla. I've been buying shares through this whole insanity.” — Jason Calacanis
    1. Calacanis: 'SpaceX, which is severely undervalued as a stock... I've been buying shares through this whole insanity.'
    2. He argues data centers can go anywhere, including to space, making SpaceX's launch/orbit capability the hedge against permitting bans.
    3. Resulting trade: stay long SpaceX ahead of its IPO and any orbital data-center buildout.
  • TSLALongCalacanis lumps Tesla with SpaceX as names he has been buying during the AI/data-center selloff, framing it as tied to the space-based compute idea. — Jason Calacanis+7.7% since episode
    Why this trade
    “SpaceX, which is severely undervalued as a stock and as SpaceX, Tesla. I've been buying shares through this whole insanity.” — Jason Calacanis
    1. Calacanis says he has been buying shares of SpaceX and Tesla through the whole insanity.
    2. He ties the thesis to SpaceX putting data centers in space, where Tesla/Elon assets are central.
    3. Resulting trade: long Tesla as part of this Elon-complex positioning.

Ideas · 4

  • Hyperscalers (Amazon, Meta, Microsoft, Google)LongChamath argues data-center slowdowns benefit the hyperscalers because they own existing capacity and power, letting them become kingmakers and advantage their own models. — Chamath Palihapitiya–
    Why this trade
    “I think what that does is it benefits the hyperscalers because they have existing capacity. I think it's very bad for the Neo Clouds because they're renters.” — Chamath Palihapitiya
    1. Chamath: 'I think what that does is it benefits the hyperscalers because they have existing capacity.'
    2. Hyperscalers (Amazon, Meta, Microsoft, Google, SpaceX) own critical power and can reprice or co-opt renter capacity.
    3. Resulting trade: prefer hyperscaler exposure over neocloud/frontier-lab tenants as permitting uncertainty bites.
  • Neoclouds (GPU/data-center renters)AvoidChamath warns the 'renter class' in data centers is very much at risk as hyperscalers co-opt capacity and permitting slows, citing Google taking over a Crusoe data center. — Chamath Palihapitiya–
    Why this trade
    “the renter class in data centers as of Tuesday is very much at risk... Google just took over a Crusoe data center. They co opted the renter.” — Chamath Palihapitiya
    1. Chamath: 'the renter class in data centers as of Tuesday is very much at risk.'
    2. Hyperscalers with power are co-opting renters (Google taking over a Crusoe data center), squeezing neocloud supply and economics.
    3. Resulting trade: avoid neocloud/renter exposure; favor owners of power and capacity.
  • ANTHROPICAvoidChamath argues frontier labs like Anthropic are dangerously short of compute and could be repriced by SpaceX's 90-day termination clause, making them the vulnerable renter class. — Chamath Palihapitiya-3.2% since episode
    Why this trade
    “If SpaceX says I'm gonna terminate this compute in ninety days, That is the king making move... And Ant would have no choice except to immediately capitulate.” — Chamath Palihapitiya
    1. Chamath: SpaceX could terminate Anthropic's compute deal in 90 days at a spot rate 'probably double... could probably triple', forcing capitulation.
    2. Frontier models are renters lacking fortress balance sheets or infinite compute and need capacity to grow.
    3. Resulting trade: avoid/short thesis on frontier-lab economics versus hyperscaler owners.
  • AI data-center developers (fully permitted projects)AvoidSacks expects capex to hold in 2027 for fully permitted projects but warns the data-center pipeline is likely closed or greatly reduced by 2028 as governors block new permits. — David Sacks–
    Why this trade
    “I think by 2028, I would expect that the pipeline will either be closed or significantly reduced and that will have a big economic impact.” — David Sacks
    1. Sacks: 'by 2028, I would expect that the pipeline will either be closed or significantly reduced.'
    2. New governors signal no new permits until legislation, and no politician has incentive to approve data centers.
    3. Resulting trade: be cautious on multi-year data-center pipeline exposure.

All-In Podcast

Anthropic S-1 Shocker, Greatest Business Ever? Google's $500B Wipeout, Midterm Predictions

Oct 30, 2026377h

The hosts dissect Anthropic's S-1 ahead of a planned ~$100B IPO at a ~$1.5T valuation, with Chamath calling it the most incredible business he's seen while flagging systemic risk from the AI complex crowding out Treasury markets. Jason buys more Alphabet shares during the show, calling the post-earnings drop oversold, and Sacks says Google looks a bit oversold but declines to recommend buying anything.

