This essay is the first in a two-part series exploring why there will be no Plaza Accord 2.0 between China and the US, and why AI plays a crucial role in the debasement of the dollar.

Every relationship teaches the same lesson sooner or later. With enough power you can force someone to change, however the change rarely lasts and rarely goes the way you planned. The older version of that wisdom, the Buddhist one, is even simpler: control what you can control, that is, your own actions and your reactions, and stop trying to control others. Nations learn it more slowly than people do, but it binds them all the same.

At present, from a monetary perspective, America and China are each doing what they can to manage their respective positions. Both nations remain locked in what looks like a dangerous imbalance: America is running ever-larger twin deficits (simultaneously a fiscal (budget) and a current account deficit (trade)), while China continues to expand its massive trade surplus. Against this backdrop, much of the commentary still looks to a grand bargain to realign their currencies: a Plaza Accord 2.0.

We think that in fact, the realignment is already under way, and it will come through the debasement of the dollar… with no accord at all!

What Plaza was

What was the Plaza Accord exactly? By 1985 the dollar was badly overvalued, the result of the so-called Volcker squeeze. To break the inflation of the 1970s, Fed chair Paul Volcker had pushed interest rates to around 20% in 1980–81. As a result, global capital poured into dollars to earn those rates, and by early 1985 the dollar had risen roughly 50%, pricing American manufacturers out of world markets and swelling the trade deficit. In reaction, Congress was drafting protectionist bills.

Something clearly had to be done. That September, the finance ministers of the US, Japan, West Germany, France and Britain met at the Plaza Hotel in New York. They agreed to push the dollar down together: their central banks would sell dollars on the open market, and Japan and Germany would let their currencies rise.

It worked—fast! The yen went from about 240 to the dollar to about 120 within three years. But in retrospect, many say it only worked because America had leverage. Both post-war Japan and West Germany were security dependents with American troops on their soil. Tokyo also had reasons of its own for playing along. As Michael Pettis recently pointed out in a Substack essay, Japan’s current-account surplus had reached 3.6% of GDP, and Japanese economists had long worried about weak domestic demand. In 1986, the Maekawa Commission, set up by Prime Minister Nakasone, called for a “historic” shift from exports to domestic demand. A stronger yen was an integral part of that plan.

Today, none of those favorable conditions holds when it comes to China. Beijing sits entirely outside the American security umbrella. In addition, unlike Japan in 1985, it runs capital controls, sets a daily fixing exchange rate, and has spent four decades studying what seems to be its reading as the lesson of Plaza from the Japanese perspective: first a currency revalued under pressure, then an asset bubble, then a lost decade. To Beijing, 1985 Japan might have been willing to play along, but it lost a great deal in the process.

On Substack, Michael argues that this interpretation is wrong. In his view, Japan’s lost decades were not the result of the Plaza Accord, but of the Bank of Japan’s response to the stronger yen. To cushion the slowdown, the BOJ cut its discount rate from 5% to 2.5% between January 1986 and February 1987. The resulting cheap credit helped fuel a property and equity bubble that burst a few years later, plunging Japan into recession.

More generally, a negotiated revaluation, like the one Japan agreed to back in 1985, means hurting your own exporters on command—which China is not in a position to even consider, out of both self-interest and national pride. And so, right or wrong, Beijing’s reading shapes policy: no Chinese leadership will ever sign a Plaza Accord 2.0. And as we recently discussed on the Hidden Forces podcast, American soft power has eroded anyway, so the old coordination playbook is harder to run than it was in the 1980s.

Of course, there’s been a recent rapprochement between Washington and Beijing. Xi Jinping’s recent state visit to Washington put the two governments back at the negotiating table. But look at what was agreed: the trade truce extended to January 10, 2027, tariff cuts on $60 billion of goods, Chinese purchases of at least 10 million metric tons of US coal in 2027 and again in 2028, a communication channel for AI incidents, and Beijing’s pledge to approve foreign financial institutions, including those with US capital. A lot was discussed, but nothing was said about the exchange rate or a revaluation of the yuan.

Still, could all this pave the way for a Plaza Accord 2.0? In 1984, a year before the Plaza Accord, the Yen-Dollar Agreement pushed Japan to open its financial markets. First, Japan had to open up; then it would have to revalue its currency. Some will read the recent progress on the US-China front in the same way, but we don’t. In our view, as we’ve argued before, any opening by China will take place only through channels that Beijing controls and can narrow at will when circumstances demand it: licenses, Stock Connect, Bond Connect, and Hong Kong. In other words, as Japan in 1984, Beijing is willing to change who gets access to its financial markets. But unlike the open economy that Japan has been for a long time, it isn’t ready to let global markets determine its exchange rate.

In a recent talk at the University of Zurich, which Marieke attended, the China analyst Yanmei Xie described a shift in how Western governments see China. For decades the West hoped to change China’s economic model. They now accept that it will not change, and ask instead how to protect themselves from it. The same logic applies to the currency. Washington cannot make Beijing revalue the yuan the way it got Tokyo to revalue the yen in 1985. That is why we expect no Plaza 2.0. But Washington may not need one after all. The yuan has been rising anyway, on Beijing’s terms: the central bank set its daily reference rate at 6.72 per dollar on October 16, against 7.19 in early January 2025. What that reveals is that the rebalancing that Plaza once forced on the yen is now happening to the yuan without a treaty.

If nobody signs a treaty, what will do the work the Plaza Accord once did? How can a dollar that buys less every year in American shops still hold its value against the yuan? And who will pay for the adjustment: American savers, Chinese households, or Europe? The answers run through Washington’s debt, Beijing’s repressed savings, and the AI build-out now competing with the Treasury for the same money.

Debased at home, strong abroad

If no treaty is coming, the mechanism for rebalancing has to be the dollar itself. Specifically, two different things can happen to the dollar:

Read more

Kevin Xu started out in politics. He worked on Barack Obama’s first presidential campaign, then in the White House and at the Commerce Department. He moved into technology as GitHub’s lead for international expansion strategy, during the years when the platform became where the world’s programmers built software together, Chinese programmers included. Then he became an investor. He is now founder and chief investment officer of Interconnected Capital, a global fund investing in the hardware and software infrastructure of the AI economy. He also writes Interconnected, a bilingual newsletter on technology, investing, geopolitics and US–Asia relations. He calls the common thread restlessness: he likes starting over in a field he knows nothing about, and he prefers the overlap between disciplines to the disciplines themselves.

His work cuts against the usual story about China and technology. The common assumption is that Beijing chose open source as strategy, to commoditize the model layer and undercut America’s frontier labs. His essay Chinese Open Source: A Definitive History tells a messier story, which we discuss at length in the episode:

• The mid-1990s: Linux reached China on floppy disks carried back from Helsinki.

• The 2000s: Alibaba and its peers adopted open source to escape the licence fees they were paying IBM, Oracle and EMC.

• An uneasy truce: the state tolerated open source with reservations. GitHub was once blocked for a day, and the country’s best engineers simply stopped working.

• After the Huawei export controls: open source became a matter of survival, and only then did the state embrace it.

The 14th Five-Year Plan’s call for globally recognized open-source projects came some twenty years after the first floppy disk crossed the border. The government, he argues, arrived late and gets too much credit for foresight.

In this conversation, Kevin explains why the Chinese government didn’t plan for the importance of open source but rather embraced it. He explains that open-weight models open only part of the AI stack. They release the weights but not the training code or data. That is still enough to free a model from the hardware and jurisdiction it runs on, which is why Nvidia wants more of them. He sorts China’s AI landscape into three camps:

• neolabs such as DeepSeek, Kimi, and MiniMax;

• open-weight labs inside Alibaba, Tencent, and Xiaomi;

• ByteDance, closed-source all along, which shows that openness is not a state mandate.

On the money side, he reads America’s data center moratoriums as pauses rather than bans. He traces the build-out’s shift from corporate cash to debt, the turn that took Google’s free cash flow negative for the first time. He expects the model layer to settle into something like the iOS–Android split in smartphones. He names China’s underrated weakness as shallow capital markets, not chips, which is why its AI firms still look to the dollar when they want to grow. And he describes a country that is climate-change-pilled rather than AGI-pilled and may win on clean power what it cannot win on compute.

His case throughout is that the US, China, and Europe are driven more by the same commercial incentives than by the grand strategies attributed to their governments. He also argues that China, with more than 100 million factory workers and tens of millions of delivery drivers, will be the first to face the social cost of the automation it is racing to deploy.

Kevin writes Interconnected and is on X at @kevinsxu.

Marieke: Welcome to the Currency of Power podcast, where we explore how money, technology and geopolitics are reshaping the global order. We speak with the people who are building, regulating and challenging the new financial infrastructure, from currencies and stablecoins to AI, trade, energy and the competition for technological power. Our goal is to understand how the financial system is changing and who gains influence as it does.

I’m Marieke Flament.

Nicolas: And I’m Nicolas Colin.

Marieke: And today we talk about AI, open source, and what’s at stake between China, Europe and the US. And we have the perfect guest to talk about this: Kevin Xu. Super nice to have you here. Thank you so much for joining us.

Kevin: Thank you so much for having me.

“I love looking at the Venn diagram of things.”

Marieke: A few months ago, I read your piece on Substack, Chinese Open Source: A Definitive History, and I found it absolutely amazing. Highly recommended; we’ll have it in the show notes, and we’re going to talk about it. Then more recently, I saw posts from you saying that you were open to recording podcasts, so I just jumped on it and reached out. It’s really amazing to have you here.

You’re the founder and CIO of Interconnected Capital, a global technology fund investing in the hardware and software infrastructure of the AI economy. You also write Interconnected, a bilingual newsletter on technology, investing, geopolitics and US-Asia relations. What I think is really unique, and I’m sure this will come through as we discuss, is your perspective, which is very cross-disciplinary. You started your journey in politics. You worked on Barack Obama’s first campaign, and later in the White House and the Commerce Department. You then became a technology operator, including as GitHub’s lead for international expansion strategy, before becoming an investor and a writer. I think that’s an incredible background for sharing a lot of thinking about what’s happening right now, because you’re at the intersection of AI, open source, chips, industrial policy and the US-China relationship. Your work is cited and published across many, many outlets. I won’t name them all here, but whatever important outlet you can think of, I think you’ve been quoted in it or helped write for it. So that’s awesome, and we are really super happy to have you with us today at Currency of Power. Thank you.

Kevin: Of course, happy to be here.

Marieke: Maybe let’s start with your journey, actually. Tell us about your journey from Obama to GitHub, and now investor and writer. What’s been the common thread?

Kevin: I think the common thread, on a personal level, is just a healthy level of restlessness and easy boredom, in a lot of ways. With a lot of these disciplines, you can dive in very, very deep. But if you put your head up above the water, sometimes you see other things on the horizon, other things that are more interesting. From a personal perspective, I love to start over. I enjoy knowing nothing and jumping into a new field. You learn a lot very quickly that way. It can feel very scary from time to time, but ultimately, at least so far, it’s all turned out pretty well for me.

The other thing, which is probably more of a personal preference, is that I love looking at the Venn diagram of things as opposed to the circles themselves. The circle itself is fine. You can dig very deep and become a longtime domain expert in a particular discipline. But I do think the intersection, the interconnection, the overlap between circles is just as important in terms of informing what’s going on in the future. I find that underexplored, and I personally just enjoy looking at those things anyway. That turned out to be more or less what I do every single day, whether on the writing side or on the investing side.

Nicolas: Does your writing inform your investing, or is it the other way around?

Kevin: It kind of depends on the day. It’s very hard to actually tie those two things together. I would say one thing: I started writing before I started investing professionally. One common thread among almost all the investors we’ve come to know and respect, whether it’s the Warren Buffetts of the world, Stan Druckenmiller, a lot of fantastic European investors, Chinese investors as well, is that they’re all very good writers. They write differently, they don’t all write the same, but they are all good, compelling writers in their own way. Writing, to me, is still a very important discipline for thinking clearly, and I don’t think any investor can be good for a long time without being able to think clearly. He or she can be good for a year or two, a quarter here and there. But to sustain yourself over the long term, you need a very clear process to flush out the silly stuff in your brain, really distill the important stuff, and then hopefully act on it in a way that makes yourself and your investors some money. So I saw that natural connection, and that led me to do both. But as far as whether today’s writing informs my investing decisions tomorrow? Honestly, not really.

Marieke: I think that’s really music to our ears, because Nicolas and I have been writing. Nicolas, you’ve been writing for over a decade. But it’s true that you can only express something clearly, and therefore do something with what you’ve thought about, if you’re able to write it down.

Should we move on to open source and your essay? Do you want to start, Nicolas?

Nicolas: Yes. I wanted to start with a very, very basic question, because we have to assume that part of our audience is not familiar with the world of software development and what open source is exactly. Can you explain what open source is about, the origin of the open source movement, how open source differs from free software and obviously from proprietary software, and why it’s been so successful in being deployed all over the world and powering, or supporting, most of the software applications we use on a day-to-day basis?

Kevin: Absolutely. Open source, in a nutshell, is actually not a kind of technology. It’s more a mode of building technology. You can literally visualize a developer sitting in Brazil logging onto the internet and another programmer sitting in France logging onto the internet, and they’re all building this thing called, say, Linux, which is the largest open-source operating system in the world right now and originated on the campus of the University of Helsinki in the 90s. And these things persist.

Now, why did this happen? Originally, it was a response to all the closed-source, proprietary software that dominated our world in the 80s and 90s. You can think of it literally as the CD we used to buy and put into our computer to get software or a floppy disk. Then it became something you could download from the internet, but you had to pay for a copy. It could be very expensive, which made a lot of tech companies, in the US in particular, a lot of money selling what we call closed-source or proprietary software.

The pushback started in academia first. I believe it was initially on the campus of MIT, where people said, “Why do I have to pay a company to update my printer software to make it print better? This situation is ridiculous. I’m a computer science professor. I can write this software myself. I’m going to write a version of it and put it out into the world for everybody to use.” That’s the free software era of open source. Free software actually predates open source slightly.

But then people said, OK, free is good, but how do we change it? How do we improve it? What’s the mechanism through which this software can evolve? Because software has to evolve almost in perpetuity. If you buy a machine, a laptop, or a printer, it just sits there until you want to buy a new one. But the software is what runs underneath. That evolved into the organic global movement we now call open source, where everybody has the right, the privilege, and the access to download, customize, redistribute, and actually sell this piece of free software that is just sitting on the internet, literally. That eventually became the dominant way we all use software, whether we know it or not. Everything you use on a daily basis has at least one or two, probably fifteen, pieces of open source software sitting inside, powering it.

Nicolas: Can you say a few more words about the legal side of things? Every open source component has a license attached to it that protects its openness in perpetuity. You can’t take something that’s open source and turn it into closed-source or proprietary software, and that’s because of the license, if I’m not mistaken.

