The Everything Code and the Four-Year Window: Raoul Pal and Cathie Wood, Revisited

Download this essay as a PDF

The full essay as a printable document: the Everything Code step by step, what Raoul Pal said in October about the four years ahead, where Cathie Wood's deflation argument fits, and where I now part company with my June piece.

In June I sat down with two long interviews from Steven Bartlett's Diary of a CEO — Cathie Wood of ARK Invest, and Raoul Pal of Real Vision — and wrote them up as one argument. Wood, I said, was describing what the AI economy would build. Pal was describing what it would pay with. I put the two side by side, drew up a blended allocation, and ended on the line that "the two greatest investment opportunities of our lifetimes may not be alternatives."

Four months later I want to come back to it, for three reasons.

The first is that Pal has given a long new interview — his sixth conversation with Kevin Follonier on When Shift Happens, released on 15 October — in which he says, in effect, that the window he described in June has narrowed. "We have about four years," he wrote a fortnight later, summarising it. "Four years to get yourself sorted before the economy starts to work in ways none of us are used to."

The second is that Pal has written a book. The Everything Code is published by Harriman House, Pan Macmillan's business imprint, and I bought it as soon as I could. I have read more than a third of it already, and I find it highly enlightening — not because it tells me what to buy, which it is mostly careful not to do, but because it gives a single, coherent account of something I have been circling for months in my own writing without quite naming.

The third is my own work. Since June I have published two scorecards on the AI-boom stocks: The Shortage Held. The Trade Didn't in October and, this morning, Beat, Raise, Sell. The conclusion of both, stated as plainly as I can, is that the companies did everything I expected and the share prices did not care. What moved them was the cost of money and the supply of it. That is, almost word for word, Pal's central claim. I did not appreciate how central until I read his book.

So this is not a summary of two videos and a book. It is my attempt to set out how I now understand the framework underneath all of them, where I think it is right, where I think it is weaker than its presentation, and what I would say differently from what I said in June.

1. What I wrote in June, re-read honestly

I want to start by being fair to my June piece and fair to its readers.

It did three things well. It identified correctly that Wood and Pal, coming from equities and macro respectively, had converged on the same broad picture: AI as a deflationary force, the existing financial system as inadequate for what is coming, a short window to position, and bitcoin as the asset on which both frameworks agreed. It explained Pal's argument about machine-to-machine payments — that autonomous AI agents will need money that settles instantly, works in fractions of a cent and needs no bank account — which I still think is one of the more interesting ideas in the whole space, and which I have since worked through at the protocol level in A Hash, a Clock and a Signature. And it was explicit that it was a synthesis of two other people's public views.

It also did three things I would not do again.

It turned convictions into allocations. I tabulated a "blended portfolio" with percentage ranges — 40–50% in core crypto, 25–35% in disruptive equities and so on. Those numbers were my synthesis of two people's public enthusiasms, not the output of any analysis of a reader's circumstances, and putting them in a table gave them an authority they had not earned.

It repeated claims I had not checked. I wrote, for example, that "65% of agentic AI payments run on Solana". I cannot now find a primary source I would stand behind for that figure, and I should not have used it.

It was written in the voice of a pitch. The summary invited readers to talk to me about investing in these opportunities. I am a doctor who has held bitcoin for a long time and thinks hard about macro, and I am not a licensed financial adviser. I said so in the disclaimer; I should have written the whole piece in a way that made it unnecessary to say.

None of that makes the underlying framework wrong. It makes my presentation of it worse than it should have been. This piece is the version I would rather have written.

2. The Everything Code, as I understand it

Pal has been developing what he calls the Everything Code for years, in his Global Macro Investor research and in talks — a teach-in at Solana's Breakpoint conference last December laid it out in compressed form — but the book is the first time he has set it out end to end. The publisher describes it as a framework for understanding economic cycles driven by debt refinancing, set inside a broader argument that the economic system is broken, that central banks make it worse by debasing currencies and prioritising government debt over individual prosperity, and that technology and crypto are "life rafts" within it.

Here is the chain of reasoning as I understand it, step by step, in my own words. Where I am adding something of my own, I say so.

