The Julia La Roche Show · · 49 min

#431 Raoul Pal: The Economy We Know Ends by 2030 — and Why We’ll Never See a Big Recession Again

With Raoul Pal

Summary

Raoul Pal joins Julia La Roche to argue that liquidity, not earnings, drives markets, and that the economy as we know it will break down by 2030. He blames the Iran war for raising rates and the dollar and tightening liquidity, which he says caused the choppy trading and the Nasdaq top about five or six months ago, and he expects rates and the dollar to ease and markets to recover, with a possible weaker economy into Q1. He argues that government debt is serviced through debasement of about 8% a year, which accounts for most of the S&P 500's 11% annual gain, and that this debasement acts as a put that prevents systemic recessions. On AI, he sees no bubble: demand is infinite, hyperscalers such as Nvidia, TSMC and Anthropic are highly cash-generative with margins of 30% to 65%, and neocloud failures are just capitalism. He also argues that AI agents will become economic actors, that crypto rails suit them, that jobs will shift toward distinctly human activity, and that AI will eventually govern better than people, which he frames as the fourth turning.

Key takeaways

  • Raoul Pal (bullish on Anthropic): Anthropic's revenue ramp is the fastest in history, reaching a $70B run rate in three years.
  • Raoul Pal (bullish on NVIDIA): Nvidia is among AI stack players earning supernormal margins of 30-65%.
  • Raoul Pal (bullish on Taiwan Semiconductor Manufacturing Company): TSMC is among AI stack players earning supernormal margins of 30-65%.
  • Raoul Pal on AI bubble & tech valuations: AI is not a bubble; AI stocks' P/E ratios are low and this is the clearest super cycle.
  • Raoul Pal on US growth & recession risk: Systemic recessions won't recur; debasement acts as a put option everyone pays for.

What was said, by company and theme

28 views from 1 speaker. Every quote is checked against the episode's transcript (how).

Anthropic

Across the shows: 106 speakers in other episodes, 140 bullish and 89 bearish calls, past 30 days

  • Raoul Pal · Real VisionBullish

    Anthropic's revenue ramp is the fastest in history, reaching a $70B run rate in three years.

    Pal cites Anthropic as the fastest revenue ramp ever, going from zero to a run rate of about $70 billion in three years. He presents this as evidence that AI demand is real and that companies across the AI stack are earning supernormal profits.

    “This is the fastest ramping of revenues in history as anthropic. Nobody's ever scaled revenues this fast in history. They went from zero to what a run rate of $70,000,000,000 in three years? It's staggering.”at 16:22

NVIDIA

Across the shows: 93 speakers in other episodes, 73 bullish and 50 bearish calls, past 30 days

  • Raoul Pal · Real VisionBullish

    Nvidia is among AI stack players earning supernormal margins of 30-65%.

    Pal says every layer of the AI stack is highly profitable, with participants earning between 30% and 65% margins. He names Nvidia as the company that programs the chips, alongside TSMC, the hyperscalers and Anthropic, to illustrate supernormal profits.

    “When you break down the AI companies, everybody involved in this stack, everybody's making supernormal profits. Everyone's making between thirty and sixty five percent profit margins. That's TSMC who make the chips, NVIDIA who kind of program the chips.”at 16:36

Taiwan Semiconductor Manufacturing Company

Across the shows: 7 speakers in other episodes, 6 bullish and 0 bearish calls, past 30 days

  • Raoul Pal · Real VisionBullish

    TSMC is among AI stack players earning supernormal margins of 30-65%.

    Pal argues every layer of the AI stack is highly profitable, with margins between 30% and 65%. He names TSMC as the company that makes the chips, one of the profitable players in the stack.

    “When you break down the AI companies, everybody involved in this stack, everybody's making supernormal profits. Everyone's making between thirty and sixty five percent profit margins. That's TSMC who make the chips, NVIDIA who kind of program the chips.”at 16:36

Bitcoin & crypto markets

  • Raoul Pal · Real Vision

    Crypto is the infrastructure layer for the tech revolution and the best long-run performing market ever.