Kevin: That is mostly correct. I won’t bore your audience too much with the legal side of open source, but there is actually a huge distribution of licenses governing different open source software, and you, as the creator, can choose what you want your license to be depending on what you want out of it. There are many licenses, and this will be very relevant when we talk about Chinese open source; that is what we call permissive, meaning you can download the software, change something, turn it into closed-source software, and sell it for money. That’s totally OK if it’s licensed under these permissive licenses. Very, very hands-off. Then you have other licenses that are much more restrictive, in the sense that you cannot do what I just described. Or you cannot resell the software on a cloud platform like AWS or Azure, for example, but you can do whatever you want with it in a nonprofit or academic setting. That variety is actually something open-source people fight about a lot internally as well. What is really open source? What is not open source enough? I won’t get into the nitty-gritty of the cultural challenges with legal licensing, but it is very much a spectrum, and it’s a live issue in the industry.

Nicolas: Super interesting. Maybe one last question on open source. My eyes were opened to open source, and I understood most of it, when I read the landmark article by Tim O’Reilly published in 2004, called Open Source Paradigm Shift, which didn’t really appear in your essay on Chinese open source. Does O’Reilly in particular, or the whole community connected to him, have some influence or some audience in China, or are those completely separate?

Kevin: I think the influence is completely there. I chose to under-represent, if you will, some of the typical Western voices in open source, because I think the story of what happened in China is fascinating in its own right. But the undercurrent of how China even started doing open source to begin with all came from the West. In fact, even today, a lot of the norms, industry standards, and best practices in open source are still mostly learned from the West. Even these licenses we talked about: DeepSeek R1, which became the “DeepSeek moment” a little less than two years ago and shocked the world. One of the reasons it became such a shock is that it was licensed under one of these super permissive licenses where you can do whatever you want, and the name of the license is literally the MIT License. It came from the MIT campus.

What’s interesting about open source is this. When we think about the geopolitics of AI, or tech in general, borders matter a lot. One country versus another, this country doing its own thing, the other country imposing export controls. That’s all fine and good. But when it comes to open source, the permeation of information and know-how is seamless across borders. It’s almost borderless by design. So to your point, they’ve consumed everything Tim O’Reilly has written, everything all the open-source leaders have written, and then made it their own and adapted it to their own country’s conditions, situation, and needs. That’s how their trajectory became a little bit different.

“It became a matter of survival.”

Marieke: Maybe this is a good turning point to deep dive into your essay, Chinese Open Source: A Definitive History At first glance, there seems to be a contradiction. How would a centralized, more controlling state embrace open source? What’s fascinating in your essay is the story of why China actually has a very long history with open source. So maybe let’s start with the history of China and open source, if you can walk us through it.

Kevin: Sure. The one top-line takeaway I hope all your listeners take away, before I get into the stories, is that this is an example where the Chinese government came very late into the picture. Open source, as the form of technology development I described, more or less existed organically in China until, essentially, the US export controls that put Huawei on the blacklist. We’ll explain the logic of that. That’s the big takeaway, and I think it’s very different from what a lot of people think about China: that the government has big strategic thoughts, tells everybody what to do, everybody just does it, and then you get a lot of EVs on the streets, batteries in your home, appliances, and whatnot. Some of that is true, depending on the sector. But for open source, the government was late to the picture.

Open source entered the Chinese economy organically in the mid-90s, not long after Linux itself was born at the University of Helsinki. In fact, I think the story is that a Chinese researcher happened to be studying at the same university, learned about Linux, carried a copy of it back to China on a bunch of floppy disks, and started playing with it.

What was attractive about open source initially is that at the time, the late 90s and early 2000s, there was already a corps of internet companies that came up in China as part of the dot-com boom, like Alibaba, Tencent, and a bunch of others that no longer exist. They were paying a ransom’s worth of fees to closed-source software companies, particularly from America, just to keep their websites running: IBM, Oracle, and EMC in particular. They saw open source as a way to get rid of Western proprietary software so they could run their own stack more cheaply and also have more control over their own software and IT stack. That’s phase number one.

Then, as that became successful and Alibaba became successful, everybody started throwing their energy toward open source as a way to keep building. The important part is that China hadn’t really contributed back to open source at this time. They were takers of whatever had been produced in the West. They adapted it, but they didn’t contribute back.

This was also the time when GitHub, where I used to work, became a mild global sensation among hackers, because it made open-source collaboration that simple. It used to be very hard. You had to be on some obscure email listserv to coordinate, then download the software, make some changes, upload it to some other source-control thingy, and somebody else had to go in and do something else to make it work. GitHub made it that simple for everybody around the globe, including China. So that’s phase number two, where open-source norms became much more democratized by an American company funded by VC money like everybody else, but with global impact.

Then you move on to a more domestic phase, where China started making its own open-source communities, its own meetups, and its own hacker groups. They started grabbing what was amazing from the West but also assimilating it into their own setting to make it work for themselves.

There was also an interesting period when open source became almost a vehicle for social advocacy, and this is where the nuance between the Chinese government and open source really came in. As we all know and appreciate, the Chinese government does not really appreciate tools that let people organize online very, very easily. That’s why you don’t have Facebook in China. That’s why you don’t have Twitter in China. And for a little while, GitHub was also in the crosshairs because of this possible social dynamic, where people could organize on GitHub, since it’s a social network, in a way, just for programmers and geeks. GitHub was actually shut down for 24 hours, and all hell broke loose in the Chinese tech ecosystem because nobody could work. Your most productive, talented programmers in universities and tech companies were just saying, “OK, there’s nothing to do; GitHub is down.”

That became a tug-of-war, an uneasy relationship, as I call it in my piece, where the Chinese government probably didn’t like some aspects of open source, to be perfectly honest. But it’s also such an important engine for the country’s most talented people to build technology that you need to not be so, I don’t know, authoritarian about it, if you will. So that’s the first half of Chinese open source. I’ll pause here to see if there are any follow-up questions before I keep rambling on to the other chapters.

source that the

Kevin: There have been attempts at a split. What ended up happening, in response to these elements of open source the government may not appreciate, is that a local alternative to GitHub started being propped up. It’s called Gitee, G-I-T-E-E. You can think of it as China’s own GitHub, which is all fine and good. And then what ends up happening is exactly what you said. Gitee is used only by Chinese programmers. Does that work? OK, fine, you can upload some code onto Gitee and do open-source collaboration with only Chinese programmers. But what does that do to help your software advance when you could get the best eyes and the best talent from around the world to look at your software, file issues, improve it, give you use cases you’ve never thought of, give you better ideas, and interact with them all seamlessly on GitHub? The magic of open source is that it’s global by default. As soon as you nationalize it, that part is gone, and it’s just another software development tool without much to it.

So that’s still a tug-of-war. It’s not a resolved question, but China is still attempting to have alternatives to the Western platforms that host important open-source software. We have GitHub, and now we have Hugging Face, which is the platform that hosts AI models, datasets, and so on. China has its own version of that too. It’s sponsored by Alibaba and called ModelScope. It’s fine. But it’s just not going to get the global traction that Hugging Face has by default.

So there’s always going to be a little bit of tug-of-war, which is why, again, for audiences who don’t follow this super closely: We hear a lot of narratives saying China loves open source and is going to push open source forever for all these big strategic reasons, whether it’s overcapacity or destroying the US frontier models. What I’m trying to say is that the relationship has always been, and still is, very, very uneasy.

Nicolas: I was about to say that China is a large country with lots of software developers. Can’t you reach critical size with China alone to deliver high-quality open-source software, without relying on developers in the rest of the world?

Kevin: That’s a very interesting question, and it gets into the second half of the history, which is when China started producing its own open-source software. Everything I’ve described up to this point, which takes us to about the mid-2010s, give or take, is China taking, taking, taking. And then came US export controls, in particular during the first Trump administration, which put Huawei in the penalty box. I’m a hockey fan, so I like to use random hockey references here and there. They got put in the penalty box. That meant open source wasn’t just a nice-to-have for building your technology company anymore. It became a matter of survival for your company and possibly your country.

So Huawei, and all the other companies that had been playing with open source longer than Huawei, because Huawei came late to the open source picture as well, all had to throw more energy into producing software using open source. They could have kept it closed for themselves, but they also saw the magic and the power of distributing open source globally. So Huawei started not only making its entire stack by itself, because it couldn’t access Western technology anymore, from operating systems to chips to databases to servers. It also open-sourced all of that through its own open-source foundation to push it out to the rest of the world. One, because you just make better software if more people have eyes on it, period. You could make it into your own little ecosystem, but who’s really going to care if no one else can try it, improve it, raise questions about it, or find bugs in it? That’s a long but very important iterative process that makes open source work the way it does.

Two, and this is where the government does come in, in the early 2020s: The government finally issued a directive from the very top saying we want China to have two to three globally recognized open-source projects, as part of its five-year plan. That was the 14th Five-Year Plan, so not the most recent one, the one right before. So the government came in 20-ish years after Linux first crossed the Chinese border and said, “We want open source; we want to promote open source because it’s such an important part of our technology self-sufficiency story vis-à-vis external pressure. It’s obviously making our companies better too. There are some problems with open source that we don’t like, but net-net, it’s a very good thing, and we now want to lean into it.” And this was all before AI. The AI proliferation that came afterwards, with China dominating the open-source AI scene, made open source even more self-evidently something they want to support and should support, even though there are still reservations about elements of how it works.

“Building God in a box is great, but what's in it for me?”

Read more

In the early 1970s, the oil sector was reorganized following major geopolitical changes, culminating in the oil shock of 1973. Today, we see a parallel in the current evolution of frontier AI labs and the massive restructuring of the oil sector in the early 1970s in the context of widespread geopolitical turmoil. If the parallel holds, AI could be approaching a fundamental reorganization of its entire value chain. It’s definitely not too early to reflect on what it would look like on the other side.

We are not claiming that AI will replay the history of oil exactly. It is just that, as the famous quote goes, “history does not repeat itself, but it often rhymes,” and we believe the history of oil provides a useful framework for considering what may happen next in AI—and by extension, in compute, energy, and finance.

Specifically, the parallel is that both industries grow capital-intensive and financially interconnected at the precise moment their foundational assumptions begin to fail. In the 1970s, the oil shocks exposed the fragility of an industry built around cheap and perceived abundant supply, leading to a profound reorganization of who controlled the assets, who could provide the services, who controlled the resources, who financed the infrastructure, and who ultimately captured the profits. AI may be approaching a similar point, leading to a fundamental reorganization across its entire value chain.

Indeed, just like the oil industry in the postwar era, the AI industry seems to have been built around the expectation that demand for compute will keep rising, that ever larger amounts of capital can be deployed in data centers and GPUs, and that frontier labs will continue to absorb that capacity through increasingly expensive models. But the physical and financial structure underneath those ever-growing assumptions is becoming harder to ignore. If compute becomes pricier while large-scale training becomes harder to build and finance, the current value chain will stop making economic sense.

The question is what happens when that value chain starts to break apart. To us, it’s obvious that the history of the oil industry provides an answer.

The oil industry before the shock

The oil industry of the postwar era was unusually integrated. A handful of large companies controlled much of the chain from finding oil underground to selling gasoline at the pump. The so-called “Seven Sisters”—that is, Exxon, Mobil, Chevron, Texaco, Gulf, BP, and Shell—all combined exploration, production, transportation, refining, and distribution at an enormous scale. This structure worked because the economics of the industry supported it: oil was abundant, demand was rising, transportation costs were falling, and these powerful companies could deploy capital across the entire chain.

The system also depended on a particular geopolitical order. America became deeply involved in the Middle East after World War II, notably through its relationship with Saudi Arabia. In 1945, President Roosevelt famously met King Abdulaziz aboard the USS Quincy, helping establish a strategic relationship that would become central to the postwar energy system. Meanwhile, the Texas Railroad Commission, the original regulator of the oil industry, demonstrated that coordinated production controls by US authorities could exert enormous influence over oil supply and prices worldwide.

Then the structure began to change. In 1956, Egypt’s President Gamal Abdel Nasser nationalized the Suez Canal, triggering a military intervention by Britain, France, and Israel that the US opposed and helped bring to an end. The episode became an early demonstration of the changing relationship between Western powers, resource-producing countries, and strategic infrastructure. Then in 1960, Iran, Iraq, Kuwait, Saudi Arabia, and Venezuela founded OPEC, and over the following decade, producing countries increasingly asserted control over their natural resources and demanded a larger share of the prize.

As a result, by the early 1970s, the old arrangement was under severe pressure. National oil companies were taking greater control over reserves, and the previously dominant Western oil companies faced governments increasingly willing to renegotiate contracts or nationalize assets. Ultimately, the 1973 oil embargo brought the system to a breaking point and accelerated its overhaul, exposing how vulnerable consuming economies had become to decisions made by producing countries. After the shock, the Western oil companies retained enormous technical and commercial capabilities, but their position in the value chain had changed for good, creating space for new businesses around services, technology, refining, trading, and finance.

The AI version of vertical integration

AI has developed its own form of vertical integration, even if the overall structure looks very different from oil. At the bottom sits the physical system: electricity generation, transmission, data centers, networking equipment, and accelerators such as GPUs, TPUs, and custom AI chips. Above this physical system sit the models, and then above the models are products and applications that turn intelligence into something end customers can actually use.

The so-called frontier labs, such as OpenAI and Anthropic, as well as large players like Google, SpaceX, and Meta, increasingly reach across several of these layers, signing enormous infrastructure deals, securing GPU capacity, developing their own chips, building or leasing data centers, developing their own models, and distributing those models through consumer and enterprise products. This structure makes sense during a period of explosive demand. An AI company that can secure more compute can train larger models, improve its products, attract more users, and raise more capital, with those new users then supporting further infrastructure investment. Ever since 2022, that feedback loop has become one of the defining features of the AI boom, with the circular financing between Nvidia, the hyperscalers, and the frontier labs revealing a high level of integration and coordination across the system.

But now, this model-centric logic is already coming under pressure. Chinese developers, led by DeepSeek, have closed much of the capability gap with leading closed models while making increasingly capable open-weight models available at much lower cost. The result is that the value of owning the model itself is becoming harder to defend as a durable moat. Open-weight models can be downloaded, adapted, and run on different infrastructure, allowing customers to shift between providers and putting pressure on the pricing and margins of closed models. The rise of these models therefore challenges one of the central assumptions behind the current AI value chain: that ever larger investments in frontier models will translate into durable economic rents for the companies that build them.

This scenario creates a further challenge to the whole loop: there is a fundamental divergence in behavior between software and physical infrastructure. A model can be copied or become obsolete almost instantly, while a gigawatt-scale data center requires power generation, grid connections, land, cooling systems, construction, networking, chips, financing—and years of planning.