Positions · 2

  • GOOGLLongGoogle is oversold after a 10-15% post-earnings drop; search still grows 12% and Cloud doubled, and Alphabet owns stakes in SpaceX, Anthropic and Waymo. — Jason Calacanis+3.4% since episode
    Why this trade
    “I just literally bought some shares. So I'm long. I just bought more shares of Google while doing the show.” — Jason Calacanis
    1. Sacks notes search revenue rose $7B YoY and Cloud doubled, so the selloff is overdone.
    2. Jason buys shares on air, citing diversifying revenue plus ~6% SpaceX and ~15% Anthropic stakes.
    3. Resulting trade: add to GOOGL on the weakness.
  • ANTHROPICLongChamath calls the S-1 the most incredible business he has seen and sees a constructive read into the IPO, though market risk tempers aggressive pricing. — Chamath Palihapitiya-21.5% since episode
    Why this trade
    “This is the most incredible business I've ever seen.” — Chamath Palihapitiya
    1. Anthropic grew revenue 14x YoY with ~85% gross margin and huge cash, an unusually capital-efficient AI business.
    2. Chamath frames the IPO as fairly priced around $1.05 where it offers an entry into future upside.
    3. Resulting trade: constructive stance around the IPO, with systemic market risk flagged.

Ideas · 1

  • GOOGLLongSacks says Google is a little bit oversold because search kept growing and Cloud doubled, so the selloff looks overdone. — David Sacks+3.4% since episode
    Why this trade
    “I tend to think Google's a little bit oversold on this news today.” — David Sacks
    1. Search revenue rose from $56B to $63B YoY while Cloud roughly doubled.
    2. The share shift below 50% reflects faster-growing segments, not search deterioration.
    3. Resulting view: Google looks oversold, though he stops short of an explicit buy.

All-In Podcast

WHOOP CEO, Will Ahmed: 50B Hours of Health Data, a 4 Million Member Data Moat, & the FDA Fight

Oct 29, 2026401h

Positions · 2

  • AAPLAvoidCalacanis argues Apple's watch and especially Apple Health software are poor experiences and that Apple is a competitor threat to WHOOP, making it a weaker wearables-health play. — Jason Calacanis+1.4% since episode
    Why this trade
    “Because Apple's gonna make a competitor. They know their watch kinda sucks, and their software sucks. Like, Apple Health is terrible.” — Jason Calacanis
    1. Host repeatedly criticizes Apple Watch UX (constant alerts, battery) and calls Apple Health 'terrible'.
    2. Apple monetizes via expensive hardware and, per the host, will build a competitor to WHOOP.
    3. Implication favors specialists like WHOOP over Apple in health/recovery tracking, a negative read-through for AAPL's health moat.
  • WHOOPLongCalacanis endorses WHOOP as a product and business, citing 4M members, strong retention, $15-30/month recurring revenue and a large physiological data moat. — Jason Calacanis–
    Why this trade
    “Great product. I give it my highest endorsement, Eight Sleep. I'm an investor in. I love that product too. Those are my two favorite health products.” — Jason Calacanis
    1. Host says WHOOP is one of his two favorite health products and gives it his highest endorsement.
    2. CEO cites 4M members, subscription model, 50B+ hours of data and 80% member AI interaction.
    3. Private company — thesis is a venture-endorsement, not a tradable listed call.

All-In Podcast

AI Cyber Panic, Ken Griffin's Bubble Warning, The Fall of France?

Oct 23, 2026543h

Positions · 1

  • France (French government bonds)ShortChamath says bond vigilantes should take French 10-year yields to ~8% to 'break the back' of the government, forcing pension/immigration/spending reform. — Chamath Palihapitiya–
    Why this trade
    “I think you need to take France to, like, 8% and just break them” — Chamath Palihapitiya
    1. He argues France spends/borrows more than any communist state ever did with broken outcomes.
    2. Government won't self-reform, so higher yields are the only forcing mechanism.
    3. Resulting trade: short French debt / expect yields toward 8% to force the reset.