    Pal calls crypto the infrastructure layer for the technological revolution. He admits it has been a poor market for the last year and a half, but argues that over time it has been the best performing market in history. That, plus the outperformance of technology, is what draws capital away from traditional assets.

    “we've got crypto, which is like the infrastructure layer for all of this. It's not been a great market for the last year and a half, but over time, it's been the best performing market in all of history.”at 12:18
  • Raoul Pal · Real Vision

    Crypto rails are almost perfectly suited to an agent economy, so crypto is a good underlying bet.

    Raoul argues that if AI agents become economic actors, which he says they already are, crypto rails were built for that. Smart contracts can define what an agent is allowed to do and not do, which he sees as a key problem since agents can go rogue. They can also identify the agent and what it is doing, and they are necessary for the payment structure. His logic is that intelligence has moved from carbon to silicon, which runs about a million times faster, while human payment systems and stock markets cannot operate at machine speed, but crypto rails can. He describes crypto as a kind of universal basic equity: infrastructure that unseen agents use, with value accruing to the rail, and anyone with a phone can buy it. He calls it an easy underlying bet, while acknowledging it is volatile and hard to hold.

    “I think of it like a universal basic equity. It's like, if you wanna participate in something, here's this infrastructure rail that's gonna be used by these unseen agents in its unseen economy, and it accrues the value to the rail, and everybody can buy it.”at 39:45
  • Raoul Pal · Real Vision

    Digital art NFTs hold value because blockchain creates digital scarcity; he is buying them.

    Raoul says he has been buying a lot of digital art and NFTs, which he thinks looks bizarre to others but is not. He argues that art has repeatedly moved across substrates, from canvas to walls and streets (Basquiat, Banksy) and now to the internet, where people live their lives. Digital goods normally fall to zero because they can be copied infinitely, but blockchain creates digital scarcity. He links art's value to culture: art captures the culture of each era (Warhol for the 1960s and 70s, Basquiat and Banksy for periods of social unrest), and the art of the internet becomes valuable because humanity is going through the largest period of change in the shortest time. He sees art as the only culture asset that holds value, unlike records, concert tickets and bar visits, which are valuable only as memories.

    “I've been buying a lot of digital art, NFTs. People think this is bizarre, but it's not. ... Anything digital can go to zero in value over time because you could produce infinite amounts of it. But blockchain makes digital scarcity.”at 40:28

US stock market outlook

  • Raoul Pal · Real Vision

    Restricted liquidity is why markets are choppy; a mid-cycle pause until liquidity returns.

    Pal frames liquidity as the main driver of markets. He says liquidity is restricted right now, which explains the choppy trading, violent sector rotations and the Nasdaq topping out about five or six months ago, plus weak crypto. He calls the current phase a mid-cycle pause in which markets have lost their bid. Liquidity is not flooding in yet, but when it returns he expects everything to reignite.

    “And liquidity right now is restricted, and that's why markets have been choppy.”at 2:23
  • Raoul Pal · Real Vision

    Only technology and crypto beat debasement-driven liquidity over time, because technology compounds.

    Pal filters markets through liquidity and concludes that only two asset types outperform it over time. Technology, which he says compounds, and crypto, which he calls its infrastructure layer, do so. That is why he sees the Nasdaq as the place to be.

    “you find there's only two assets that actually outperform liquidity over time, and they're technology because technology compounds over time.”at 7:09

US growth & recession risk

  • Raoul Pal · Real Vision

    Tight financial conditions will weaken the economy into Q1 next year, weaker than expected.

    Pal says the earlier rise in rates and the dollar, which tightened liquidity, will knock on to the real economy with a lag. That means the economy may be weaker than people expect into Q1 of next year, even though markets should recover as rates and the dollar ease.

    “it's gonna knock onto the economy a bit into q one of next year, so the economy is gonna be a bit weaker than people expect potentially.”at 3:11
  • Raoul Pal · Real Vision

    Systemic recessions won't recur; debasement acts as a put option everyone pays for.