The result is a growing tension between the speed of software development and the speed of physical infrastructure. Across the compute-energy stack (energy grid → data centers and hardware → models → integration), each successive layer relies entirely on the foundational tier below it. In other words, the economic viability of the model layer depends on the cost and accessibility of the underlying physical infrastructure. Consequently, a significant portion of the sector’s financial architecture rests on critical assumptions regarding the future availability and affordability of power and compute capacity.

The coming compute shock

Another feature of the current cycle echoes the 1970s: enormous amounts of capital are flowing into AI infrastructure based on expectations of continued growth. The largest US hyperscalers are investing hundreds of billions annually in data centers, chips, networking, and power. For 2026, Amazon, Alphabet, Microsoft, and Meta together plan roughly $725 billion of capital expenditure, according to a Goldman Sachs aggregation reported by Yahoo Finance—Amazon $200 billion, Alphabet $175–185 billion, Microsoft about $190 billion, and Meta $115–135 billion. McKinsey estimates that global data center investment will reach roughly $6.7 trillion to $7 trillion cumulatively by 2030, including approximately $4.3 trillion in servers and storage, $1.3 trillion in power, cooling, and network infrastructure, and about $1 trillion in land, site, and construction costs.

In the 1960s-1970S, the oil industry made similar bets in the years before the shocks, building ultra-large crude carriers, ports, refineries, pipelines, and other projects around assumptions about cheap and plentiful crude. Then when the underlying economics changed almost overnight, some of those assets became liabilities. AI infrastructure has a similar vulnerability because a multi-gigawatt campus can take years to plan and construct while the economics of the hardware inside it can change much faster.

This creates a critical financing problem. An investor committing capital to a data center today has to make assumptions about how demand for compute will develop several years into the future. That means forecasting electricity prices, GPU economics, hardware utilization, training demand, inference demand, and the value of the resulting capacity. As we know, GPUs also have relatively short economic lives compared with power plants or buildings, and successive generations can dramatically increase the performance available for a given dollar of capital. Small changes in any of these assumptions can therefore have a large effect on returns.

There is also considerable uncertainty about what will drive that demand. Training and inference have different compute requirements and different growth dynamics. Training and other R&D have long been major sources of demand, as companies have invested in ever larger models and increasingly costly training runs. Inference is different: it depends on how widely models are used, how often they are queried, and how much compute each task requires.

The assumption is that inference will eventually account for a much larger share of total compute as AI moves from model development to widespread use. But that shift has not yet happened to the extent often assumed. On the contrary, training’s share of compute has remained relatively stable so far, even as inference has grown. We therefore do not know how the balance between training and inference will evolve, which makes it harder to forecast the demand for the infrastructure being built today.

Now, there is still an important difference between AI and oil. OPEC countries could deliberately restrict the physical supply of a scarce natural resource. AI has a more complex supply system. New chips can be designed, data centers can be built, models can become more efficient, and software can extract more work from the same hardware. The supply of compute can therefore expand not only through new physical capacity but also through technology and better use of existing resources. Unlike oil scarcity back in the 1970s, compute scarcity is thus partly a question of technology and capital allocation, rather than simply a result of geopolitical tension.

Yet that does not make the physical constraints any less real. Electricity still has to be generated, transmission capacity built, chips manufactured, data centers connected to the grid, and capital committed years before the resulting capacity can earn a return.

Overall, the very integration that drove the first phase of the AI boom could become a source of strain. As infrastructure becomes larger, more expensive, and longer-lived, while models and demand evolve much faster, the economics of owning every layer of the stack may begin to diverge. That is where fragmentation can start to create value.

From oil majors to the new AI value chain

Read more

Marcin Zarakowski trained as a lawyer at the Jagiellonian University in Kraków and moved to Switzerland to work for layer-1 blockchain foundations—first the Lisk Foundation, then BSV, where he became general counsel. His job was to sell blockchain to enterprises and governments, and the objection he kept hitting was the same: what happens when someone loses their keys or has their wallet drained?" In traditional finance there is always someone to call. In decentralized systems, there is no one.

Patrick Prinz grew up in Linz, started in corporate and investment banking at Deutsche Bank and Citi, moved to strategy consulting at PwC Strategy&, and spent several years on the buy side in South Asia, East Africa, and Zurich. He came to digital assets through high-net-worth clients and saw the same contradiction Marcin saw from the infrastructure side: an industry built on libertarian ideals was becoming institutional at speed, and the compliance and incident-response apparatus had not followed.

Together with Roman Bieda, who previously built blockchain-analytic tools at Coinfirm, they founded Recoveris in Zug in 2025. The firm traces stolen crypto, coordinates freezing requests across exchanges and stablecoin issuers in multiple jurisdictions, and runs source-of-funds analysis for financial institutions onboarding digital-asset clients. Half the twenty-five-person team builds product—including an AI model designed to keep pace with criminals who already use AI at every stage of a scam. The other half investigates: a former Argentine prosecutor who set up her country’s crypto task force, a former Guardia di Finanza officer with a decade on darknet markets, and analysts from the investigation departments of large exchanges. Recoveris counts the US Secret Service and Interpol among its users, has a joint distributed-ledger analysis offering with EY, and is a strategic partner of the ICC FraudNet.

In this conversation, Marcin and Patrick walk through the case that broke the ZondaCrypto scandal—a Polish exchange whose CEO went on camera to prove reserves and ended up pleading on live video for the disappeared founder to send back the private keys. They describe the race between investigators and criminals after a theft, where assets move in seconds and mutual legal assistance between countries takes six months. They explain how the choke points work: centralized exchanges with fiat rails, Telegram groups of five hundred compliance officers, and the blacklisting power of stablecoin issuers like Tether and Circle. They detail VC-impersonation scams—rigged proof-of-liquidity sessions, CCTV cameras reading seed phrases, and follow-on recovery scams run by the same perpetrators—and explain why criminals have largely abandoned privacy coins in favor of stablecoins.

Both argue that the future of financial intelligence is blockchain intelligence, that Europe needs its own analytic capabilities and euro-denominated stablecoins rather than extending US jurisdictional reach through dollar-denominated tokens, and that the shift from cryptocurrency to tokenized traditional assets is where the real scale—and the real risk—will land.

Recoveris is on X at @RecoverisTeam. Marcin is on LinkedIn; Patrick is on LinkedIn.

Marieke: Welcome to Currency of Power podcast, where we explore how money, technology and geopolitics are reshaping the global order. We speak with the people building, regulating and challenging the new financial infrastructure — from currencies and stablecoins to AI, trade, energy and the competition for technological power. Our goal is to understand how the financial system is changing, and who gains influence as it does. I’m Marieke Flament.

Nicolas: And I’m Nicolas Colin.

Marieke: Today we will deep dive in the fascinating world of law enforcement and on-chain traceability with Marcin Zarakowski and Patrick Prinz, co-founders of Recoveris. Crypto was built and sold as a system of financial sovereignty and irreversible transactions. But when money is stolen, laundered through bridges and mixers, or linked to sanctioned networks, reversibility reappears — through the ability of stablecoin issuers, exchanges, law enforcement and forensic firms to trace, freeze and ultimately return value. And Recoveris works exactly in that new terrain, translating public blockchain data into intelligence that can move private companies and public authorities to act. So what does this tell us about power in the emerging tokenized financial system? Marcin, Patrick, welcome to the show!

Marcin: Thank you very much. Good to see you.

“What happens if a user loses their private keys?”

Marieke: Awesome to have you guys here. Let’s start with the foundation of Recoveris. Like what did you see and what triggered you to actually create Recoveris?

Marcin: I’m happy to start, and I think it stems from my personal experience in the world realm of blockchain digital assets. I’ve been in the blockchain space for nine years now. I used to work for two layer-one blockchain foundations. And what we were initially trying to do was sell blockchain. What do I mean by that? We were talking to different companies—blue chip companies, enterprises, but also the public sector—trying to convince them to utilize this technology, to build certain use cases using DLT.

And one of the obstacles we saw for adoption—or what stifles the adoption of digital assets and blockchain technology, especially among financial institutions—was the question of what happens if a user loses their private keys, if the private keys get stolen, or if access to certain tokenized assets, or even their identity, is compromised. What then?

Using the analogy of somebody forgetting the PIN to their credit card or the login credentials to an online banking, there’s always a possibility to issue a new one. There is somebody you can call. Whereas in decentralized systems, there’s nobody you can call. So that was one of the ten challenges we faced when “selling” blockchain.

And then I started really investigating further what happens if somebody’s crypto is being stolen. And as crypto was getting more and more adopted, these situations were happening more and more often. Some of them were reaching public opinion via different news.

And what I realized is that if you’re a victim of a scam or a fraud or your crypto is being stolen, in principle, to simplify, you have two options. One, you try to trace the assets. Obviously, it requires some skills and tools, so you will not do it alone. You will hire experts for that. You may go to blockchain analytics firms. So you will trace the assets. But what then? You know where they are, but what is the next step?

The other alternative is you go to law enforcement, and you try to file a criminal complaint. Let’s assume for a second that you’re lucky enough that the person you talk to understands what crypto is and how it works. But the first thing they will do—they will need to do the tracing. And I thought this was a huge waste of time.

So our idea was to build a one-stop shop where somebody whose assets were stolen can come, and we can do the investigations, we can trace the assets, and we can also help and coordinate with the recovery. These were the origins of Recoveris. We saw huge demand, unfortunately for the people but fortunately for us, for our services.

From then we built on, and we are not just tracing assets that are stolen and that are moved by the criminals, but we are also tracing backwards to check the origins of the funds. The purpose here was more for source of funds purposes—so whenever assets are to be onboarded in the financial system, certain checks need to be done during the due diligence process. So we can also do the same analysis on the blockchain. The transactions are there. The data is there. We just need to properly document it and analyze that. This is also another growing segment.

And to finish, while doing all this, we realized that in order to scale it up, because the demand was growing, we cannot just hire more world-class experts. A, because it’s just simply not efficient. B, because talent is very scarce—we are competing with leading companies, and we are competing with the public sector. And finally, with the use of AI and AI being used by criminals, we cannot simply have humans analyze transactions of AI agents that can be much faster and much more complex. So we decided to start automating the whole investigative workflow. We started building products that help companies like ours or the users of investigative tools, and we also started training an AI model, which I guess we’ll have time to discuss later.

Patrick: And if I may add, Marieke and Nicolas, I’ve observed over the yeayears—Ime from the financial industry but spent quite a significant time dealing with digital assets for high net worth individuals. I found it very contradictory for a while. There was this initial world of kind of very libertarian thinking, almost anarchistic.

Marcin: Yes.

Patrick: And if you compare it to where we are today—it’s highly institutional. As a matter of fact, I challenge you to name me one layer 1 blockchain that doesn’t have institutions as their main target audience now for growth. So what does that imply? It implies you have to meet the requirements; they have to be an issuer and to be a facilitator as a regulated entity. You have to deal with compliance and AML; you have to deal with incident response, audits, etc. That’s the name of the game today and that’s what we are catering for.

Marieke: That’s super interesting, you both come at it from different angles and what you are saying is that ultimately for adoption to exist, there is a certain set of tools and systems that are actually needed. Before we get into some of your case studies—which I think are absolutely fascinating—I think it’s really interesting also to understand all the different skill sets that you have to have combined across the table. If I understand it correctly, there is a part which is investigation. There is a part which is legal, so you need former law enforcement professionals, and people who are also building a product because you’re talking about a platform. Give us an idea of what makes the make-up of teams at Recoveris, the functions you have to actually make all of that come to life.

Patrick: Marcin, before you go into the details of our team, I just want to give an analogy. I sometimes think about the automobile industry and the change from the combustion engine to EV. It’s not just that the car works slightly differently; the entire supply chain looks different. And here we are looking at a paradigm shift in that technology, the infrastructure that is used to exchange value. That implies that all the stakeholders along the value chain have to adapt the tools they use and processes but also know-how.

Marcin: So today, we are a team of 24, 25 people. And half of that is product. So basically engineers, software developers, and product managers that are working on the tools that we are building in-house to automate the workflow. And also we have an AI expert that is building the model and training the model.

The other half of the team is mostly composed of investigators, and these people come from former law enforcement. We have a former prosecutor from Argentina. She set up the whole crypto task force for the whole country where all the cases were flowing in. We have people from Guardia di Finanza, who spent more than a dozen years there, working in cybercrime units, taking down some of the largest darknet markets. We have people that used to work for investigation departments of large VASPs. And also we have people like our co-founder Roman, the third one, who used to build blockchain analytic tools. He knows the limitations of those tools, and he knows where they can excel.

But coming back to your question, when you look at a blockchain-based digital asset, you can argue that everything is public; every transaction is stored in a publicly accessible immutable ledger. However, then this would make it pretty easy to trace. So one could argue that everything is indeed traceable. But on the other hand, there are some factors that make it much more complex.

First, a blockchain-based digital asset allows you to transact value peer-to-peer, meaning there is no centralized intermediary that facilitates the payment, unlike traditional financial systems. We don’t have banks, corresponding banks, payment service providers, or credit card issuers. There is nobody to call when the fraud happens, and there is nobody that can stop certain transactions. This really complicates the issue.

The second factor is the speed in which digital asset transactions can be done. The value can be transferred not in a matter of hours or days—it can transfer in a matter of seconds, which is also unprecedented.

And thirdly, almost every case becomes cross-border immediately. Because we can have a victim that is based in Switzerland. We can have criminals, perpetrators, sitting somewhere in one of the scamming compounds in Southeast Asia. And in a matter of seconds, the assets are stolen from the victim and land on multiple different crypto exchanges—starting from Seychelles, the UAE, Europe, the US, and also some other, let’s say, jurisdictions where it’s really hard to enforce anything.

In order to coordinate this, of course, first you need to do the proper tracing, so necessary blockchain analytics tools are required; therefore, staff and people that know how to use them properly are also required. But the element of coordination of these cases, of multiple stakeholders in different jurisdictions speaking different languages and being accustomed to different ways of operating—this is important, and time is of the essence. The most important part for us in every single case is to freeze the assets as soon as possible.

And there is a constant race between law enforcement and the victim on the one hand and the criminals on the other side. They want to steal the assets and off-ramp into fiat as soon as possible. This off-ramping happens only at centralized crypto exchanges that have fiat payment rails. And on the way there, they will use different obfuscation techniques to make our work harder. So there is this race.

So you really indeed need people with different skill sets and different profiles that not only can trace the assets but also are able to coordinate these cases, can understand legal frameworks in different jurisdictions, and also have networks of contacts with some exchanges. I’m not going to reveal the secret—sometimes we reach directly on WhatsApp or on Telegram to law enforcement agents that we used to work on several cases with in another country, or we try to reach out via different channels to compliance people in a crypto exchange asking them for a courtesy freeze, making it plausible with evidence that the assets they’ve just received are really illicit. So all this really requires a different skill set.

“Is it still a human being or is it an agent representing me?”