Ideas · 7

  • PANWLongArora argues cyber budgets (4-10% of IT) must rise because AI-driven vulnerability hunting makes breaches persistent, and perimeter/virtual-patch products like Palo Alto's are the first line of defense. — Nikesh Arora-14.1% since episode
    Why this trade
    “you can actually put it in a perimeter defense product like ours or Cloudflare has one” — Nikesh Arora
    1. Arora says AI finds exploits far faster than every ten days and enterprises carry unpatched tech debt.
    2. He highlights perimeter defenses (Palo Alto, Cloudflare) as the virtual-patch layer any attacker hits first, implying rising security spend.
    3. Resulting trade: long cyber vendors best positioned to capture that budget growth.
  • NETLongArora names Cloudflare's perimeter defense as a virtual-patching layer absorbing the surge in attacks (millions of attempts), implying share of rising cyber spend. — Nikesh Arora-6.9% since episode
    Why this trade
    “Cloudflare has one, or other firewalls have one. You put the patch there. It's a virtual patch” — Nikesh Arora
    1. He describes massive hacking attempts during the PHP crisis that flow through CDNs first.
    2. Arora explicitly cites Cloudflare as a perimeter product that can block the vulnerability.
    3. Resulting trade: CDN/edge security names benefit as enterprises shift spend to proactive defense.
  • CRWVAvoidReferenced as a data-center/AI-capex name down 10% as AI spending concerns hit, illustrating how capex-dependent AI infrastructure stocks swing on bubble fears. — Jason Calacanis-6.7% since episode
    Why this trade
    “people who are either making data centers like CoreWeave down 10%” — Jason Calacanis
    1. Chamath warns AI spend is circular and concentrated in money-losing startups that need to diffuse.
    2. Data-center builders like CoreWeave are most levered to that concentrated AI capex.
    3. Resulting trade: caution/avoid on pure AI-capex plays until spending broadens.
  • METAAvoidFlagged as down 6% after earnings as investors question huge AI spend without clear productivity return—the exact 'where is the productivity' worry Griffin raised. — Jason Calacanis-1.2% since episode
    Why this trade
    “having a huge AI spend, Meta, which is an extraordinary one there, it's down 6% after earnings” — Jason Calacanis
    1. Griffin says companies struggle to convert AI spending into productivity.
    2. Meta is cited as a massive AI spender punished on earnings.
    3. Resulting trade: near-term caution on heavy AI-capex spenders.
  • AMDAvoidNamed among AI-linked names down ~5% on the data-center/capex pullback, reflecting sensitivity to AI spending sentiment. — Jason Calacanis+0.3% since episode
    Why this trade
    “AMD down 5%” — Jason Calacanis
    1. Chamath argues AI spending must diffuse beyond startups to sustain the trade.
    2. AMD is a chip/capex beneficiary levered to that spending.
    3. Resulting trade: near-term caution as the concentrated AI capex cycle is questioned.
  • NVDALongArora notes the profit pool 'lives in NVIDIA and the memory industry' and that demand is insatiable with chips sold out to 2028—structural tailwind despite diffusion concerns. — Nikesh Arora-24.3% since episode
    Why this trade
    “they all live in NVIDIA and and the memory industry... There's not enough compute in the world to satisfy us” — Nikesh Arora
    1. Arora says Jensen can't deliver chips until 2028 and demand exceeds supply.
    2. Today's AI profit pool sits in NVIDIA and memory; compute scarcity persists.
    3. Resulting trade: long NVIDIA as the scarce compute supplier through the buildout.
  • SPCXLongJason cites Elon gifting every American child a SpaceX share at ~$150-170, framing SpaceX equity as an appreciating 'Invest America' asset and model for forced long-term ownership. — Jason Calacanis-6.2% since episode
    Why this trade
    “every American child... a share of SpaceX worth a 150, 160, $170 it's been trading at, which is a $10,000,000,000 gift” — Jason Calacanis
    1. Musk is reportedly gifting one SpaceX share to each American child (~$10B).
    2. Jason argues forcing long-term equity ownership makes people capitalists and creates value.
    3. Resulting trade: constructive on SpaceX equity as an appreciating private-market asset.