Step 1 — Growth has three sources, and two of them are failing

The foundation is close to an identity. An economy's output grows because there are more people working, because each of them produces more (productivity), or because more is being borrowed against future output (debt). Over the long run, the first two are what make a society richer. The third brings consumption forward.

For most of the twentieth century, rich economies grew on all three. Then two things happened. Populations aged: in much of the developed world, and in China, the working-age population is now flat or shrinking. And measured productivity growth slowed, from the high rates of the post-war decades to something much more modest. Growth did not stop. It came increasingly from the third source.

The Everything Code's first claim is that this is not a cyclical blip but a structural condition. If the labour force is not growing and productivity is slow, an economy that wants to keep growing — and governments that want to keep funding pensions, health systems and defence — have to keep borrowing more. Pal points to global debt of roughly 140–160% of GDP and rising.

As a physician, I recognise the shape of this argument. It is the logic of a compensated failure: a system that keeps its output up by leaning harder and harder on a reserve, until the reserve becomes the thing everyone has to manage.

Step 2 — The debt has to be refinanced, on a schedule

Debt does not just accumulate; it matures. When rates fell to zero after 2008, governments refinanced enormous amounts of debt at near-zero rates, much of it at maturities of three to five years. In an interview with Tom Bilyeu in 2024, Pal put the average at around three and a half years: debt comes due, is rolled into new debt, and the interest bill arrives on schedule.

This is the second claim, and the most original. If a large stock of debt has to be refinanced every few years, the refinancing itself creates a cycle. When a wave of debt comes due, the system needs liquidity to roll it. Central banks and treasuries supply that liquidity — by buying bonds, by managing the issuance calendar, by easing bank capital rules — and that liquidity flows into assets. Then the wave passes and liquidity tightens until the next one.

At Breakpoint, Pal argued that when the US Treasury extended the average maturity of its debt to around 5.4 years in 2022, it stretched the cycle — and, by extension, the crypto cycle, which he has long tied to the same rhythm — from roughly four years to something longer, pushing the projected peak into late 2026.

Step 3 — The only way through is to expand the money supply

This is where "debasement" comes in, and it is the step most people bristle at, so I want to state it carefully.

If debt is near 100% of GDP for governments and well above that for the private sector, and if nominal growth is low, then the interest on the debt absorbs a large share of the growth. Pal's illustration to Bilyeu was simple: at about 2% growth and 2% interest, all of the growth goes on interest. There are only a few ways out of that arithmetic: default, austerity severe enough to shrink the debt, sustained high growth, or growth in the money supply that allows the debt to be serviced in money worth less than when it was borrowed.

Pal's argument is that the last option is the one democracies always choose, because it is the only one that does not require a government to tell voters something unbearable. So debasement is not a failure of policy. It is the policy, pursued deliberately because the alternatives are worse.

I would add one qualification of my own. "Debasement" is sometimes used as though it meant hyperinflation. It does not, in Pal's usage. It means that the money supply grows faster than the real economy, persistently, so that the purchasing power of money falls against assets even when the measured inflation rate looks moderate.

Step 4 — When money grows faster than the economy, assets outrun wages

This is the consequence that makes the framework feel personal, and it is the part of the book I have found most illuminating so far.

If the money supply grows faster than output, the extra money has to go somewhere. Some of it shows up in consumer prices. Much of it shows up in the prices of assets: houses, equities, land, collectibles, and anything else scarce. Wages, which are tied to the real economy, do not keep up. So people who own assets get richer, and people who earn salaries fall behind, even when both are doing everything right.

Pal calls the result a kind of money illusion. Measured against the size of central bank balance sheets, he told Bilyeu, the S&P 500, real estate and gold have "gone nowhere": they have simply risen with the tide of money. He claimed that central bank balance sheets were about 97% correlated with asset prices. And his summary of what that means for ordinary people is bleak: "nobody's getting any richer".