Read more

What made the 1970s pivotal was not only the oil shocks that reshaped energy markets and geopolitics. It was also the reinvention of the global financial system after the Nixon Shock of 1971, when the US ended dollar convertibility into gold—a pillar of the Bretton Woods system established in 1944.

This reinvention went far beyond gold. It transformed who could participate in financial markets, what instruments were traded, how risk was managed, and where global finance was conducted.

Over the 15 years that the reinvention lasted between 1971 and the London Big Bang in 1986, the world moved from a post-war world of fixed exchange rates, capital controls, regulated deposit rates, relationship lending, and nationally segmented markets to a dollar-based fiat system of floating currencies, volatile interest rates, freer capital flows, offshore funding, securitized credit, and rapidly expanding futures, options, and swap markets.

By 1986, financial risk—especially currency, interest-rate, and credit risk—could increasingly be priced, transferred, and leveraged through markets rather than remaining mainly on bank balance sheets or constrained by regulation.

In December 2025, in our article 2025: Year of the Trump Shock, we made a parallel between the Nixon Shock and the Trump administration’s then-emerging financial agenda—especially its push to place dollar stablecoins at the centre of a new monetary architecture. Our argument was that this agenda could have consequences far beyond crypto: it could reshape the global financial system itself. The past few weeks have felt like an acceleration of our thesis, and with it a crisper understanding of what the new system will look like.

To put the acceleration in context, just consider a few very recent events:

🇺🇸 On July 31, 2026, the US Treasury joined Japan to buy yen in support of the JPY, an event we discussed recently in The Dollars Japanese Problem. Then on August 19, 2026, the Treasury announced that it would increase 10–30 year Treasury buybacks starting in September, an attempt to tame the 30-year yield, which hit a 19-year high of 5.34%.

With those two actions, the Trump administration is moving toward what economists call yield-curve control: a form of financial repression in which long-term interest rates are shaped by state intervention rather than set solely by markets. The approach remains soft—for now—but it is state-managed nonetheless. It marks a significant shift in who steers the Treasury market and, therefore, in how the US economy is financed.

🇨🇳 Simultaneously, Beijing is advancing its own version of the Great Reset through a continued strategy of “commercial embeddedness.” On 10 August 2026, Deutsche Bank was the first European bank appointed as RMB Clearing Bank for Europe, enabling it to handle Chinese yuan payments directly on the continent. This opens up new payment routes that allow businesses and national economies to bypass the US dollar and Western networks such as SWIFT. Beijing continues to focus on making its payment systems too convenient to ignore and making the yuan indispensable, even if it has not yet won the world’s full trust.

Foreign investors still worry about China’s strict government controls. Beijing’s capacity to tightly manage domestic liquidity was on display in mid-August, when the People’s Bank of China conducted an unusual ¥349 billion overnight reverse-repo operation to ease mid-month funding pressures in the banking system. The move supported a bond rally, with the 10-year Chinese government-bond yield falling by about 1 basis point to 1.68%, its lowest level since July 2025. That is quite a contrast with the US.

🇺🇸 Finally, something else is happening in America. Within days, Washington had signalled a broader ambition: to bring the infrastructure of the AI and digital-asset economy within US market rules—and, with it, to turn intelligence itself into a commodity. On August 19th, the CFTC, the regulatory authority that oversees commodity markets, opened a request for comment on derivatives linked to computing power—treating “compute” as a prospective commodity market—and Chairman Michael Selig described compute as perhaps the most important commodity of the intelligence economy. Meanwhile, the SEC has proposed a dedicated “Regulation Crypto Assets” framework for certain investment contracts involving crypto assets, while the Treasury made a proposal to implement the GENIUS Act’s licensing and cross-border rules for payment stablecoins. The emerging US policy direction, beyond the general ‘pro-crypto’ stance, is clearly an effort to onshore and regulate the market infrastructure of the next financial and technological cycle.

These signals and developments are, in our view, among the defining macro stories of 2026. The 1970s produced a Great Financial Reset that redrew the map of global finance—from New York to the Caribbean, London, and Hong Kong. Today, the actions of governments and markets from Washington and Frankfurt to Abu Dhabi and Beijing suggest that another reset is underway—one that will produce a markedly different financial system on the other side.

🇺🇸 America’s debt and the mechanics of soft yield curve control

We all know the drill: in Western economies, central banks have the authority to set short-term interest rates, that is, the rate at which banking establishments can borrow from the central bank itself; long-term interest rates, on the other hand, are set by the market. Those longer yields reflect investors’ assessment of the risk of lending to a sovereign over time.

The ‘spread’ is the difference between long-term interest rates and short-term interest rates. It is not only a vote of confidence in the central bank but also a vote of confidence in the fiscal authorities and the government. If the spread is high, it clearly signals either a lack of confidence in the fiscal trajectory or expectations of higher inflation, or both.

Over the past 18 months, the Trump administration has repeatedly pressed the Federal Reserve to cut interest rates. In their view, lower policy rates can reduce borrowing costs, encourage households and companies to spend and invest rather than hold cash or short-dated fixed-income assets, support asset prices, and lower the government’s near-term debt-servicing costs.

Trump reiterated that preference recently in widely discussed remarks concerning Switzerland. Meanwhile, Kevin Warsh, the recently appointed Federal Reserve chair, has repeatedly indicated an openness to a more accommodative monetary-policy stance.

The standard objection has been voiced all along: the Federal Reserve controls short-term interest rates; it does not directly control the long end of the Treasury curve. If investors doubt the credibility of the US’s fiscal trajectory—because they expect persistently large deficits, rising debt issuance, or an inflationary policy mix—then long-term yields may remain elevated even as the Fed cuts its policy rate. The result would then be a wider spread between short- and long-term rates: easier money at the front end, but persistently expensive financing for mortgages, corporate investment, and the government at the long end.

In other words, rate cuts can support cyclical growth, but they cannot by themselves restore market confidence in public finances. If fiscal credibility deteriorates or the inflation outlook worsens, the bond market may effectively offset the Fed’s easing by demanding a higher term premium.

And well, here we are: while the Fed most recently decided to hold short-term interest rates despite facing above-target inflation in the context of tariffs, multiple shortages, and the Iran war, long-term interest rates reached a cycle high of 5.34% on 19 August—a level not reached since before the Great Financial Crisis in 2008. Clearly the rise in long-term rates points to growing investor concern about the US’s fiscal trajectory and expectations of higher inflation. In other words, investors are demanding greater compensation for the risks of holding long-term US government debt: the risk of default, the risk of a depreciation in the value of their claims through inflation, or both.

This summer rise in long-term yields appears to have prompted growing concern within the Trump administration. The Treasury intervened twice to contain it:

• As we discussed in The Dollar’s Japanese Problem, the US Treasury stepped in to support the Japanese yen and prevent a chain reaction in global bond markets. A sharp fall in the yen could have forced Japan to sell large amounts of US Treasuries to defend its currency, pushing US interest rates even higher. Rather than selling reserves, Treasury Secretary Scott Bessent appears to have relied mainly on public statements to calm the market.

• The US Treasury then announced it would increase buybacks of long-term government bonds as the 30-year yield hit record high levels. While Treasury buybacks are a regular part of America’s public debt management, increasing their size provided additional liquidity to a market where investors were becoming more reluctant to hold older, less liquid Treasuries. This temporarily helped support bond prices and limit further increases in long-term yields.

What does this tell us? As the global financial system undergoes a massive transition, America seems to be quietly turning to a strategy known as financial repression: using the financial system to keep borrowing costs under control and limit the movement of capital. Given the turbulent reset, governments face two adverse reactions from markets: (i) bondholders demand higher yields to compensate for the increased risk, while (ii) investors try to move their wealth out of the country altogether, creating capital flight. To maintain stability while the monetary and financial system is being upgraded, governments need to contain both—and intervene:

• Capital flight can be prevented by restricting the ability of investors to move money across borders, using regulations, capital controls, and other measures that keep domestic savings within the national financial system. The US is not currently imposing capital controls in the traditional sense, but the Trump administration has begun to restrict the movement of capital in more targeted ways, particularly by limiting US investment in China and restricting Chinese access to sensitive US assets. It has also signalled that it wants to expand these restrictions further. In other words, Washington is not yet trying to trap capital inside the US, but it is increasingly seeking to control where US capital can go and who can own US assets.

• Yield curve control, meanwhile, is implemented through direct intervention in the bond market, with the government or central bank stepping in to buy long-term bonds and keep yields from rising too far. That is broadly what we are now seeing from the US Treasury (see above).

There is a direct and important connection between the two. By forcing yields down, yield curve control makes government debt less attractive to investors, which can increase the incentive to move capital elsewhere. The more governments suppress yields, the greater the need for other forms of financial repression to prevent capital flight. Over time, the two measures reinforce each other: controlling the bond market makes capital controls more important, while controlling the movement of capital makes it easier for governments to keep bond yields down.

It’s too early to tell where this is going. On one hand, if the Trump administration is serious about reining in the deficit, it will have to confront powerful constituencies. A major contribution to the US skyrocketing deficit, indeed, is what economists call the “voracity effect.” The idea is that when borrowing is unusually easy, every constituency, whether domestic or foreign, has an incentive to capture some of that borrowing capacity for itself, either through more spending or lower taxes.

And so, if the US government can borrow at what looks like an attractive rate, why not borrow a little more? That logic, repeated across both domestic and foreign constituencies, can turn cheap borrowing into a self-reinforcing process. Defense contractors, the healthcare industry, wealthy taxpayers, America’s allies, and increasingly even Bitcoin holders demanding a Bitcoin strategic reserve all have an interest in securing a larger share of the resources made available by the US government’s ability to borrow. What starts as extraordinary borrowing power therefore becomes a political resource, which in turn creates yet more incentives to borrow.

This is part of the deeper problem facing the US. Global demand for dollars and Treasuries has given Washington an extraordinary capacity to run deficits while allowing the US to attract the foreign capital that funds innovation and sustains its financial system (an “exorbitant privilege”). But the same system also makes it easier to postpone difficult decisions about spending, taxation, income distribution, and the country’s industrial base. As long as demand for US debt remains strong, the political pressure to restrain borrowing remains weak (an “exorbitant burden”).

That paradigm, however, may be changing. The rise in long-term Treasury yields suggests that the demand for US debt may be becoming less elastic, even as the cost of servicing the existing debt continues to rise. The Treasury can try and intervene to improve market liquidity and contain yields in the short term, but those measures do not address the underlying fiscal problem. If the administration wants to change that trajectory, it will eventually have to confront the constituencies that have benefited from the borrowing capacity of the US government.

There’s even a political rationale for going in that direction. Recently resurfaced videos of Vice President JD Vance from 2023—before he took his current office—revived the critical argument that the dollar’s global reserve currency status acts as a “resource curse.” This constant global demand keeps the US dollar artificially strong, which systematically hollows out American manufacturing and hurts American workers by preventing the domestic economy from rebalancing.

Since Vance is arguably a contender for the Republican nomination in 2028, this perspective suggests the Republican Party could develop an intriguing political synthesis in the isolationist, fiscally conservative tradition of Robert A. Taft (another prominent Ohioan). Vance’s modern Taft-like framework would advocate for both isolationism (arguing we have had enough of the rest of the world using our currency as a reserve) and strict fiscal conservatism: by voluntarily renouncing the “exorbitant privilege” of the global dollar, the US would no longer have an unrestrained, foreign-funded debt machine to rely on, effectively forcing Washington to rein in its massive fiscal deficit once and for all.

💵 Stablecoins as systemic debt satellites

We are far from a consensus. Vance’s argument treats the dollar’s reserve role as a constraint on fiscal discipline and industrial renewal. The administration’s stablecoin strategy, by contrast, seeks to strengthen that role by expanding global demand for dollar assets and Treasuries. Indeed, other developments from this summer suggest that the US is in fact determined to keep borrowing ever more from the rest of the world, specifically by making it easier to issue dollar-denominated stablecoins and flooding the global economy with them.

To understand this counterforce, one must look at the underlying plumbing of how a “digital dollar” actually works in the new regime instituted by last year’s GENIUS Act. Because stablecoin issuers are required to hold US Treasuries as reserves, growth in stablecoin issuance can create substantial additional demand for US government debt. This means that stablecoin issuers have quietly transitioned from speculative, crypto-native tools into systemic “debt “satellites”—privately run funding vehicles that help the US Treasury finance its mounting fiscal deficits.

Recognizing these developments, the Treasury has moved aggressively to bring these digital rails under formal state control. On August 17, 2026, the Treasury issued a Notice of Proposed Rulemaking (NPRM) to implement Section 3 of the GENIUS Act, establishing a formal licensing regime for payment stablecoins slated to take effect on January 18, 2027. Scott Bessent himself was characteristically frank about the geopolitical goal of this regulatory land grab, explicitly stating that formalizing stablecoins is designed to “cement the role of the U.S. dollar as the world’s reserve currency” and, echoing Trump’s own words, ensure America remains the undisputed “crypto capital” of the world.

This new role is well illustrated by Tether, the largest player in the industry. Its stablecoin, USDT, is backed largely by US Treasury bills and other US government assets, meaning that global demand for USDT translates into demand for US government debt. As Tether’s circulation has grown, so has the pool of Treasuries held against it. Tether is therefore not merely a crypto company issuing a digital dollar: it has become a large, privately run buyer of US government debt, effectively acting as one of those “debt satellites” of the US Treasury. Its growing integration into the regulated financial system only strengthens this role. On August 13, 2026, Tether completed a historic milestone by securing a clean, unqualified audit opinion from KPMG US on the financial statements of its primary issuing subsidiary, Tether International, S.A. de C.V., for the year ended December 31, 2025. The audit verified that Tether’s reserves exceeded its liabilities by $6.814 billion.

Yet, even as Washington attempts to use stablecoins to export dollar demand globally, Silicon Valley’s dream of entirely borderless digital money is colliding with the reality of localized sovereignty. As Airwallex’s Jack Zhang has observed, central banks around the world are refusing to let stablecoins bypass their domestic laws. Instead, financial systems are becoming increasingly localized: Brazil has absorbed virtual assets directly into its foreign exchange laws, Vietnam is setting up a regulated crypto market while continuing to prohibit crypto payments, and the Philippines has indefinitely locked down its licensing regime. In the eurozone, the ECB has warned that broad domestic use of foreign-currency-denominated digital assets, including dollar stablecoins, could cause digital currency substitution and weaken monetary-policy effectiveness. It has also stressed that stablecoin growth requires safeguards against financial-stability risks, while its preferred path is properly regulated, EU-governed, and euro-denominated settlement assets.

This global pushback is a direct defense mechanism against American financial hegemony. When the US Treasury attempts to solve its domestic debt crisis by exporting digital dollars, other nations face an existential threat: the risk of their own citizens abandoning local currencies for the safety of the digital greenback. To prevent this capital flight, foreign governments have a powerful incentive to fence in these digital assets. By forcing stablecoins to comply with strict domestic rules and localized fiat gateways, sovereign states are ensuring they maintain control over their domestic monetary borders—effectively shattering the dream of a unified, borderless digital dollar.