All-In Podcast

Arvind Krishna & Nikesh Arora: Quantum Threat, Open vs Frontier AI, AI Cyber Warfare & SaaS Collapse

Oct 21, 2026591h

Nikesh Arora and Arvind Krishna argue that post-quantum cryptography must be adopted now because IT stacks are broadly breakable by ~2029-30, a spending tailwind for infrastructure/security. They expect open-weight models to capture 75-90% of tokens eventually, front-running significant disruption in the SaaS application layer, while Nvidia is cited as the modern vertically integrated control point. Arora explicitly says cybersecurity valuations are elevated but rationalizing, and broadly signals caution on software/SaaS terminal value, even as he expects cyber consolidation via M&A.

Positions · 2

  • NVDALongArora calls Nvidia the modern vertically integrated stack, letting it move profit between layers as competition shifts. — Nikesh Arora-27.3% since episode
    Why this trade
    “the modern day version of that is NVIDIA... he's a modern day version of the vertically integrated stack” — Nikesh Arora
    1. Arora frames NVDA as the contemporary vertical integrator controlling the full stack and funding its ecosystem.
    2. Control of the stack lets it commoditize layers and shift profit pools, preserving leverage.
    3. Treat as an explicit long on the vertical-integration advantage.
  • Software / SaaS sectorAvoidArora warns terminal value of software is unclear and that huge SaaS disruption is coming as product-market fit gets upended by AI. — Nikesh Arora–
    Why this trade
    “there's gonna be huge disruption in the SaaS space over the next few years” — Nikesh Arora
    1. Arora says it's very hard to understand the terminal value of software.
    2. AI-first rearchitecture means incumbents' product-market fit is being disrupted by newcomers.
    3. Conclude there will be massive disruption in SaaS over the next few years; avoid broad software exposure.

Ideas · 2

  • PANWLongArora says customers shouldn't stitch 40 cyber tools and is consolidating the industry via M&A, supporting platform winners. — Nikesh Arora-5.6% since episode
    Why this trade
    “we're working as fast as we can to take disparate parts of our industry and find the way to stitch them together” — Nikesh Arora
    1. Arora describes consolidating disparate security point solutions into one platform for customers.
    2. He notes cyber currently enjoys better valuations because of the AI moment.
    3. Arora is discussing his own company's strategy rather than issuing a fresh buy call, so view stays implicit.
  • Post-quantum cryptography infrastructureLongKrishna and Arora argue post-quantum readiness is a now-mandated IT refresh, forcing enterprises to replace non-quantum-safe stacks. — Nikesh Arora–
    Why this trade
    “You should be replacing your IT stack today with quantum safe or quantum proof infrastructure right now” — Nikesh Arora
    1. Krishna says quantum could break today's cryptography around 2030 and post-quantum code already exists.
    2. Arora says ~50% of IT infrastructure isn't quantum safe and must be replaced gradually, not overnight.
    3. This implies a multi-year enterprise security/IT refresh tailwind, though no specific ticker is named.

All-In Podcast

Dylan Field: Is AI Replacing Designers?, Design is the New Code & Democratizing Creativity

Oct 19, 2026642h

Dylan Field (Figma CEO) argues that as AI commoditizes coding, design, brand and point-of-view become the differentiator ('design is the new code'), and that the SaaS-apocalypse thesis is too binary — Figma's business accelerated to ~48% YoY growth after Anthropic's Claude Design launch. Market-relevant views are limited: he is constructive on Figma (FIG) and complimentary of Cursor as a user-obsessed software winner, while noting OpenAI's math breakthroughs; no explicit buy/sell calls were made.