    Pal argues governments will always print money when assets fall, as they did in 2020 and 2022, so credit collapses cannot cascade through the system. That debasement works like a put option costing about 8% a year. Sector recessions will still happen, but the economy will not see big systemic wipeouts again. The 2008 recession was the last real one.

    “That's what the debasement is. It's a put option that we all pay to stop the economy falling apart.”at 26:25

AI infrastructure & data centers

  • Raoul Pal · Real Vision

    Hyperscalers can easily fund the AI buildout; their leverage is negligible and their ability to pay is zero risk.

    Pal argues the AI build-out is driven by the most cash-generative businesses in history. Hyperscalers carry so little debt that their ability to pay for data centres is not in question. The market's concern about who funds the data centres and the debt issuance is, in his view, a red herring.

    “Most of this is being driven by the most cash generating businesses in all of history, the hyperscalers. They have more money than God. The amount of leverage they have is so negligible, it's irrelevant.”at 15:00
  • Raoul Pal · Real Vision

    Debt-funded neoclouds may fail, but hyperscalers will absorb their data-centre assets.

    Pal concedes that some neocloud operators borrowing to build data centres will go bust, which he calls capitalism. He expects the assets to be bought by hyperscalers. Systemic risk from that does not follow, because the stronger players hold the cash and will not stop the build-out.

    “And, yeah, fine. A bunch of those will go bust, but a bunch of these assets will just get bought by the hyperscalers who have more money than God”at 17:06

AI: overall outlook

  • Raoul Pal · Real Vision

    AI has infinite demand because more intelligence creates demand for more intelligence.

    Pal argues AI is the only good humanity has created with infinite demand. More intelligence requires still more intelligence, and agents already in use demand more of it. He adds that intelligence per unit of energy is doubling every 4.6 months, which he says explains why the build-out cannot slow.

    “[Intelligence is] the only good we've ever created that has infinite demand. It never stops. Because the more intelligence you build, the more intelligence it requires.”at 15:18
  • Raoul Pal · Real Vision

    AI will become the system of government, making better decisions than humans.

    Raoul argues that the coming change is not a change of political system but a change of the whole system because of AI. He describes it as a complete substrate shift for intelligence, moving from carbon to silicon, which changes how society is governed and what humans do. In his view, AI is going to make better decisions than people, and this is where government is heading. He presents this as the fourth turning, the biggest such shift humanity has gone through. He offers no hard data for the governance claim; the case rests on the substrate argument and his view that the signs are now obvious.

    “The system is going to be AI. It's gonna make better decisions than us. This is where government is going.”at 42:37

AI & jobs

  • Raoul Pal · Real Vision

    Wage growth won't rise again as AI agents do the work, while productivity explodes.

    Pal argues that by 2030 AI and robots will be economic actors doing work humans used to do. That makes productivity explode, but wages stop rising because the work is no longer done by people. He says the economy as we know it will start breaking down by then.

    “it won't have all changed by 2030, but we'll start to see everything breaking apart. Things like wage growth won't ever go up again, but productivity will explode because it's not us doing the work.”at 36:48
  • Raoul Pal · Real Vision

    Human attention and experiences will gain value as AI takes over work, so jobs won't vanish.

    Raoul says people worry about losing jobs to AI, but he does not, because humans can do something AI can never do: be human. Human attention and experiences are immensely valuable, possibly more so than paying for an accountant or lawyer. His evidence includes sports teams and music tours, which grow in value every year. Chess is a key example: no human has beaten a computer in about 35 years, yet the chess industry is the biggest it has ever been because people like watching flawed humans compete with each other. He points to the rise of podcasts and content creators, who capture human attention, and to holiday and leisure spending, which keeps rising because people want experiences. He concludes that the economy will not lose all its jobs; people will simply do different things, and there is plenty of money to be made from what humans find valuable.