Furthermore, this Trump administration’s embrace of stablecoins masks a brutal, unresolved conflict:

• On one side of this fault line stands the Wall Street establishment for whom, as explained to us by Michael Pettis a few months ago, dollar primacy is a way to make money and therefore non-negotiable; in this context, if stablecoins can serve as efficient, high-tech vacuum cleaners sucking up global capital to buy Treasuries, they are a welcome addition to help fund the massive national deficit of which Bessent, a Wall Street man himself, is in charge.

• On the other side stands Silicon Valley, which, interestingly, views crypto very differently: as a borderless escape hatch from traditional finance.

Vance sits at the intersection of two incompatible projects. As vice president, he must support an administration that continues to run large deficits and depends on the Treasury’s ability to place ever more government debt—much of it with foreign investors. That aligns him, for now, with Treasury Secretary Scott Bessent, whose central task is to ensure that the US can finance its borrowing at tolerable rates.

Read more

America has a weak spot: it relies on the rest of the world to keep buying its debt. We’ve just had another example of this.

To the financial press, the recent joint US-Japan intervention to support the yen was a heartwarming show of transpacific solidarity. The US Treasury and Japan’s finance ministry had been coordinating on exchange rates for months — Finance Minister Satsuki Katayama said she and Bessent spoke around ten times, including a three-and-a-half-hour session over dinner during his May visit to Tokyo — and the effort culminated last Friday in the first joint US-Japan yen-buying intervention since 2011. A Reuters photo caught Bessent’s notepad at a Friday cabinet meeting reading “To Do: Buy Japanese Yen (JPY) $5–10 bil.” The coordinated buying, run by Japan’s currency diplomat Atsushi Mimura in close step with the Bank of Japan, helped firm the yen from around ¥162.80 to ¥157.80 to the dollar, pulling it back from a four-decade low. Some analysts even draw parallels to Washington’s support for Argentina’s peso under President Javier Milei, back in September 2025 when the country was battling currency instability ahead of key midterm elections.

But behind the scenes, this was less a philanthropic rescue mission than a cold, self-interested defensive bailout. Indeed, the mechanics of this intervention expose the fraying structure of the US dollar system. Washington now has to draw on its European allies and push its own financial institutions deeper into the management of a foreign-exchange crisis in an effort to keep its debt-recycling machine from grinding to a halt.

The imperial debt trap and the hostage crisis in the Treasury

For decades, America has benefited from a structural privilege that comes with issuing the world’s dominant reserve currency. The rest of the world wants dollars and, above all, dollar-denominated assets, allowing the US to run large external deficits while foreign savings flow back into American financial markets.

This creates a closed, circular capital flow:

• The US imports physical, tangible goods from net savers such as China, Germany and Japan.

• The US pays for these imports in dollars, which then flow back into the global financial system.

• For surplus countries, there is an unusually deep and liquid market in which to recycle those dollars: US financial assets, above all US Treasuries.

This loop lets Washington run a huge $1.8 trillion deficit and keep total public debt at 122% of GDP without facing the consequences of hyperinflation or soaring borrowing costs at home. But this great privilege carries with it an existential vulnerability: the US has become structurally dependent on foreign central banks to continuously purchase and hold its debt. This is a creditor hostage situation.

In this context, the recent plunge of the Japanese yen toward 164 against the dollar has brought the Bank of Japan to a tipping point. Tokyo faced a painful choice: accept further yen weakness and the inflation it would bring, or intervene in the currency market by finding dollars with which to buy yen. One obvious source of those dollars was Japan’s enormous stock of US Treasuries, but selling them at scale would have put further pressure on the Treasury market.

Indeed, had Japan liquidated its entire Treasury portfolio, the sudden surge of supply would have sent US yields rocketing higher. The yield on the 30-year US bond had already hit more than 5.2 percent, its highest level in nearly two decades, leaving Washington unusually sensitive to any new source of selling pressure. The US government cannot easily absorb a further sharp rise in its debt-servicing costs.

That is why Scott Bessent came along. “Excess volatility” in currency markets was undesirable, he said. In fact, the intervention did something else: it protected America’s own debt market from the consequences of a yen crisis. Japan, the world’s biggest foreign holder of US Treasuries, would otherwise have been forced to sell dollars and Treasuries to defend its currency.

This desperate stabilization effort discloses two profound things about the geopolitical hierarchy of the dollar system.

Takeaway 1: Japan remains a sovereign vassal

First, Japan is still a vassal of imperial America, in the great tradition that SCAP established in 1945.

General Douglas MacArthur, as Supreme Commander for the Allied Powers (SCAP), oversaw a far-reaching US-led occupation that demilitarized and democratized Japan and reshaped its political, social and economic institutions.

Over the following decades, this arrangement became a central feature of the Pacific financial system rather than a simple one-way bargain. Japan’s growth created large private and institutional savings pools, and Japanese investors became major purchasers of US assets—including Treasury securities—helping finance U.S. borrowing alongside domestic investors and other foreign creditors. Japan remains the largest foreign holder of Treasuries, with about $1.2 trillion recorded at the end of 2025.

And so, when the bill was due, Japan’s monetary sovereignty was completely subordinated to Washington’s domestic market needs.

In a healthy sovereign relationship, a nation in crisis has the freedom to use its own foreign reserves to defend its currency. Japan can, of course, sell its Treasury holdings. The problem for Washington was that Japan doing so at scale would have added selling pressure to the US Treasury market at a time when long-term borrowing costs are already high.

Accordingly, Treasury Secretary Scott Bessent has publicly pressured the Federal Reserve to expand its Foreign and International Monetary Authorities (FIMA) Repo Facility to paper over this restriction. From Japan’s perspective, the FIMA facility is a golden cage: it provides temporary liquidity by allowing Japan to borrow physical dollars against its Treasury holdings, without selling the underlying bonds.

In other words, the intervention amounts to a desperate backdoor bailout. By calling for the Fed to expand its balance sheet to provide a backstop for a transpacific crisis, Bessent has pushed the central bank deeper into a problem that sits at the intersection of monetary policy, foreign exchange and Treasury debt management. The boundary between the Federal Reserve’s monetary role and the Treasury’s need to protect its debt market is becoming harder to maintain.

Takeaway 2: Europe is the punching bag

Secondly, Europe is the punching bag as usual.

The most cynical, calculated part of this transpacific drama was the intervention mechanism: the US Treasury sold euros to buy yen during the intervention; and according to the Financial Times, the ECB was informed only after the transaction had been executed.

A country intervening in FX markets will usually buy or sell in its own currency. But doing it here would have added to the selling pressure on the dollar or tightened domestic US dollar liquidity. This is the reason why the US Treasury decided to sacrifice the euro instead.

Indeed, the US Treasury sold an estimated $5 billion to $10 billion of euro-denominated assets from the Exchange Stabilization Fund (ESF), which consumed roughly 40% of the US’s total euro holdings to that point ($10 billion of the $26 billion total euro assets held by both the Federal Reserve’s SOMA and the ESF).

The result was a classic case of transatlantic financial asymmetry. Washington could shore up its key Asian creditor by dumping euros, pushing the immediate cost of the intervention onto Europe’s currency rather than its own. The episode exposed how little say Europe has over the use of its currency in a monetary system still dominated by the dollar.

Japan’s other financial strategy: bringing crypto inside

Read more

Historically, financial repression — where the state dictates the movement of capital through various control mechanisms — was considered a phenomenon relevant only to developing economies. Those had weak institutions and captive savers, unlike the advanced world that had spent five decades dismantling exactly those controls.

These days, however, that distinction is going away. Financial repression is returning to the rich world, driven by necessity. What we are witnessing is a late-cycle response to a fundamental crisis: national debt levels have reached a point where they can no longer be outrun by productivity or curbed through traditional cost-cutting. And when a system-wide expansion plateaus, the state must transition from market-led allocation to a coordinated — often coercive — effort to ensure the survival of the financial superstructure.

Total debt-to-GDP now runs to roughly 255% in America, above 300% in France, and near 400% in Japan. Two levers are most often contemplated to solve the equation: austerity, or an increase in productivity. Both seem somewhat blocked. Austerity died politically in the populist backlash of the 2010s; no government that wants to survive an election will now try to cut its way out. And the productivity surge that is meant to grow nations out of their liabilities has not yet arrived, AI technological promises notwithstanding. Consequently, it’s good old inflation that has emerged as the sole remaining lever to shrink debt-to-GDP ratios. If history is any guide, it’s now all but certain that most governments around the world will try and inflate away the real value of state liabilities while nominal GDP continues to rise.

That said, inflation only works as a debt strategy under one condition: savers cannot escape it. And financial repression is precisely how economists label the apparatus that stops savers from escaping inflation: a technical matrix built to keep domestic savings from fleeing into hard assets or foreign markets. Broadly speaking, it consists in five different levers:

1. Impose negative real interest rates. Hold nominal rates below inflation: 1% paid to the saver against 4% inflation is a 3% annual levy on wealth, which liquidates a quarter of real purchasing power over a decade without a single line in any budget.

2. Make sure the money has nowhere to hide. Through modifications to Basel III, pension rules, and insurance solvency frameworks, regulators “nudge” — the polite word for compel — domestic institutions into holding low-yield sovereign bonds regardless of what a free market would price them at. Captive buyers do not demand a premium for financing the state; that is the point of making them captive.

3. Repress wages, which results in repressing domestic consumption. Germany did it to defend export competitiveness: after the Hartz reforms, output kept growing and pay stopped, and the gap surfaced as corporate profit. China did it harder and for longer, through suppressed deposit rates, a managed currency, and wage restraint, and now has the lowest household share of GDP in the world.

4. Rebrand the whole thing as innovation policy. Britain’s Mansion House Compact channels pension money into private equity; France’s Tibi initiative points domestic savings toward strategic sectors. Both are sold as forward-looking industrial ambition, but they’re really financial repression in disguise — guaranteed capital to destinations an open market might decline to fund.

5. Run it long enough that it stops looking like policy. Japan is the proof of concept. For three decades it has kept government yields far below what an open market would have demanded, by ensuring that domestic banks and pension funds absorb the debt — the central bank alone now holds roughly half the outstanding stock. This is what a successful developed-world cage looks like from the inside: a quiet, permanent arrangement in which the saver funds the state and everyone calls it prudence.

As these walls are now rising everywhere in the world. They create an existential threat to any nation dependent on the free movement of international capital.

The American Dilemma: Dependency and the Funding Shock

For decades, America has enjoyed “exorbitant privilege,” a position where its structural trade deficit is automatically funded by capital inflows from foreign holders of dollars seeking US assets. This automatic recycling of global savings into Treasury bonds and American stock has historically allowed the US to fund itself without domestic capital controls.

The arrangement worked for two reasons: America offered the world’s deepest and most sophisticated capital markets, backed by the rule of law, and foreign savings could move there freely. Remove either condition and the system ceases to function as it did. Whatever becomes of America’s relative appeal, the second condition is already beginning to fail. America remains one of the world’s most attractive destinations for investment, which is precisely why more governments are trying to stop domestic capital from flowing there.

Indeed the idea that capital should move freely is now being dismantled, wall by wall, by a growing number of nations. Every sovereign cage erected abroad shrinks the pool of voluntary buyers for US debt, because savings trapped in Frankfurt or Tokyo or Beijing to fund domestic deficits are savings that will not turn up at a Treasury auction. Curb the financial flows, and the automatic bid America has leaned on for half a century starts to thin, resulting in a slow subtraction and, ultimately, a funding shock.

Instead of waiting for market forces to voluntarily align, Washington is choosing to act proactively. This is the pivot that defines the rest of the story: from a defensive privilege that depended on others’ goodwill to an offensive instrument that manufactures the incoming flows so as to keep funding the US economy. Increasingly, that instrument runs on rails made of software.

The Offensive Shield: Software-Defined Narrow Banking

Read more

Venture capitalists like to analyze technology in terms of waves. For them, each new wave is an opportunity to reposition, find interesting companies before anyone else starts paying attention, and generate precious alpha as a result.

The current wave, for instance, is the rise of agents, enabled by everything AI now contributes to software development. Before that came others: cloud computing, mobile internet, social media, B2B SaaS, the metaverse, and many more. Some produced great companies and outsized returns, while others proved disappointing.

In the early 2000s, one massive wave helped pave the way for a Silicon Valley renaissance: open source software. The dotcom crash of 2000 had been interpreted by many as a clear signal that there was no “there” there, that the idea you could build profitable businesses on the internet was merely an illusion. But then a small vanguard of software engineers started to share software libraries and lines of code with their peers for free, with license to reuse and modify. Over time, that movement contributed to both a massive collapse in the cost of software development and a dramatic improvement in the quality of code. It eventually fueled the rise of a new generation of startups and led to a decade during which, to borrow Marc Andreessen’s words, software would eat the world. A massive success!

One reason open source became a movement is that, in 2004, Tim O’Reilly, a respected publisher and conference host, brought it into perspective by explaining why sharing code for free made economic sense. His landmark article Open Source Paradigm Shift, based on a keynote he first delivered at the annual technology conference of private equity firm Warburg Pincus, drew on historical parallels and business strategy concepts.

Above all, it explained how you could still make money from something shared for free: open sourcing a layer in any value chain commoditizes that link and displaces competition toward another layer where one or more players hold a sustainable advantage. If you have something that others don’t, your best strategic move is to open source an adjacent layer and force everyone to compete with you within the layer where you hold the upper hand.

O’Reilly was writing about software, but the logic he described is indifferent to its raw material. One could say it even applies to money in general, and stablecoins in particular. And this is exactly the reason why we have been revisiting O’Reilly’s landmark article recently. On June 30, 2026, a consortium of more than 140 companies, including Visa, Mastercard, Stripe, Coinbase, BlackRock, and BNY, announced OpenUSD: a stablecoin nobody owns, with no fees to mint or redeem, no volume caps, and the reserve yield handed back to partners rather than kept by a single issuer. In other words, an open source stablecoin.

The announcement immediately drew two reactions. First, the trade press read it as a shot at Tether and Circle, and the market agreed with unusual speed, sending Circle’s stock down more than 17% within hours. Then came the backlash: several named “partners” turned out to have signed nothing; the governance is a curated board of incumbents, not a commons anyone can join by showing up; the word ‘open’, critics argued, is doing marketing work rather than describing a structure.

Both reactions are fair, but they are also too small. Whether OpenUSD ships, stalls, or collapses under its own consortium politics is a question about one product only. What we’re interested in here is the broader perspective. Stablecoins were already open-source money of sorts: they run on permissionless, open-source blockchains — Ethereum, Solana, and their peers — that nobody owns and anybody can build on. By contrast, what had not been open until now was the layer immediately above the rails: issuance, that is, where Tether and Circle sit.