One of the book's endorsements, from Micky Malka of Ribbit Capital, captures what the book is trying to do with this: explain how the system works, "why it isn't your fault", and what to do next. That is, I think, the emotional core of the book, and the reason it will reach readers well beyond the crypto audience Pal usually speaks to. A great many young professionals — including, I suspect, many of the junior doctors I teach — feel that they did everything they were told and are still further from owning a home than their parents were at the same age. The Everything Code offers a mechanism for that feeling.

Step 5 — If liquidity drives everything, the risk curve tells you where to stand

The final step turns diagnosis into strategy. If asset prices are driven mainly by liquidity, then the question for an investor is not "which company is best?" but "which assets rise most when liquidity rises?". Pal's answer is that the further an asset sits out on the risk curve — the more speculative, the more dependent on future growth — the more it responds to liquidity. Technology equities respond more than the broad market. Crypto responds more than technology. Within crypto, smaller networks respond more than bitcoin.

That is the logic behind his long-standing view that crypto and technology are the only assets that have clearly beaten debasement, and it is why his own portfolio is concentrated where it is. "It's all noise," he said at Breakpoint, "unless you're talking about liquidity."

The book as an object

Two things about the book itself are worth knowing.

It is short — the paperback runs to a couple of hundred pages — and it reads, as Malka put it, "like a long dinner with Raoul". It is aimed at the intelligent general reader, not the professional macro trader, and it is better for it.

And it is, by the publisher's own description, "human-originated and AI-facilitated". Harriman House says the ideas and voice are Pal's, but that the manuscript was produced with a proprietary tool, Weave, grounded only in Pal's own published work, speeches and frameworks, and then fact-checked, reviewed and edited like any other title in its list. Pal's explanation was characteristically direct: he "would never have the time to write this book" while juggling four roles. Given everything I have written this year about AI in clinical practice — about what it means to vouch for a document you did not draft line by line — I find this both interesting and honest. The publisher discloses it on the cover and in an author's note. It is, in a small way, a working example of the world the book describes.

3. What Pal said in October

The new interview is long — about an hour and fifty minutes — and wide-ranging, with sponsors and digressions of the kind every successful crypto podcast now carries. Here is what I take from it, drawn from published summaries of the episode and from Pal's own posts about it.

Watch the interview: Raoul Pal: How to Get Rich in Crypto Before AI Changes Everything — When Shift Happens, E190

He has done almost nothing, deliberately

The most useful thing in the interview, for most readers, is the least dramatic. Pal says he has made only two trades all year — one in January and February, one in July when prices fell — and that both followed the same rule: add only when an asset is about two standard deviations below its long-term log trend. "I've done nothing except buy when I got two standard deviations oversold versus a log trend and I added more and I bought some Zcash. That's all I've done."

He pairs that with an argument about cash that I think is underrated: that keeping a reserve for deep sell-offs does more for long-run compounding than trying to call tops, because buying at the bottom of a drawdown means your portfolio reaches new highs before the market does. And he puts a number on his own record. Reviewing 21 years of recorded positions at Global Macro Investor, he found a win rate of about 65%, with winning trades averaging around 50%. He attributes it to pattern recognition rather than intelligence.

That is a far more disciplined picture than the one most people have of him, and a far more useful one to copy than any particular coin.

He thinks Ethereum is badly mispriced

Pal called Ethereum "wildly undervalued" and offered a thought experiment I found genuinely clarifying: ask what would disappear if the network were switched off. His answer was about 70–76% of all stablecoins in existence, a similar share of decentralised finance, most of the on-chain real-world assets, and the digital art market — a loss he estimated at $1–1.5 trillion of value that, in his view, the market does not price because it focuses on Ethereum's cash flows rather than its role as infrastructure. He also said that almost 90% of his own ETH is now held as digital art, and that he is starting a digital art fund, Epoch1.

He is still underwater on SUI, and says so

In June I reported Pal's very large conviction in SUI — he had described it as dramatically undervalued relative to Solana and had at one point said he had moved the bulk of his portfolio into it. In October he said his average entry is near $1.70, that the position is currently at a loss, and that he remains committed, citing what he calls "economic density" — value locked per active user — as evidence that SUI, Solana and Ethereum held up best in the recent pullback. The episode's chapter list includes, pointedly, "Was the 70% SUI position sizing wrong?" and "Facing the millions of people who followed him on the SUI trade". He also said he now refuses to disclose position sizes.