Ideas · 3

  • FIGLongAI commoditizes code but raises the value of design and brand — exactly Figma's core — and the business kept accelerating despite the Claude Design scare. — Speaker 0–
    Why this trade
    “You guys grew 48% year over year, and the business basically accelerated its growth.” — Speaker 0
    1. Field argues 'design is the new code' as code gets cheap, differentiation shifts to design/brand and point of view.
    2. Figma is the core design-and-build tool for the world's best product teams, so that shift accrues to its workflows and pricing power.
    3. He notes growth accelerated to 48% YoY after Claude Design, undercutting the 'Figma is dead' narrative — a constructive long stance.
  • Anysphere (Cursor)LongCursor 'should not have grown the way they did' against frontier labs, because they obsessed over the user — a repeatable playbook for software winners. — Speaker 1–
    Why this trade
    “Cursor, like, by all means, should not have grown the way they did... They've had an incredible success.” — Speaker 1
    1. Field says frontier labs chased Cursor with their own models, yet Cursor out-executed by focusing on customer needs.
    2. He calls the outcome 'incredible success' and a template that will be repeated across companies.
    3. Implies durable value in user-obsessed application layers versus raw models — supportive but not an explicit buy call.
  • OPENAILongOpenAI's 700+ paper math drop was 'factually' the greatest day of discovery, a show of frontier capability edge in formal domains. — Speaker 1-15.6% since episode
    Why this trade
    “the math drop last night from OpenAI of 700 plus papers on all sorts of very famous problems.” — Speaker 1
    1. Field describes the OpenAI math release as a 'show of something' well beyond prior model progress.
    2. Speaker 0 calls it the greatest day of discovery in human history; Field agrees 'factually, yes.'
    3. Frontier model leadership in verifiable domains (math, cybersecurity) reinforces OpenAI's position — implicit positive, not a trade recommendation.

All-In Podcast

CoreWeave Crash, AI Trade Cracks, Anthropic IPO Red Flags, Bill Gates Doomer

Oct 17, 2026709h

The hosts dissect the CoreWeave-led selloff in neoclouds, with Chamath warning the name sits near a breakeven point if debt costs keep rising while Sachs and Ben Horowitz argue the demand backdrop remains intact, and Ben explicitly says A16Z avoids the neoclouds. Chamath makes the clearest action call—'get long energy'—while Sacks and Chamath frame OpenAI and Anthropic as near-mandatory defensive hedges and discuss Anthropic's looming IPO. Bill Gates' AI-doomer stance and Anthropic's fear-mongering round out the episode.

Positions · 3

  • Energy sector (long energy)LongChamath doubles down that power is the ultimate moat and investors should 'get long energy' given a structural US electron deficit that constrains AI buildout. — Chamath Palihapitiya–
    Why this trade
    “I'm telling you guys, you gotta get long energy.” — Chamath Palihapitiya
    1. US faces a ~1.4 terawatt-hour shortfall by 2050 even if everything planned is built on time.
    2. Hyperscalers may be unable to secure power, breaking the cloud-channel model.
    3. Therefore own the land, power and shell—i.e. go long energy.
  • ANTHROPICLongChamath calls Anthropic's $16.8B quarterly revenue 'breathtaking' and says the company will do great, though it should raise $200-250B at a lower valuation to fund compute. — Chamath Palihapitiya-28.7% since episode
    Why this trade
    “these guys are gonna do great. They're a great company.” — Chamath Palihapitiya
    1. Anthropic is one of two foundation labs with real revenue and a compute duopoly with OpenAI.
    2. He believes the business is exceptionally strong and a successful IPO is likely.
    3. He'd prefer a larger war chest priced to clear for survival through a correction.
  • OPENAILongBen Horowitz says A16Z is 'long both' foundation labs with more stock in OpenAI and is not worried at all about demand for its IPO. — Ben Horowitz-15.5% since episode
    Why this trade
    “We have a lot more stock in OpenAI. I would say, you know, philosophically, we're long both of them” — Ben Horowitz
    1. OpenAI and Anthropic together take ~60% of global compute buildout and dominate foundation-model revenue.
    2. A16Z holds more OpenAI stock than Anthropic; he calls them a duopoly on capability and revenue.
    3. He views owning OpenAI as a defensive hedge against disruption across portfolios.