Whatever OpenUSD’s fate, it is the first serious attempt to push the openness one layer up the stack, and to make issuance open source. The question worth asking is not whether this particular consortium succeeds. It is: how far up does the openness travel — and what happens to the money underneath it, and to every currency that refuses to play the same game?

The commoditization play, twenty years on schedule

When O’Reilly wrote his essay back in 2004, the goal was merely to explain why Linux was eating the software industry from below. His core claim was that software turns into a commodity the moment it has to speak a shared protocol, whether HTTP, SMTP, or TCP/IP.

Two decades on, the article reads as a template for building a competitive advantage on top of a commoditized layer in any value chain. A commodity, by definition, is something interchangeable, sourced from more than one producer, that a customer can swap for a rival without noticing the difference. Once a layer becomes a commodity in this sense, the profits don’t disappear into thin air; instead, they typically relocate to an adjacent layer.

O’Reilly borrowed the mechanism of that relocation from the late Clayton Christensen, then a professor at Harvard University and the famed author of the book The Innovator’s Dilemma. One of Christensen’s well-known theoretical concepts, which he labeled the “law of conservation of attractive profits”, holds that when a product becomes modular and commoditized at one stage of a value chain, attractive profits reappear at an adjacent stage.

This happened several times in the history of computing: IBM commoditized the personal computer in 1981 by opening its architecture to clone manufacturers, then lost control of the industry to Microsoft, which owned the layer just above the hardware. Dell later commoditized the hardware itself and won the market by embracing thin margins rather than fighting them. And of course, open source software later did the same to proprietary software. By commoditizing the act of writing code, it shifted competition to the adjacent, user-facing layer, where firms compete on user experience and network effects.

Stablecoin issuance today sits where proprietary software sat in 2003: it is closed, and it is concentrated, with Tether and Circle holding roughly 83% of the market between them. Their business model is disarmingly simple: take deposits, buy (mostly) short-dated Treasuries, keep the yield — whether to fund other ventures as Tether does, or to redistribute it to partners as a marketing fee as Circle does. Operating the issuance layer is lucrative precisely because it has not been commoditized: USDT and USDC are not perfectly interchangeable, and switching costs in integrations, liquidity, and compliance let both issuers hold on to reserve income that a true protocol-level market would compete away.

That reserve margin is exactly what OpenUSD is aiming at. The consortium doesn’t even need to be truly open — a genuine commons where anyone can join and shape the rules — to do the damage. It only has to open the standard wide enough to break the current Tether–Circle duopoly. A fixed set of members who then compete against each other inside walls they built together would be enough to turn issuance into a commodity.

Of course, OpenUSD is, for now, only an announcement. Still, Circle’s stock fell the instant the market grasped which way issuance was heading — toward the losing side of Christensen’s law, where a service becomes a commodity and its margin collapses. Nothing about how issuance works has changed, and nothing has yet shipped; the market is simply pricing in what issuance might become. Watching investors reprice Circle in a single afternoon, on the strength of a press release, was watching the shift O’Reilly described twenty years ago arrive exactly on cue.

How far up does the openness go?

Again, the stablecoin rails were open from birth. Permissionless blockchains are open-source software in the strictest sense: public codebases, forkable, maintained by distributed communities, owned by no one. The monetary instruments that run on them, however, are the opposite of open source. Tether and Circle built proprietary stablecoin businesses on a commons — the oldest move in the open-source playbook, and a profitable one for as long as nobody opens the layer you occupy.

OpenUSD’s value proposition is to open that layer too. Even if it fails, the idea will not die with it; the economics that produced it — 83% concentration, reserve yield held by two firms, switching costs as the only moat — will produce another attempt with a different logo. Then there will be attempts to commoditize even beyond issuance.

The agent frameworks now being built to let software transact autonomously are, for the most part, already open source. Settlement and compliance infrastructure could follow the path of web servers. Wallets and distribution, in turn, could be commoditized in much the same way that web browsers were. In the end, if these software-dominated layers are all commoditized through open source, Christensen’s law suggests that profits will migrate to whichever adjacent layer still depends on scarce resources. That would confine scarcity to the physical world: robotics, logistics, energy, and the machines that carry out what agents decide and stablecoins settle. As stablecoins become open infrastructure, layer by layer, they will become embedded in those physical systems, serving as the default settlement rails for real world commerce while competition and value creation shift to the assets and operations that cannot be commoditized.

This leads to the final question. If the rails, issuance, wallets, settlement, compliance, and agent frameworks all succumb to commoditization through open source, what becomes of the currency itself? Is the dollar simply the next layer in the stack, or is it something fundamentally different?

The dollar is the protocol beneath the stack

Read more

We have come to Tether from opposite ends.

Marieke spent years at Circle, Tether’s rival, and carried the industry’s scepticism about it almost as a professional identity: the opaque reserves, the Bitfinex entanglement, the attestations that fell short of audits. Later, running the NEAR Foundation, she became a Tether customer and worked to bring USDT to the NEAR protocol. She now advises startups and non-profits that Tether funds. Finally, in October 2025 she went to the Plan B conference in Lugano and left convinced that something large was under way, something most of finance was missing.

As for Nicolas, he works in venture capital and is now watching Tether burst into the space, leading large rounds in companies such as Neura Robotics, with no obvious link to stablecoins1.

We have reached the same conclusion by these different routes. Tether is one of the most consequential financial institutions of the decade, but mainstream finance still does not treat it as one. What follows is our case for taking it that seriously.

The Machine That Prints Its Own Float

Tether issues USDT, a digital dollar. Customers hand over conventional dollars to get it, or convert other assets into dollars to get it. Tether puts those dollars into short-dated US Treasury bills, held through Cantor Fitzgerald, the Wall Street firm that buys and holds the bills for it, and keeps the yield. None of that yield reaches USDT holders, and none ever has.

For most of Tether’s life, when its holdings were tiny and interest rates were low, this earned little. But at today’s rates, on close to $190 billion of tokens outstanding, the arithmetic changes. Tether reported more than $13 billion in profit for 2024 and around $10 billion for 2025, on a staff of roughly three hundred. No branches, no loan book, no credit risk worth the name. Some call it a hedge fund that happens to issue a stablecoin.

As for Circle, Tether’s most direct competitor, it earns the same spread between a zero-yield token and interest-bearing reserves. But the two firms differ in what they do with the money. Circle spends it on credentials: compliance, institutional relationships, the standing to serve as regulated infrastructure. As a result, it is listed, MiCA-compliant for both USDC and EURC, and holds its reserves in BlackRock money market funds. It also hands a large share of its yield to distribution partners such as Coinbase, a cost Tether does not carry.

Beyond its core product, USDC, Circle has now actively started to build the entire stack required for money to move: from proprietary purpose-made Layer 1 blockchain ARC, to Circle Payment Network and more recently Circle Nanopayment, a gateway for stablecoin payments.

In short, Circle goes deep in building the stack that the future of money requires. Meanwhile, Tether seems to spend its money on everything else.

Tether’s (Bleak) Worldwiew

Read more

Jess Hoversen grew up in Chicago. At ten, walking past the Chicago Board of Trade on LaSalle Street, she decided she would work there one day. At twenty she did, starting as an intern at the exchange and then trading currencies as the pits gave way to the screens, mostly in financial futures and options. She moved to MF Global as an FX and bond strategist, and watched the firm’s European debt position bring it down in 2011. Wanting to serve in government, she joined the CIA, and spent thirteen years as an economist across the Agency and the Departments of State and Treasury, latterly as a Treasury and Office of Foreign Assets Control (OFAC) attaché in Bogotá. She is now chief economist at Column, a nationally chartered bank and technology company in San Francisco, where she writes the newsletter Hegemoney.

Jess is one of the few people to have seen dollar power from both sides of the desk, as a trader who priced the currency and as a national security economist who helped wield it. On the floor she learned that currencies move on interest rate and growth differentials. At MF Global she saw a third force. When a colleague called the euro Europe’s answer to centuries of war, she realized a currency is also an instrument of statecraft. Her work centres on what sustains that leverage for the dollar. Most commentary asks whether the dollar is losing its reserve status; she thinks that is the wrong measure. The better question, she argues, is where the dollar is dominant rather than why, and whether anyone in Washington is still working to keep it there.

In this conversation, Jess explains why the fixation on reserve status misreads the dollar, and why its use as a vehicle currency, the rail two countries reach for when neither wants the other’s money, is the better gauge. She reads the surge in central bank gold as a sign of a new geoeconomic paradigm rather than a transactional rival, and argues that payment infrastructure now sits alongside economic size and deep markets as a pillar of dollar power. She dates the turn to alternatives to 2014 and the response to Crimea, draws the parallel with the eurodollar market of the 1960s, where even eurodollars had to touch the US system whilst many stablecoins now may not, and reads China’s aim as survival and alternatives, with Panda and Dim Sum bonds as the signposts to watch. She warns the United States to steel-man both sides of the trade-imbalance debate before repeating Europe’s mistake on austerity. Her case throughout is that the dollar erodes slowly, one per cent at a time, and that Washington has been too complacent to contest it.

Jess writes Hegemoney and is on Substack, and on X at @JessicaHoversen.

Marieke: Hi, I’m Marieke Flament.

Nicolas: And I’m Nicolas Colin.

Marieke Flament: And this is the Currency of Power podcast, the companion to the Currency of Power newsletter, where we dig into currencies, power, and the people connecting the two.

Today’s guest is someone uniquely positioned to discuss the power the dollar holds and what could realistically alter it. Jess Hoversen is the chief economist at Column, a nationally chartered bank and technology company in San Francisco. Jess writes analysis on geoeconomics and the dollar at Hegemoney, a Substack newsletter we’ve been really enjoying. For 13 years, Jess served as a senior economist at the Departments of Treasury and State and at the CIA. She’s also a former FX strategist and derivatives trader. Jess has years of experience in sanctions, international currencies, financial infrastructure, SWIFT, international payment systems, tech, and the macro conditions around the dollar. The perfect guest for Currency of Power. Jess, great to have you with us.

Jess: Thank you both so much for having me. It’s a real pleasure to be here, and I really appreciate the opportunity to talk with you today.

“Currencies are not just a tool of commerce, they are a tool of leverage”

Marieke: Thanks, Jess, for joining. We’re really looking forward to the conversation. Today’s episode should run in three parts. First, we’ll go into your background and your story, which is always interesting to hear. Then we’ll go into why dollar dominance is actually more durable than most people think, and the mechanics of the power that comes with the dollar, which you write a lot about. And then we’ll talk about the pressure points, the things that could challenge that power, and the role that tokenization and stablecoins play in the whole landscape. So the main theme is what actually sustains dollar power and what could realistically alter it. Let’s get started.

Jess, fascinating background: Treasury, State, CIA, lots of different things. Now you write on geoeconomics, you author a newsletter, and you work at Column. Tell us about you.

Jess: All right, let’s go. When I was 10 years old, my mother and I were walking across the street on LaSalle in Chicago, where I’m from. I looked at this beautiful building and I said to my mom, “I’m going to work there one day.” It was the Chicago Board of Trade (CBOT). She said, “Wait, what? You said you were going to be a ballerina.” She thought I wasn’t serious about anything, but I really was. I’ve always been fascinated by money and economics, and I knew really young that this was something I was going to do. So when I turned 20, I started an internship at the CBOT and I never looked back. I fell in love with markets and currencies, and I knew I was going to spend my professional life learning as much as I possibly could about them.

I started off as an FX trader. We did a lot of trading on the screens, because I was in the markets at a time when they were transitioning from pit trading to screen trading. It was still an exciting time to be a derivatives trader. The agricultural pits were still full. The majority of derivatives trading happened in those pits, and it was so exciting to hear the roar of the crowd, literally, when trades would go. I did a lot of financial market options and futures, a lot of FX futures and options, and a little bit of commodities, but mainly financials.

After about four years, I realized I needed a deeper understanding of how markets work to be a better trader. So I left the floor, sadly. I had a fun jacket and everything, just like the movies. My badge number was HOVE, H-O-V-E, which I loved, because all traders have a little acronym for their name so you could write down who you traded with, and mine was HOVE. I still have the jacket. They don’t let you keep the badge, unfortunately. Then I started a job with the now defunct MF Global. I think we were the only US firm to collapse in the European debt crisis, so that is a real badge of something. I don’t know if it’s a badge of honor, but it’s a badge of something.

Through that position I learned a lot more about market mechanics. What drives currencies? What drives financial markets? What are the factors? Currency markets tend to have two main drivers: interest rate differentials and growth differentials. But the deeper I got in my career, the more I realized there’s something else, and that something else was national security and geoeconomics. So, having been on the floor, having studied markets and worked as an FX strategist and a bond strategist, I needed something else. I also knew I wanted to spend time in government service. I really thought I wanted to join the Air Force, but I’m afraid of flying, and that’s a prerequisite. So I didn’t, but I wanted to serve my country. How do I do this? I put in an application and joined the Agency.

That was a fascinating part of my life. I spent 13 years total in government, moving around between different organizations, and that really fulfilled that third bucket of what drives currencies. Because if it’s not growth or interest rates, it’s something else, and that something else is geoeconomics. I spent a lot of time trying to understand the nexus between national security, capital markets, and the US dollar. I really capped off that experience by working as the Treasury and OFAC attaché down in Bogotá, Colombia, because that felt like the tactical implementation of dollar power.

When we think about why the dollar is a powerful tool of coercion, it’s because we can block access to our financial markets. Working for OFAC gave me that tactical experience: how do we run an investigation, what do we target, how do we deeply understand the executive orders that give OFAC the permissions to target certain entities and individuals, and how do we understand the evolution of sanctions. I want to say it was around 2014 or 2015, Jack Lew wrote a really interesting piece in Foreign Affairs. I think it was one of the first times a US policymaker started talking about the potential for over-weaponization. That was also right around the time of the BNP Paribas penalty, $9 billion. I think it was 8.9, but everyone likes to round.

Nicolas: Jack Lew was Secretary of the Treasury back then.

Jess: Yes, a sitting Treasury Secretary writing about the potential for over-weaponization. Watching that evolution and working at OFAC really helped me understand that concept much more: how we use sanctions, why we use them, and their impacts.

Then I joined Column. After 13 years in national security, I actually thought I was going to retire at the Agency. I thought I’d stay there the rest of my life, walk out with a small celebration of donuts and coffee, and go off into the sunset somewhere in West Virginia.

Then, through Happenstance and mutual friends, I met William Hockey, the founder of Column and also the founder of Plaid. I had never heard the private sector’s national security mission articulated in such a powerful way. I thought, Column is the epicentre of national security in the financial system. William really understood that there are three major theatres of economic competition: technology, energy, and finance. Column’s entire mission is to build a better bank, and better banks are what will put us at the forefront of that financial competition mission. After just one meeting, I knew I had to do this. It felt like that extension, because when you spend so much time in national security, you still crave mission. Column’s mission was so strong that I knew this was the next step, and it felt right. So here I am, writing Hegemoney.