    “one thing humans can do that AI can never do and will never be able to do is be human. ... their real currency is attention. Attention and experiences.”at 43:22

Middle East & Iran

  • Raoul Pal · Real Vision

    The Iran war pushed rates, the dollar and oil up, tightening financial conditions and draining liquidity.

    Pal traces the liquidity squeeze to the Iran war. The war raised rates and the dollar and pushed oil higher. He says that combination tightened financial conditions, which lowers liquidity and drives the rotations and weak markets he describes.

    “So rates went up, the dollar went up, and those things with the increase in oil tightened financial conditions, which lowers liquidity.”at 2:29

US rates & the Fed

  • Raoul Pal · Real Vision

    Rates and the dollar will ease, unwinding tight conditions and helping markets find their footing.

    Pal argues the tightening from higher rates and a stronger dollar is already reversing. Rate expectations are coming down as markets wobble, which he says stabilises markets. Markets should get their footing as soon as the dollar and rates ease further, which he thinks is happening now.

    “Now the good side of that is that forces rates to come lower and the dollar to weaken a bit anyway.”at 3:05

Banks

  • Raoul Pal · Real Vision

    Banks must lend to fund the biggest capex boom, so bank-system liquidity is reigniting.

    Pal says there are only two sources of new money: central bank printing and the banking system. Central banks have stopped printing, so new liquidity has to come from banks lending. The reason is the biggest capex boom in history, which needs heavy borrowing. Financial conditions such as the dollar and rates determine how much banks lend.

    “which we're trying to reignite right now to get banks to lend because we've now got the biggest CapEx boom in history.”at 6:07

US deficit & government debt

  • Raoul Pal · Real Vision

    Governments service debt by debasing money about 8% a year, which lifts asset prices.

    Pal says ageing populations cut growth, so governments piled on debt. Debt as a share of GDP is almost perfectly inverse to labour force participation, which he calls the greatest chart in macro. Rates were cut to zero and debts refinanced, and interest is now paid through liquidity and money printing. Pal says global debasement runs at about 8% a year. That accounts for roughly 8 of the 11% annual gain in the S&P 500, with about 3% from earnings growth.

    “So the S and P goes up by about 11% a year, and the rate of debasement globally is about 8% a year. So you've got 8% is just driven by this lowering of the denominator and about 3% being driven by improvements in the company's earnings”at 6:44

US housing

  • Raoul Pal · Real Vision

    Home prices outrun incomes because currency debasement inflates scarce assets faster than earnings.

    Pal says debasement changes the denominator on scarce assets such as houses, while earnings are cyclical and do not grow the same way. Home prices therefore keep climbing faster than what people earn, however hard they work. That is why people hold two or three jobs and feel unable to get ahead.

    “It's like house prices just become impossibly expensive. However hard you work, your earnings don't go up as much as the house prices do.”at 9:25

Gold & precious metals

  • Raoul Pal · Real Vision

    Gold does not outpace debasement, so it doesn't build real wealth over time.

    Pal says debasement sets the baseline return on everything. Real estate, gold and the S&P 500 do not reliably beat it. Traditional portfolios such as the 60/40 therefore leave investors at the same relative wealth, which he sees as the reason people turn to technology and crypto.

    “Because real estate doesn't go up faster than debasement. Gold doesn't go up faster than debasement. The S and P 500 doesn't really go up faster than debasement.”at 12:36

Market sentiment, positioning & volatility

  • Raoul Pal · Real Vision

    The November 5 sell-off was narrative-driven noise from thin liquidity and leverage, not meaningful.

    Pal says the sell-off happened because there was little liquidity and the markets were leveraged. The catalyst was the midterm fear that Democrats would be anti-AI and anti-data centre. He does not think that is meaningful. It is a narrative driven by liquidity, and falling rate expectations helped stabilise markets afterwards.

    “Is it anything meaningful? I don't think so. ... So it's a narrative. It's liquidity driving it because it always is.”at 14:02

US–China AI & chip race

  • Raoul Pal · Real Vision

    The AI race with China can never stop; neither side can afford to lose.

