Bitcoin outlook, 2026-09-13

What Dalio actually wrote

On August 21 Ray Dalio, who founded Bridgewater Associates and has spent close to fifty years studying government debt, published an essay called "How Countries Go Broke: The Dynamic Behind What Is Happening Now." It carries the title of his June 2025 book and applies that book's template to the present moment. One sentence travelled further than the rest: he expects "non-government-produced monies like gold and Bitcoin to do relatively well." His general advice was to diversify across asset classes and countries with strong income statements and balance sheets, "underweighting debt assets like bonds, and overweighting gold and a bit of Bitcoin."

Two details tend to fall off in the retelling. The percentage belongs to gold alone: "Having a small percentage, maybe 10-15%, of one's money in gold can reduce a portfolio's risk, and I think it would also raise its return." Bitcoin gets "a bit," with no figure attached. And Dalio has not changed his mind about Bitcoin. In late July he said he holds "about 1% of my portfolio in Bitcoin because there's different kinds of money, and the money that you can't print, that's one kind," while adding that he prefers the gold bars. In May he wrote that Bitcoin "lacks privacy. Transactions can be monitored and potentially controlled, which is why central banks aren't looking to hold it."

That combination is the reason to read him. The allocation survives its own author's scepticism.

The week it landed in

The essay arrived in a specific week. On August 18 total US federal debt outstanding passed $40 trillion for the first time, closing the day at $40.05 trillion. On August 17 the 30-year Treasury yield settled at 5.31 percent, which Bloomberg reported as the highest since 2007. On August 19 the Treasury announced it would more than double its long-end liquidity support buybacks, from $2 billion to at least $4 billion an operation in the 10-to-20-year and 20-to-30-year sectors, running from September 9 to November 4. And Bitcoin closed its best week since March 2023, up about 23.6 percent, adding roughly $14,800, the largest one-week dollar gain in its history.

The order matters. That rally started when long yields fell after the buyback announcement, two days before Dalio published. He did not move the price, and nothing in his essay claims to.

One number in the essay deserves a note. Dalio writes that the debt is "about $32 trillion," while the headline figure that week was $40 trillion. Both are right and they measure different things. On September 10 the Treasury reported $40.05 trillion outstanding in total, of which $32.36 trillion is debt held by the public and $7.68 trillion is held inside the government itself. Dalio is using the public figure, which is the one that has to find a buyer.

The arithmetic, in his own numbers

The essay puts the problem as a balance sheet. Revenue this year will be about $5.5 trillion, expenses about $7.5 trillion, leaving a shortfall of about $2 trillion. Against $32 trillion of debt, roughly six times revenue, the interest bill is about $1 trillion, close to 20 percent of what comes in. Each year the gap is borrowed, and the interest line grows faster than receipts and starts eating the rest of the budget.

At some point some creditors decline to roll over at the offered yield. From there, without a fiscal adjustment, Dalio sees two exits. In the first, yields rise until someone finally buys. That works, because there is always a clearing price, but it squeezes credit, breaks markets and produces defaults. In the second, the central bank creates money and buys what private demand will not, which makes the debt problem look solved while the currency loses purchasing power.

His claim, drawn from the 35 cases over the past 100 years he studied for the June 2025 book, is that for governments whose debt is denominated in a currency their own central bank can create, the first route hurts so much that the second one is chosen. That is a statement about incentives rather than a prophecy. Nobody wins an election by engineering a depression to protect the currency.

Why the Bitcoin line is reached by subtraction

If broad devaluation is the expected path, ordinary currency diversification does less work than it looks like it does. You are no longer escaping the cause; you are spreading across several expressions of the same cause. So the filter narrows. Not the best asset of the moment, but something much more specific: monetary, liquid, transferable, scarce, and not somebody else's liability. Gold and Bitcoin are what is left at the end of that subtraction, and that is where Dalio names them.

Nothing in the reasoning requires liking Bitcoin, and that is the point worth keeping. An argument by elimination has a property an argument from conviction does not: it does not rest on the enthusiasm of whoever makes it. To contest it you have to attack the filter or the balance-sheet arithmetic, not the author's taste. It also explains why the allocation is small. Inside the part of a portfolio meant to protect against devaluation, gold is already sitting in the seat, and Dalio's own list of Bitcoin's defects keeps the rest of the share down.

The objections, and what the current data says

It trades like a technology stock. Dalio's argument is that when investors are squeezed elsewhere in a portfolio they sell Bitcoin to cover, which weakens the haven case. The recent record cuts against it. In a research note of August 27, Grayscale put Bitcoin's 90-day correlation with the Nasdaq 100 at about 33 percent, down from above 60 percent at the start of the year, while its correlation with gold rose above 50 percent from close to zero. BlackRock, writing on August 17, said that "These periods of elevated risk correlation have been episodic rather than structural." The objection describes a regime, not a fixed property, and the regime has been moving.

The market is small and therefore manageable. True, and it stays true even when the comparison is done carefully. The World Gold Council counts about 216,000 tonnes of gold above ground; at $4,385 an ounce on September 11 that is roughly $30.5 trillion, against about $1.55 trillion of Bitcoin, close to twenty to one. But most of that gold is not investable: the same breakdown puts about 45 percent in jewellery and about 15 percent in industrial, medical and other uses. Counting only bars, coins, funds and official reserves, roughly $11.9 trillion, the ratio is nearer eight to one. Still a wide gap for anyone who has to place billions, and still a constraint on adoption rather than a defect in the asset.