I respect the candour. I also think it is the most important lesson in the interview, and I come back to it in section 6.

He bought Zcash, on a privacy thesis

Pal said he bought Zcash at around $200–300, including during a flash crash caused by a supply scare that turned out to be false, and that his thesis is rising demand for financial privacy — which he links to governments seizing the assets of sanctioned parties.

He thinks liquidity went somewhere else

On the cycle, Pal's explanation for why crypto and technology have lagged the growth in money supply this year is that liquidity flowed into AI infrastructure instead. He expects rate cuts and renewed capital spending to redirect it towards crypto, and argues that crypto is cheap relative to gold and the Nasdaq. One chapter title — "The BANANA ZONE that never happened" — is an implicit admission that the explosive phase he predicted for this cycle has not yet arrived.

And the window has shrunk

The headline, which Pal amplified in a post on 31 October, is the timeline. He now places what he calls the "economic singularity" — the point at which AI agents and robotics break the models economists use, reshaping wages, productivity and capital markets — at around 2030, earlier than the 2030–2032 range he used before. He calls the integration of AI and crypto the largest mispricing in markets, because AI agents could turn a finite base of human users into an effectively unlimited one. And he frames the next four years as the period in which to prepare:

"We have about four years. Four years to get yourself sorted before the economy starts to work in ways none of us are used to … you'll still be able to invest after that, but the rules will be different."

His prescription, in the same post, was "technology and scarce assets".

4. Cathie Wood, revisited

Watch the interview: Cathie Wood on Diary of a CEO

Wood's framework, as she set it out on Diary of a CEO, starts from the opposite end. Where Pal begins with money and debt, Wood begins with technology and cost curves.

Her central idea is convergence: five technology platforms — artificial intelligence, robotics, energy storage, multiomic sequencing and blockchain — maturing at the same time and accelerating one another. Her central mechanism is Wright's Law, the observation that each doubling of cumulative production of a technology reduces its cost by a roughly constant percentage. Apply that to five platforms at once and she arrives at productivity growth of the order of 4–6% a year, a rate not seen since the late nineteenth century.

Her central consequence is deflation. If AI and robotics keep driving down the cost of producing goods and services, then prices fall, interest rates follow, and long-duration growth assets — the companies whose value lies furthest in the future — are re-rated upward. Bitcoin fits into this, in her account, as the scarce asset in a world of abundance: a fixed supply in a world where everything else gets cheaper.

ARK's own numbers have moved since June, and in an interesting direction. Its 2026 research, as reported in May, projected that bitcoin's market capitalisation could reach about $16 trillion by 2030, driven largely by its taking around 40% of gold's market value, plus nation-state and corporate treasuries and on-chain financial services. And ARK cut its forecast for bitcoin's role as a safe haven in emerging markets by 80%, because stablecoins have grown so fast in developing countries.

That last revision is the one I would underline for any reader in Kenya. It is exactly what I described in my Malipo Rail write-up: ordinary Kenyans who want to protect themselves against the shilling overwhelmingly reach for dollar stablecoins, not bitcoin — and the Central Bank of Kenya has now taken powers to control the availability of foreign dollar-linked stablecoins. ARK's revision is, in effect, an admission that the emerging-market version of the bitcoin story is being taken by a different asset.

5. Where they agree, and where they quietly contradict each other

In June I emphasised the agreement. I still think it is real, but it is worth being more precise about it.

The contradiction is in the third and fourth rows. Wood says the future is deflationary: AI makes everything cheaper, so rates fall. Pal says the future is debased: governments must keep expanding the money supply to service their debts, so money keeps getting cheaper against assets.

These are not actually incompatible. You can have both at once: falling prices for things, and rising prices for assets. In a sense, that is a fair description of much of the last fifteen years. But they imply different things for the cost of money. Wood's world is one in which rates fall because inflation does. Pal's is one in which rates are held down because debt requires it. In the October of my scorecards, neither was happening: the US ten-year yield stood above 5%, inflation figures were accelerating, and the market spent the month pricing the possibility of a rate increase. The memory and AI-infrastructure stocks I follow lost a fifth of their value in three weeks while their earnings rose.