    Pal argues the AI race happens at national level as well as corporate level. The US and China are in a race that cannot stop, because slowing down means losing. He says that slowing AI through regulation would hand China superintelligence, which he sees as a one-party world.

    “The US is in a race with China, and it can never stop. Nobody can afford to win. Nobody can afford to lose, and nobody will win.”at 18:12

AI regulation & safety

  • Raoul Pal · Real Vision

    Nobody can stop AI; the doom narrative looks like Anthropic-backed regulatory capture.

    Pal says AI is too big and too powerful to stop, and government cannot keep pace with an intelligence that doubles every 4.6 months. He reads the viral AI-doom narrative as likely paid for by Anthropic and others to win regulatory capture for monopolies. He says the open-technology view prevailed, since slowing AI would let China win.

    “But I think it's too big and too powerful to stop. ... I don't think there's any way anybody's gonna stop it”at 20:05

AI bubble & tech valuations

  • Raoul Pal · Real Vision

    AI is not a bubble; AI stocks' P/E ratios are low and this is the clearest super cycle.

    Pal says a bubble is not possible because demand is infinite, the build-out is funded by highly cash-generative companies with little leverage, and every component of the stack makes money. Price-earnings ratios across these companies are low because the market doubts the super cycle. He calls it the most obvious super cycle he has seen, and only a massive liquidity withdrawal could stop it.

    “So there's no way that this is a bubble. We haven't got going yet. And if you look at the the price earnings ratios or anything of these companies, they're all low right now because the market's so skeptical of this being a super cycle.”at 22:57

Robotics & autonomous vehicles

  • Raoul Pal · Real Vision

    AI-brained humanoid robots will be stronger, cheaper and more numerous than humans; invest in robots.

    Pal expects AI brains to be placed into robots that are stronger, cheaper and faster than people, and that infinite numbers of them will be made over time. He says that will reshape the economy, so as a hedge against the uncertainty he suggests investing in the robots.

    “We're gonna put these AI brains into robots, which people aren't ready for. ... we're gonna make infinite numbers of them over time.”at 32:40

Enterprise AI adoption & AI software

  • Raoul Pal · Real Vision

    AI agents can form businesses, hire staff and raise capital instantly, with no registration.

    Pal argues that agents remove the friction of setting up a business. They have no contracts, registration or slow paperwork, so they can form businesses, hire employees and raise capital immediately. He points to meme coins raising a billion dollars in an hour as a precedent. Eventually these agents will change how markets operate.

    “Your AI can build a business instantaneously. It can have employees instantaneously. It has no contracts. It has no registration.”at 35:01

Across the shows

What other episodes Chorus covered said about the same companies in the 30 days up to this one.

Topics

  • AI agents as economic actors
  • AI arms race
  • AI as governing system
  • AI bubble risk
  • AI compute demand
  • Iran war financial tightening
  • November 5 sell-off
  • bank lending for capex
  • chip-stack profit margins
  • crypto as agent-economy infrastructure
  • crypto as long-run performer
  • debt servicing via debasement
  • digital art and NFTs
  • gold vs debasement
  • halting AI development
  • home prices vs incomes
  • humanoid robots
  • hyperscaler capex funding
  • job displacement
  • liquidity and market direction
  • near-term growth slowdown
  • neocloud financing risk
  • rate expectations
  • revenue growth
  • systemic recession risk
  • technology outperformance
  • wage growth under AI

More from The Julia La Roche Show

All The Julia La Roche Show episodes

Ask your AI about this episode

Chorus gives Claude, ChatGPT and other assistants every view from this episode and every other show Chorus follows. Ask who agrees, who changed their mind, and what else was said about the same companies.

Setup guide

A weekly email with the calls that mattered from the most important voices in investing. Unsubscribe any time. Privacy Policy

Chorus reports what speakers said. It isn't investment advice. How Chorus works · About Chorus