Marieke: Tell us the link between the two. Help us understand the link between the newsletter and the mission of Column.

Jess: Column started as an infrastructure bank, designed to empower fintechs to build on top of the US financial system. William is a strong patriot, as are the employees of Column, and the mission evolved from “How do we enable fintechs” to “How do we enable the dollar.” How do we build this bank so we can be the tip of the spear in financial competition and national security?

Our core business, the bank, is building that faster, better bank. I think 90% of all our payments are processed in under 24 hours. You can’t be competitive without efficient, fantastic technology. That’s the first piece: the infrastructure for the dollar.

The second is access. We’re a correspondent bank. Despite the many attempts to move away from correspondent banking, it’s still like gravity. I like to say the payment law of gravity is that all dollars have to be cleared on US-regulated infrastructure, and despite the changes, correspondent banking has been the mechanism that fulfils that law of gravity. Column is actively creating that access point.

The third piece is research and outreach, and that’s where Hegemoney comes in. William recognized there are a lot of ways to talk about the dollar, as you both know, since you write about it so prolifically. I love your newsletter, which is another reason I’m excited to be here. But it’s another thing to frame it for a technology audience, for policymakers, which I have a ton of experience doing, and for the financial community.

For one example, you often hear that the dollar’s reserve currency status is going away. It’s thrown around so carelessly and flippantly. But if we look at the roles of the dollar, there are multiple avenues. How do we actually diagnose dollar dominance? After many conversations with William and Sean, our editor, we sat down and said we need to create a newsletter that talks about the dollar the way we think it should be talked about. So Hegemoney is the third point of the trident in Column’s national security strategy.

Nicolas: I have two small questions to make sure our conversation is accessible to a broad audience. Can you explain more about correspondent banking? It’s a very US-specific dimension that no one in Europe really understands. We clear payments here within the European Union very fast and without all those complications. So I know correspondent banking, but can you explain what Column provides in helping with that?

Jess: That’s a great point, and I often forget that Europe has a highly integrated system. I guess we like tradition, what can I say. We inherited it from the Medicis, so we’ve got to hold on to it.

Correspondent banking is essentially when a US bank provides a dollar account for a foreign bank so they can clear payments through us. We act as a gateway to the US payment system, whether it’s ACH or Fedwire. For foreign banks to clear dollars, they need access to the Fed payment system, and the Fed does not hand that out prolifically. It’s usually US financial institutions, or banks like Deutsche that have access through their US branches or subsidiaries. Generally it’s a US firm that actually has access to what we call the Fed master account. Foreign banks don’t have that, so they need an account at a US institution.

It has been in decline. Since 2010 or so, the BIS data shows correspondent accounts have shrunk by 20 or 30%. We’ve seen alternatives pop up to fill the gap, but at the end of the day the payment law of gravity when it comes to settling dollars is that you must clear through a US-regulated financial system.

Nicolas: My other question is at a higher level. Can you define geoeconomics for us? Everyone’s vaguely familiar with geopolitics, and over the past month everyone’s starting to get interested in this new vertical, geoeconomics. Gillian Tett is writing a lot in the FT about it. How do you define it?

Jess: I think it’s the intersection between economics and politics. That sounds cliché, but if you think about almost every war ever fought, most of them are a resource struggle. When we think about competition, growth, and even collaboration and multilateralism, it still comes down to how we create prosperity, and prosperity is a function of economics. As the world shifts into a new paradigm, whatever we call it, I look forward to whoever finally names it, because it’s no longer Bretton Woods. It’s not post-Bretton Woods or the Washington Consensus. It’s a new world order. Mark Carney tried “principled & pragmatic” policy, P3, which I love. Why are we not using that? It was fantastic. But we’re not. In this new world order, relationships appear to be more transactional, and we’re no longer assuming everyone is acting in good faith.

Marieke: Fascinating. Let’s go into the depths of the topic, since you’ve already started on it. Dollar dominance. I don’t know if you can share more, but when did you actually realize, hands-on, that the dollar was a system of leverage? You alluded to it a little. I’d love it if you could walk us through that moment when you realized it’s way more than just a currency.

Jess: I love that question. When I was a derivatives trader and even an FX strategist, I thought this was a market function. Currencies tie the world together. It’s how we do trade, it’s what links us, it’s the flow of funds. But it was actually when I was at MF Global that I realized currencies are a political tool. MF Global had, it’s not a secret, a very large European debt position, which is what ended up taking the firm down. Those bonds paid out completely; they never defaulted, they were never restructured.

John Corzine was the former CEO, and we were talking about the position. I’d been pretty neutral on the euro, not bearish, only because despite the economic struggles, the ECB was not engaging in the same level of quantitative easing as the Fed. From a balance sheet perspective, it was difficult to see how the euro weakened substantially. John said something: “The euro was Europe’s answer to centuries of war.” And I thought, you’re right, absolutely. That was it for me. The euro created a binding constraint, because why would you go to war with another country that shared your currency, whose fate directly impacted the economic fate of your own country?

It was really that moment, around 2011, which tracks. I realized I was missing this entire element of understanding currency markets. Currencies are not just a tool for commerce and finance, they are a tool of leverage.

Nicolas: I’m curious. I know we’re focusing on the dollar today, but you had that direct experience of the euro and trading on European markets, confronted with a crisis in 2011 that almost brought down the euro. How do you see the euro playing out? What are the tensions within the eurozone? What’s the role of Germany? Are we making the most of it, or do we need to find something else to make the eurozone work? I know those are big questions.

Jess: Let’s talk about the euro. I’m going to say something that I feel like everybody says, but it’s important: without debt mutualization, the euro can’t advance. You can’t have monetary unity without fiscal unity. The euro has prevented Europe from going back to war. It has brought everyone to the table. But I remember during the Greek financial crisis thinking that Greece essentially had a pegged currency. It was a floating currency, yes, but when your economics don’t match those of the strongest player in the union, it’s really difficult. There’s no adjustment mechanism, and they lack the fiscal adjustment. So that should be the priority for the euro, debt mutualization, and I’m one of a million people saying this.

Beyond that, it will be interesting to see to what degree Europe pursues a euro stablecoin. What I worry about with CBDCs and stablecoins is that we create a bifurcated infrastructure, where there’s one payment system for CBDCs and one for stablecoins. Because CBDCs are reserves, they will always be valued at one. Stablecoins can go under one. When you’re settling trillions of dollars, that matters a lot. That’s going to be difficult for interoperability. They’re effectively two separate currencies, so I worry about a bifurcation.

With the United States, we’re banned from pursuing CBDCs until 2030. The executive order is very clear. We’re full throttle on stablecoins. On CBDCs, we’re not advancing that until 2030, and I believe that’s both retail and wholesale, from what the EO seems to suggest. Europe has the opportunity not to be locked out of any payment system if it pursues both options, because it’s moving quickly on a digital euro. If it also pursues stablecoins, that would preserve its access to other types of settlement and payment systems.

“My local dive bar does not let me buy beer in gold dust”

Marieke: Let’s go back to the dollar. You write a lot about the power of the dollar and the misconceptions people have, for example that it’s all about reserve currency status. Walk us through the main misunderstanding people have about the power of the dollar.

Jess: I think the main misconception I read on FinTwit, and I hate that I’m this online, I could not have imagined in my 40s that I’d be so online and getting all these memes, but here we are. When I look at how the dollar is perceived, there’s a real misguided focus on reserve status.

Why was reserve status important in 1945? Because everything was pegged to the dollar and then to gold. The system required it. Even after the fall of Bretton Woods and the collapse of the institution in 1973, we went to a series of pegged exchange rates, which made reserves even more important.

Then came the Asian financial crisis in 1998. In the period from 1973 to 1998, countries needed reserves because that’s how you intervened in your currency and how you imported goods. Every development economist knows, and you probably learned this in Macro 101, that a country needs three months of reserves to be considered stable. That’s an indicator built into so many models.

But after 1998 and the collapse of the pegged exchange rates, I’m not saying reserves aren’t important, but they’re not as necessary. We are not protecting pegged exchange rates as much as we did before 1998. So what do reserves actually signify now? There was the surge of surpluses, the savings glut from around 2005, which Ben Bernanke talked about, and Brad Setser and Maurice Obstfeld have written on this at length. Because of globalization, the surge in global trade, and the interconnectedness of commerce, we did see a rise in reserves. So what do reserves reflect in that moment? They reflect the currencies that countries were using to engage in trade. That’s an important shift: reserves became an indicator of the currency being used in everyday trade.

Let’s talk about gold. If you’ve opened an article in the last five years, you’ve seen the surge in gold holdings. Now it’s about 25% of total reserves. When we look at reserves, other assets, SDRs, gold, and foreign exchange, gold is about 25%. That’s pretty significant. Ten or fifteen years ago it was only 10%, so that’s a huge rise. But why does it not matter as much? Because it doesn’t suggest the dollar is being used significantly less as a vehicle currency. What it reflects is the movement into a different era. We’re in a different paradigm, a geoeconomic-driven paradigm, and countries are protecting themselves from potential weaponization by holding gold. But as I wrote recently, my local dive bar does not let me buy beer in gold dust. Maybe they’d figure it out, I can’t say they’d say no, but gold is not a transaction currency. Not yet. There’s talk of tokenization, but you can’t use gold transactionally at a large, meaningful scale.

So why do I bring all this up? When we think about dollar dominance, we have to look at it holistically: reserves, vehicle currency, and funding currency. That’s how we diagnose the dollar’s status. Vehicle currency, to me, is the most transactional and useful measure of dollar dominance, because it shows that countries need the dollar to engage with one another. If I want to import avocados, a significant portion of US avocados come from Mexico, and that trade is done in US dollars. It shows that in order to get the things I need, I have to use this currency. That’s why I think vehicle currency status is more important than reserves these days. Reserves tell you what countries are using for their trade; vehicle currency is what they’re actually using.

What’s also fascinating is that when I sat on the FX desk at MF Global, you could see that the spread between exotic currencies was so wide that it always made more sense to go through the dollar. FX markets have changed now: they’ve become more electronic and digitalized, and it’s become easier to access pools of liquidity, so those spreads have shrunk. But generally, if I wanted to trade Brazilian real to Indian rupee, it just makes more sense to go through the US dollar than to try to make a market between those two. So if you think of vehicle currency as a two-pronged spear, the first prong is the currency you and I use to transact, and the second is the currency I use to drive from one currency to the next. Those are powerful and sticky.

Finally, on funding currency, our markets are enormous compared to other financial markets. They’re not comparable. Especially without mutualization in Europe, an Italian sovereign bond is very different from a German, Belgian, or Spanish one. The funding currency angle is really interesting because it shows people are willing to fund their businesses and financing needs in US dollars. So those are the three tools we need when we look at not why the dollar is dominant, but where the dollar is dominant. Because why the dollar is dominant is a very different story.

“We need an offensive dollar strategy”

Nicolas: What about sanctions? You know that space well, having been a practitioner in the US government.

Let me make a quick detour first. I used to say that when Trump was re-elected in 2024 and made it clear he intended to impose tariffs and wage trade wars on various partners, including allies, it forced us to dust off our macroeconomics textbooks. I learned macroeconomics at Sciences Po 25 years ago and never used it professionally until last year, when suddenly everyone was talking macroeconomics. When you revisit macroeconomics, you also revisit the events that led us to this paradigm shift you’re alluding to.

A key turning point seems to be 2014, when Russia invaded Crimea and the US responded with sanctions. Those sanctions were traumatic from a Chinese perspective, because the Chinese realized, if they do this to Russia today, maybe they can do it to us tomorrow. So they’d better work on making their economy, currency, reserves, and banking system immune to US sanctions.

How do you balance the two? Countries like China want to become more immune to US sanctions, but still everyone wants to use dollars for everyday trading because it’s so efficient and convenient. Paul Krugman made the comparison, quoting someone else [Charles Kindleberger], that when a French speaker and a Spanish speaker who don’t know each other’s language meet, they discuss in English. That’s the lingua franca that makes it possible for everyone to interact, and the dollar is the same.

So how do you balance the two? Are there tensions at the moment? Are there counterparties that refrain from using the dollar because there’s macro or systemic pressure from government authorities who want to avoid exposure to sanctions?

Jess: I love this question, and I want to circle back to the first thing you said, because 2014 is such an underappreciated date.

Read more

On 2 June, Alphabet, Google’s parent company, sold new stock for the first time since 2005. It announced an $84.75 billion raise, the largest in history and more than twenty times the roughly $4 billion it took in back then. Warren Buffett’s Berkshire Hathaway took $10 billion of it at a discount.

Alphabet had cash, just not enough of it. It held roughly $127 billion at the end of March, less than it now plans to spend in a single year. When the build outruns the balance sheet, even a company this rich has to raise the rest in the market. That build, of course, is AI.

Only a few days later, SpaceX went public in the largest IPO in history, valued at $1.75 trillion. OpenAI and Anthropic are preparing to follow. Washington and the firms’ own investor materials cast all three as national infrastructure, foundational capability the United States must dominate whatever the cost. The framing lifts them above ordinary products and into the language of strategic necessity.

Those events sit awkwardly inside the usual story about the US market, which says American stocks rise because American firms are the best in the world and that the market exists to fund them.

None of that seems to hold anymore. The US stock market now works less as a place where companies raise capital on their results and more as the world’s destination for spare money. Global capital comes here because nowhere else is big enough to hold it. Prices follow the volume and direction of those flows more than the earnings beneath them, pushed by the index funds and automated machinery that spread money across the market without judging value.

The newer development sits on the debt side. Foreign money has long helped fund the federal government, so that link is old. What is new is the machinery now taking shape to deepen it: dollar stablecoins, whose design turns every token issued into fresh demand for US government debt.

To see what moves the market, then, look past the earnings picture to the monetary one: money supply, capital flows, and the government’s borrowing needs.

The world’s surplus savings have nowhere to go but America

Let’s begin with where the money comes from. Michael Pettis, our most recent podcast guest, has argued for years that capital flows drive trade flows rather than the reverse. Surplus economies such as China, Germany and Japan hold down household consumption to support their industries, and the savings they cannot spend at home have to go somewhere. As Michael puts it,

If I am able to control my external account... then whether you like it or not, your economy will adjust to my industrial policy.

And so, most of those savings come to the United States, because no other market is deep or open enough to absorb them, directly or indirectly. The routing hardly matters: Chinese or German savings parked elsewhere displace other capital towards the United States, because the world’s surpluses must be matched by deficits, and America’s capital markets are both the most welcoming and the largest of all.