Quantum computing. A long-horizon risk with known mitigations and no evidence of a near-term capability. We have covered it separately and do not treat it as the load-bearing objection.

There is no privacy. This is the strongest of the four, and it is the one that is different in kind. Market size and correlation are conditions of adoption that can change. A public ledger is not a phase. A central bank holding Bitcoin would, once its addresses were identified, be publishing the size of its position and every subsequent move to its counterparties. For an institution whose effectiveness depends on nobody knowing what it will do next or when, that is a real problem, and Dalio's version of it holds up.

Testing that objection against the best-documented case

The clearest recent case of sovereign reserves being immobilised is Russia. Its international reserves peaked at $643.2 billion on February 18, 2022, and within days about $300 billion became inaccessible. The Congressional Research Service records that the main funds not frozen were the reserves held in China and the gold stored in the central bank's own vaults. In the European Union about €210 billion is immobilised, of which some €185 billion sits at a single Belgian depository, Euroclear.

No public ledger was required for any of that. What made the freeze possible was that intermediaries held the assets under jurisdictions that joined the sanctions. Four things are worth keeping apart: visibility says what can be targeted, custody says who can move the asset, jurisdiction says who can give the order, and coercion decides whether the order can be carried out. Being visible is not the same as being seizable, though visibility makes targeting easier wherever custody, jurisdiction or the key holder offers a control point.

Give Dalio what is his. Self-custody removes the depository. It does not remove identification of the funds, it does not remove physical or legal pressure on whoever holds the keys, and it does not remove sanctions at the points where value enters and leaves the system. A public balance is identifiable in a way a vault in Moscow or Zurich is not.

Then look at what happened to the gold that escaped. Its usefulness narrowed anyway. Russia's central bank began selling gold in late 2025 and sold 44 tonnes in the first half of 2026; by July 1 its gold reserves stood at 2,283 tonnes, the lowest since February 2020, with the proceeds helping cover a widening budget deficit. What was reduced was not possession but access to the deepest markets and convertibility into the main reserve currencies. Escaping seizure and keeping full use are two separate problems, and opacity solved only the first one.

The counter-argument on the other side, made by Michael Saylor in May, is that the transparency is a feature rather than a defect: it is what allows the holding to be verified, and therefore used as collateral, without trusting anyone's accounts. Both things can be true. The same property that makes a central bank's position legible is what makes an individual's reserves provable.

The text is fourteen months old. The world moved toward it

The framework in the August essay is not new. It is the one Dalio published on June 3, 2025, and the essay carries the book's title. What changed is the surroundings. Since the end of June 2025 total federal debt has risen about $3.8 trillion, from about $36.2 trillion to about $40.0 trillion. The 30-year yield reached 5.37 percent on September 10, its highest of this year. Foreign holdings of Treasuries fell in June, led by Japan, whose holdings dropped about $26.4 billion, or 2.3 percent, to $1.1167 trillion.

A conviction has to be maintained. It needs confirmations and reasons to keep believing. A subtraction needs nothing: it holds as long as nothing comes back onto the list.

What would break the thesis, in his own terms

Dalio names the conditions, and they are worth writing down because they are the honest test. First, bring the US budget deficit to about 3 percent of GDP, from what he projects at about 7 percent. Second, stabilise debt service against revenue instead of letting it crowd out the rest of the budget. Third, get private demand back to absorbing the issuance without the central bank buying what nobody else wants.

On the first, his 7 percent is his own projection and it is above the official one: the Congressional Budget Office's February baseline puts the fiscal 2026 deficit at $1.9 trillion, or 5.8 percent of GDP. Either way the gap to 3 percent is wide, and the direction since the book was published has not been toward it. If those three conditions did hold, sovereign debt would come back onto the list, and the subtraction that leaves gold and Bitcoin at the end would stop.

Three numbers we could not stand behind

Several figures are circulating with this story that we could not confirm, so we are not publishing them. A 90-day Bitcoin to Nasdaq correlation of 89 percent in May: we found no source for it, and the series we can source starts the year above 60 percent. A gold to Nasdaq correlation of 52 percent, described as the highest since October 2020: the measured move we can verify is Bitcoin's correlation with gold rising above 50 percent, which is a different pair. And a 30-year yield of 5.33 percent: the Treasury's own par yield series shows 5.31 percent on August 17 and 5.37 percent on September 10, so we use those.

What it means for our view

Bitcoin is trading near $77,300 tonight, with the day's low at $76,516. That keeps it above the $75,600 line our recent notes have treated as the bears' burden, and below the $81,000 to $84,000 band that capped May and marks January's breakdown level. Our three horizon views are unchanged and refreshed today with the same views on all three.

Dalio's essay does not change a view, and it is not evidence that the thesis is right. What it does is describe the mechanism by which the demand he expects would arrive: not from people who like Bitcoin, but from allocators working through a filter and finding little else at the end of it. Two dated events sit in front of us this week. The Senate's cloture vote on the CLARITY Act is set for September 15 and needs 60 votes. The Federal Reserve meets on September 16, and after August inflation came in at 3.4 percent, futures markets moved to about 90 percent odds of a rate increase. A central bank raising rates into a long end that keeps selling off is, in miniature, the tension the essay is about.

This is general market commentary, not investment advice or a recommendation to buy or sell any asset.