Ideas · 7

  • CRWVAvoidChamath says CoreWeave is only ~$5B from breaching debt covenants and faces a 'negative reflexive loop' as rising rates push borrowing costs into 10-15% range. — Chamath Palihapitiya-17.8% since episode
    Why this trade
    “My team's estimates is that they're plus or minus about 5,000,000,000 away from that breakeven point.” — Chamath Palihapitiya
    1. CDS-market implies a ~55% five-year default probability and a 15% debt yield on one facility.
    2. Half of the $35B box debt still needs financing at higher rates, eroding fixed-contract margins.
    3. Result: risk/reward is 'a coin flip' and he would not want exposure until financing is resolved.
  • NBISAvoidBen Horowitz says A16Z is deliberately not invested in any of the neoclouds because the debt-financed arbitrage model 'scares me having been through it.' — Ben Horowitz-34.2% since episode
    Why this trade
    “we're not invested in any of these Neil clouds either for that reason because that whole thing scares me having been through it” — Ben Horowitz
    1. LoudCloud history shows long-term capacity contracts can be destroyed by collapsing input prices.
    2. Neoclouds are 'Rube Goldberg financial machines' dependent on debt markets independent of demand.
    3. Therefore A16Z avoids the entire neocloud cohort, including Nebius.
  • IRENAvoidChamath groups IREN with CoreWeave and Nebius as debt-funded box businesses exposed to rising risk-free rates and credit-spread widening. — Chamath Palihapitiya-37.5% since episode
    Why this trade
    “when you're dealing with companies like Corweave or Nebius or IRA, etcetera, that have to borrow a lot to fund these boxes” — Chamath Palihapitiya
    1. Neoclouds borrow heavily to fund GPU 'boxes' against fixed-price contracts.
    2. Rising 10-year yields push borrowing costs up and erode margins across the group.
    3. He warns any model thriving on a financing spread 'gets cornholed' if rates hit 5.5-6%.
  • MSFTLongOpenAI → Chamath and Ben argue OpenAI/Anthropic are a defensive hedge because they threaten almost every incumbent, and MSFT is cited as the creditworthy anchor tenant backstopping the buildout. — Chamath Palihapitiya-1.7% since episode
    Why this trade
    “you need to essentially have Microsoft, Google, Meta, and NVIDIA backstop all of this in a worst case scenario” — Chamath Palihapitiya
    1. Sachs and Chamath say the labs threaten nearly everything in a portfolio, so investors 'kind of have to' own them.
    2. Microsoft is repeatedly named as the investment-grade tenant whose credit underpins the compute boxes.
    3. MSFT offers listed exposure to that AI-infrastructure backstop.
  • NVDALongChamath argues NVIDIA is in 'complete ball control' as all packaging and HBM capacity routes through it, making it a structural winner of the compute shortage. — Chamath Palihapitiya-25.5% since episode
    Why this trade
    “NVIDIA has done an incredible job. Okay? Because if you look one layer beyond NVIDIA, all the packaging capacity is them.” — Chamath Palihapitiya
    1. NVIDIA controls packaging and most HBM allocation, forcing even Google to compete for memory.
    2. Persistent power/compute scarcity keeps GPU prices elevated.
    3. This entrenched position supports a constructive long view on NVDA.
  • GOOGLLongGoogle is cited as an investment-grade compute tenant that backstops neocloud financing, and its new model is 'getting close' to the frontier, supporting a constructive view. — David Sacks-14.3% since episode
    Why this trade
    “The new Google model is getting close.” — David Sacks
    1. Sachs and Chamath name Google among the creditworthy tenants underpinning the boxes.
    2. Chamath notes Google is even competing for HBM, showing deep AI investment.
    3. Sachs says Google's new model is closing the frontier gap.
  • AAPLLongSacks sees Apple's new home-automation device as the secure-network 'brain' that solves the IoT threat, a positive product signal. — David Sacks+3.8% since episode
    Why this trade
    “Apple just announced they're getting into the whole home automation.” — David Sacks
    1. Sacks argues a home 'brain device' will secure networks against IoT worms.
    2. Apple just announced it is entering home automation with such a device.
    3. That positions AAPL at the center of the secure smart-home build-out.

All-In Podcast

Elizabeth Holmes' Husband Billy Evans: The Trial, Prison & Nathan Fielder's Documentary

Oct 15, 2026757h

This episode is a personal interview with Billy Evans, Elizabeth Holmes' husband, discussing her Theranos fraud conviction, prison life, and a forthcoming documentary. The only market-relevant content is Evans' disclosure that he is building a new photonics/diagnostics startup called Hemanthus using Raman spectroscopy, and his mention of his prior work at Luminar building LiDAR. No hosts or guests make any buy, sell, or avoid calls on any stock, ETF, or tradable asset, and no listed beneficiary is explicitly recommended.

No actionable calls in this episode

Automated extraction of statements by podcast hosts and guests - not investment advice, accuracy not guaranteed.