That is not a refutation of either framework. Pal would say it is exactly what his framework predicts when liquidity tightens. Wood would say that the deflationary forces have not yet overwhelmed the cyclical ones. But it is a reminder that both frameworks describe long-run tendencies, and that a reader can be completely right about the decade and still lose a third of their money in a quarter.

6. Where I now part company — and where I don't

I want to be careful here, because it would be easy to read what follows as dismissive, and it is not. I have read more of Pal's work this year than I have of almost any other macro writer, and the book has changed how I think about my own scorecards. But a framework earns trust by surviving its weakest points, so here are the ones I see.

The diagnosis is stronger than the deadline

The demographic trend is measurable. The debt ratios are measurable. The maturity profile of government debt is published. The correlation between liquidity and asset prices can be computed, and argued about. These are the parts of the Everything Code I find most convincing, because they can be wrong and mostly are not.

The "four years" is a different kind of claim. Pal himself gave no model, date or measurable milestone for it in the post that announced it. "Economic singularity" is his own term, not a term of art. And a deadline that is always about four years away has a long history in financial writing: it is the right length to feel urgent and the wrong length to be checked. I take the direction of the claim seriously — AI agents and robotics will change the relationship between labour and income, and probably faster than most economists assume — but I would not organise my finances around the date.

Correlation is not mechanism

The 97% correlation between central bank balance sheets and asset prices that Pal cited in 2024 is striking, but two series that both trend upward for fifteen years will correlate strongly whether or not one drives the other. The stronger evidence for the liquidity thesis is not the level correlation but the timing: whether turns in liquidity lead turns in asset prices, consistently, out of sample. Pal's work at Global Macro Investor has always been built on exactly that kind of lead-lag analysis, and his track record suggests it works more often than not. But a reader should know that the headline number is the weakest form of the evidence, not the strongest.

The SUI trade is the lesson, not the footnote

I do not say this to score a point. Pal has been unusually open about it, and his record over twenty-one years is far better than mine. But the most valuable thing in the October interview, for me, was watching one of the most careful macro thinkers alive explain why a position he once sized enormously, and talked about to millions of people, is underwater, and whether the sizing was wrong.

Every framework in this piece — Pal's, Wood's, mine — can be right about the direction and still ruin someone who sizes it badly. Pal's own discipline in 2026 (two trades, buy only at two standard deviations below trend, keep cash for the drawdown) is a better guide to action than any of his individual conviction calls, including the ones I reported in June.

The incentives are visible, and should be

Pal runs a media and research business; the interview I am discussing carried eight sponsors; the book is, among other things, a product; and his post announcing the four-year window linked to it. None of that makes him wrong, and he is more transparent about his positions and mistakes than most. But a reader should hold in mind that the people who explain the system to us also earn a living from our attention, and that urgency sells. I say the same of myself: this blog has carried my own convictions, and in June it carried an invitation to talk to me about acting on them.

Kenya is not the United States

Almost everything in the Everything Code is described from the vantage point of the dollar system — US debt, the Federal Reserve, the US Treasury's maturity profile. For a reader in Nairobi, three things change. Your salary and your costs are in shillings, so the first debasement you feel is the shilling's against the dollar, not the dollar's against assets. The asset most Kenyans actually use to escape it is a dollar stablecoin, which is precisely what ARK has just conceded and what the Central Bank of Kenya now has power to restrict. And holding or trading crypto through Kenyan businesses now sits under a licensing regime whose transitional period ends tomorrow, 4 November, with no licences yet listed. None of that makes Pal's framework less true. It makes acting on it more complicated here than his audience in London or Miami will experience.

7. What I am taking from all of this

So what, concretely, has changed in how I think?

I now treat liquidity as the first question, not the last. My two AI-boom scorecards were organised around earnings, margins and guidance — the language of company analysis. They were right about the companies and wrong about the stocks, and the gap between the two was almost entirely the cost and supply of money. If I were setting signposts again, the first rows would be the US ten-year yield, central bank balance sheets and the debt-refinancing calendar, and only then the earnings dates. That is Pal's framework applied to my own mistake, and it is the single most useful thing I have taken from the book so far.

I separate the debasement hedge from the speculation. The Everything Code makes a strong case for owning some scarce assets as protection against the steady loss of money's purchasing power. That is a different decision, with a different size and a different time horizon, from betting on which network or company will win. In June I blended them into one table. They belong in separate conversations.

I take discipline more seriously than conviction. Two trades a year, at prices two standard deviations below trend, with cash held back for the drawdown, is a strategy almost anyone can follow and almost nobody does. It is worth more than any individual call in either interview.

I hold the window loosely. I think the transition Pal and Wood describe is real, and probably faster than consensus. I do not think anyone can date it, and I am wary of any argument whose practical conclusion is that the reader must act now.

And I will keep writing about it as research, not as an invitation. The disclaimer at the end of this piece is not boilerplate. It is the point.

8. Where I land

Four months ago I wrote that the two greatest investment opportunities of our lifetimes might be the same opportunity seen from different angles. I still think Wood and Pal are describing the same transformation — technology driving the cost of almost everything towards zero, while governments keep driving down the value of money to carry the debts of an ageing world. Read together, they are a remarkably clear account of why the decade feels the way it does.

What I would change is the conclusion I drew. The lesson of the Everything Code is not "buy these assets before the window closes". It is that in a system built on refinancing debt, the most important variable for anyone's savings is something they cannot see on a company's income statement: the supply and price of money. Understand that, size anything you do accordingly, and keep enough in reserve to buy when everyone else is forced to sell. That is less exciting than the version I wrote in June. I think it is closer to what the book is actually trying to teach.

I will write again when I have finished it.

This article is a research and educational summary of publicly stated views and of a published book. It does not constitute investment advice, and I am not a licensed financial advisor. It is not a recommendation to buy or sell any security or digital asset. Readers should consult a licensed financial adviser and do their own due diligence before making investment decisions, and readers in Kenya should be aware that the regulatory position of virtual-asset service providers is changing. Past performance, including the performance of any investor discussed here, is not indicative of future results.

References and further reading

The book - Raoul Pal, The Everything Code: Seizing the Opportunities at the End of Economic History and the Dawn of the Exponential Age. Harriman House / Pan Macmillan, 2026. - Pan Macmillan, Macro analyst Raoul Pal and Harriman House announce human-originated, AI-facilitated book The Everything Code, 20 August 2026. - Goodreads, The Everything Code.

The interviews - When Shift Happens, Raoul Pal: How to Get Rich in Crypto Before AI Changes Everything — E190 (YouTube), 15 October 2026; episode page and chapters. - Diary of a CEO, Raoul Pal: "We Only Have 6 Years Until Everything Changes". - Diary of a CEO, Cathie Wood.

Reporting on the October interview and Pal's posts - Scuttlebutt, Real Vision CEO on ETH: "it would remove …" — summary of the When Shift Happens interview, 16 October 2026. - KuCoin News / Benzinga, Raoul Pal calls Ethereum "wildly undervalued", 16 October 2026. - AIera, Raoul Pal's AI warning: four years before the rules change; Gokhshtein, Raoul Pal warns economy to change in four years, favours scarce assets, 31 October 2026.

The Everything Code before the book - Tom Bilyeu, Why you can't afford anything — Raoul Pal (transcript), September 2024. - Solana Compass, Raoul Pal's macro teach-in at Breakpoint 2025, December 2025.

ARK Invest - MetaTrader News, Bitcoin market cap could reach $16 trillion by 2030, ARK Invest explains how in new report, May 2026.

Earlier pieces on this site - The Machine Economy: What Cathie Wood and Raoul Pal Are Telling You to Buy Right Now, 25 June 2026 - The Shortage Held. The Trade Didn't: Investing in the AI Boom, Three Months On, 12 October 2026 - Beat, Raise, Sell: Investing in the AI Boom After the Third-Quarter Results, 3 November 2026 - A Hash, a Clock and a Signature: The Lightning Network, by Hand, and the Malipo Rail, 2 November 2026