Oil exporters belong in the same story, even if they arrive by a different route. Kuwait, Qatar and Norway run surpluses because they earn a resource windfall and bank much of it rather than spend it. The source differs from China’s or Germany’s, a rent in one case and repressed demand in the other, yet the result is identical: saving that outruns domestic investment and has to go abroad. Most of their oil sells in dollars, which feeds a steady global demand for the currency. The dollars pile up and flow to the same place as every other surplus.

It’s true that the dollar’s grip on oil is loosening at the edges. A growing share of trades now settles in other currencies, mostly yuan, through channels that sidestep the Western financial system entirely. The shift is real but slow, since the convenience and depth of a dollar oil market are not easily undone. Even so, if that anchor keeps eroding, one of the steadier sources of demand for the dollar erodes with it.

Wherever the surplus comes from, whether from oil exporters such as Qatar or manufacturing powerhouses such as China, it converges on one destination. The United States runs a current account deficit of about $1.1 trillion a year, which supplies the world with dollars that have to be invested, and those dollars come back as purchases of American shares, bonds and property. The country works as a giant investment platform, an arrangement we can call “America Capital Partners”. By the end of 2025, foreigners held about $27.5 trillion more in US assets than Americans held abroad, close to 90% of GDP.

The dollars then spread across US assets, Treasuries, corporate bonds, shares and property. Foreign money lands first and heaviest in government debt, holding about a third of US Treasuries but under a fifth of the equity market. That weighting drives a kind of overflow mechanism. By taking so much of the bond supply, foreign buyers hold down the risk-free rate, which pushes domestic investors out of bonds and into shares in search of return. Foreign capital lifts share prices indirectly, through the rate, while the marginal buyer of US stocks stays American. Public-debt financing and asset-price inflation turn out to be two outlets for the same global plumbing.

Passive money and an inelastic market turn those flows into prices

A flow of money does not become a price on its own. It needs a mechanism, something that turns the inflow into higher prices. Two pieces of research explain how that works in today’s market.

We owe the first to Michael Green, another podcast guest. In his work on passive investing and its impact on the market, Michael estimates that passive strategies, index funds and their variants, now account for close to 45% of the US equity market. Unlike active investors, index funds make no judgment about what a company is worth. They are required to buy more of whatever is already largest, at whatever price the market sets. They are, by design, price-insensitive.

Take the SpaceX listing on 12 June. The stock did not join the big indices on day one, but the providers had already cleared its path. FTSE Russell rewrote its rules in May to fast-track giant IPOs and will add SpaceX to the Russell 1000 from 26 June, with MSCI following on 29 June. Once those dates arrive, every fund tracking the indices has to buy, whatever it makes of the company. Only the S&P 500 held the line: its profitability and seasoning rules bar a firm that lost $4.3 billion last quarter, so SpaceX stays out until at least 2027.

AkademikerPension, a Danish fund managing about $25 billion, saw the trap coming. Judging the stock grossly overvalued and catastrophically governed, with Musk holding some 85% of the votes, it struck SpaceX from its whole portfolio by hand, knowing the passive machinery would otherwise load the position for it. Most institutions did not bother. How ironic: escaping the system takes an active decision, and a passive investor, by definition, will not make one.

So with the trackers committed to buy and SpaceX fast-tracked almost everywhere, the stock arrives with weeks of forced buying lined up behind it. Musk need only rally early retail buyers, an art he masters to perfection, to make SpaceX a heavy index weight, and the passive funds then have to follow him up.

The second finding is what price-insensitivity does to a market. Work by Xavier Gabaix and Ralph Koijen, known as the inelastic markets hypothesis, finds that each dollar flowing into equities raises total market value by roughly five, on their estimate. In a normal market, rising prices attract sellers and deter buyers, keeping valuations anchored to reality. But when nearly half your buyers must purchase more of something simply because it got bigger, that stabilising mechanism breaks down. The market becomes inelastic, and a relatively small flow can move prices by a large amount.

AQR, a quantitative investment firm, puts numbers on the same point over a longer span. Its researchers Antti Ilmanen and Thomas Maloney found that US shares beat the rest of the developed world by about 4.7% a year over the 35 years to 2024, and that only a small slice of that gap came from American firms actually earning more. Most of it, 3.8 of the 4.7 points, came from rerating: investors agreeing to pay more for each dollar of those profits, so the share price climbs faster than the profit beneath it. The American lead is mostly investors paying up. The inelastic multiplier is what lets it run: money flows in, prices rise, the gains pull in more money, and the loop repeats. The market has run it for three decades, and AI is its current fuel.

The loudest story pulls in the most money

How should a public company play this game? If one dollar of inflow moves five dollars of value, the marginal dollar is worth chasing, and the surest way to win it is a louder story rather than a better set of accounts. The flow reacts to attention faster than it checks the numbers. So the rational move for any large company is to manage its own narrative and pull the flow towards itself.

Keynes saw this long ago, describing the market as a beauty contest in which the prize goes to whoever guesses what the crowd will favour. Robert Shiller gave it a modern name, narrative economics, the spread of contagious stories that move prices on their own. What people file under meme stock is the same mechanism seen from outside.

Elon Musk has turned it into a repeatable operation, using X, the platform he owns, to push Tesla, SpaceX and Dogecoin to more than 200 million followers. Each post is a flow event. When a dollar of inflow becomes five dollars of market value, controlling the story stops being mere communications and becomes the most leveraged instrument a chief executive has. The system is built to reward exactly this.

The gains have narrowed to a handful of AI names

Put the two together. The world’s surplus arrives in the US and flows into the largest, most liquid vehicles available, which means index funds. Those funds buy mechanically and without restraint, and because the market is inelastic, each dollar they deploy has an outsized effect on prices. The result is a market where valuations are driven less by what companies earn and more by the volume of money that has nowhere else to go, amplified by the loudness of the story being told.

One sector is playing that game hardest and winning it: AI.

Read more

Michael Pettis began his career on Wall Street, specializing in Latin America in the years after the debt crisis of the 1980s. By the late 1990s he was a managing director at one of the big Wall Street firms, running its Latin American capital markets. A few years later he found that the work had become repetitive, and he wanted to study an economy he knew little about, so he moved to China to teach finance. He taught first at Tsinghua University and then, for more than two decades, at Peking University. He remains based in Beijing, where he is a non-resident senior fellow at the Carnegie Endowment for International Peace. He is the author of several books, most recently Trade Wars Are Class Wars, written with Matthew C. Klein.

Michael is one of the few Western economists who follows the Chinese economy from the inside. His work centres on one question: why does China save so much and America so little? Most answers point to national character, but Michael points to policy. When a country holds down the household share of income through suppressed interest rates on savings, a weak currency, or wage restraint, the surplus it exports is the other side of the consumption it has taken from its own workers. A trade dispute, on Michael’s view, is a dispute over income between countries.

In this conversation, he explains why China’s overinvestment and its strong infrastructure are the same story, why infrastructure is a cost of growth rather than a source of it, and why financial repression pushed down the Chinese household share through the 2000s. He uses Europe, from Germany’s labor reforms to Spain’s credit boom, to show how these imbalances spread, and argues that deficit countries do not choose to deindustrialize but have it forced on them. He sets out his doubts about the dollar’s exorbitant privilege and about stablecoins, and reviews a century of precedents to reach the question he thinks the next round of talks will turn on: not how everyone wins, but who pays for the adjustment.

Michael writes for Carnegie’s China Financial Markets newsletter and on Substack, and is on X at @michaelxpettis.

We usually share these conversations in video, but a problem uploading the HD files means this one is audio only. Apologies for the change, and thank you for your understanding.

Nicolas: Hello, I’m Nicolas Colin.

Marieke: And I’m Marieke Flament.

Nicolas: And this is the Currency of Power podcast, the companion to the Currency of Power newsletter, where we dig into currencies, power, and the people connecting the two. Today our guest is Michael Pettis.

“Infrastructure is not a source of growth, it’s a cost of growth”

Michael, welcome. You’re a non-resident senior fellow at the Carnegie Endowment for International Peace, and you’re the author of several books, one of them being Trade Wars Are Class Wars with Matthew C. Klein. I think it’s the most recent, and it’s very much about the topic we will be discussing today, which is macro imbalances, macroeconomics in general, China exporting so much and the US consuming so much, and all the consequences of that situation all over the world, including, maybe, Europe.

We’d like to start with your journey. You’re based in Beijing. You’re a rare Western economist commenting on the Chinese economy from the inside, from the ground. Can you tell us what brought you to China in the first place, and what made you decide that you wanted to stay?

Michael: I started my career on Wall Street, where I specialized in Latin America. This was right after the global debt crisis of the 1980s. By the end of the 1990s, I was a managing director at one of the big Wall Street firms, in charge of Latin American capital markets.

At some point, it had become a little bit boring, in the sense that when I first started, everything we did was very new, had never been done before, so you had to put together the entire transaction. By the end of it, it was very cookie cutter. Every deal was the same as every previous deal. So I thought I would take some time off and learn something that I didn’t know much about. India and China were the two really big economies that I had very little familiarity with, so I decided to go to China for two years. I didn’t want to go as a banker. I had been teaching at Columbia University while I was working on Wall Street and I thought I could use that to teach a couple of years in China and get a chance to learn the economy from a point of view different to that of a banker.

I was very lucky. I was offered a position by one of the top two schools, Tsinghua University. But after two years, I decided that, although I’m a New York fanatic, I wasn’t ready to move back to New York yet. So I switched to Peking University, because people told me it was a very different school. Not really, but at the time that’s what I believed. And I’ve been there for 24 years.

Nicolas: Do you speak the language? Did you learn Mandarin?

Michael: Terribly. I can order a beer, I can tell a taxi cab what to do, but not much better than that.

Marieke: It’s really interesting, because you came at economics through a background as a finance person. You’ve been in the guts of the system. What different outlook does that give you when you look at teaching economics?

Michael: I stay away. I’ll tell you the truth, I’m very skeptical about academic economics, particularly of the so-called Anglo-Saxon variety, what’s taught in the US and the UK. I think it does a really good job of describing a world that frankly doesn’t exist, sort of an Adam Smithian world in which no one is big enough to intervene in the markets. But that’s not really what the world looks like. So my focus tends to be much more practical. The most famous course that I gave at Peking University was called the PBOC Shadow Committee, where the students pretended that they were the monetary committee of the Central Bank of China. We went through regular events occurring in the world, and we really spoke about how that affected investment, how that affected businesses, things like that. So as an economist I would fit much more in the world prior to the 1960s and ‘70s, which I think is the last time we were really good at economics, because I think we took a wrong turn in the 1970s towards a much drier, less useful understanding of economics.

Nicolas: That’s funny, because when Trump was re-elected in 2024 and conveyed the message that he was really serious about imposing tariffs and waging trade wars this time, I realized we all had to dust off our macroeconomic textbooks. I used to study macroeconomics at Sciences Po in 2000 and never used it for 25 years. Suddenly it’s become the main thing, and really a discipline you need to master to understand what’s going on. Maybe that’s the same impression.

Michael: It’s what I tell my friends. We have to reread Keynes. We have to read Joan Robinson, who I think is particularly important here; I think she did some of the best work on trade. We have to reread Ragnar Nurkse, we have to read Michał Kalecki, all of those ways of understanding macro, particularly within a globalized world, which is very different from macro in a non-globalized world.

Nicolas: So what does being based in China give you that other economists or analysts don't have? What do you learn? What do you observe on the ground that's difficult to decipher from the outside?

Michael: I’m very skeptical of the argument that the people who really understand an economy are the ones in the middle of the economy. I don’t think there is that much historical evidence that supports that. What happens is that if you live in a country you get many of the small things right that people who don’t live in your country might not get. But there’s no evidence that you get the big things right. If you look at the criticisms of the US in the 1920s, they mostly came from Europe. It was only in the 1930s that American economists started to recognize those criticisms.

If you look at something that’s very close to me, because my career started out of that, the huge borrowing and current account deficits of Latin America in the 1970s, it was mostly foreign economists, not a lot, but mostly foreign economists, who said this is completely unsustainable. In Japan, it was often Americans and Europeans who were the earliest to recognize problems in the Japanese growth model. In the late ‘80s, when Paul Krugman started saying there really are problems with this model, the standard response is something that we hear today: you’re a foreigner, you’ll never be able to understand Japan.

And I think we hear that a lot about China. But as I said, there is no evidence that an outsider is more likely to be systemically wrong than an insider. In fact, you could argue that maybe outsiders can be a little bit more open about some of the issues.

Now, having said that, I don’t want to say that my having spent 24 years in China is totally worthless. You do get a very different view. For example, I mentioned the small things. They’re not small, but they’re very specific. People have been telling China again and again that if you have a problem of deflation, it’s very simple, expand your money supply. That’s what they used to say to Japan in the ‘80s and ‘90s: it’s very simple, if you expand your money supply, everyone knows more money means inflation. And yet that hasn’t happened. Money has grown quite quickly in China, as it did in Japan in the late ‘80s and ‘90s, and yet they suffer from even deeper deflation.

The reason there is when it’s very useful to be close to a system. One of the first things you notice about the Chinese banking system, which was true about the Japanese banking system, is that unlike the banking system in the US and Europe, very little credit expansion is directed towards consumption. It’s almost all directed towards investment. So when you expand the money supply and expand loans, which is what happens in a really bank-centered system, then what you’re really doing is expanding supply more than you’re expanding demand. And as anyone can tell you, that’s disinflationary, not inflationary. So those are the kinds of things that you need to know about China. I’m not sure living in China gives you a particular advantage there, but if you specialize in trying to understand the Chinese economy, those are the advantages that you get.

Marieke: So what do you think we get wrong in the West about China? You mentioned deflation, the fact that the banking system is actually not the same. What other examples come to mind, things where there's currently the wrong perspective on what's actually happening?

Michael: I think sometimes people overestimate the power of Beijing to get things done. There is a perception that if Xi Jinping says jump, everybody in China jumps. It’s not really true.

Historically, there has always been a conflict between the power of the central government, whether it’s in Beijing or Xi’an or wherever it has been historically, and the power of local governments. That continues to be a very big issue in China. For example, a few days ago, the South China Morning Post had an interesting article about, I forget which industry it was, one of the industries suffering from involution, where Beijing has made it very clear you have to cut down capacity. And capacity is actually expanding, not cutting down, because local governments want to continue expanding domestic production. So all the yelling and screaming from Beijing doesn’t seem to have had as big an effect as you might have expected.

The other thing that I think people get very wrong is that they have two different stories about China, which they think are in opposition. You can argue that the bear case about China is massive overinvestment, obviously in the property sector, which has started to correct, but also in manufacturing and in infrastructure. And then you have the bull case for China, which is counterposed to that, and the argument there is, no, China is doing great. Look at its infrastructure, it has the best infrastructure in the world. Look at its technology, it’s catching up to the US in many areas and surpassing it in others. So it must be doing very well.

And the point there is that those two stories are not at all inconsistent. They could very well be the same story. When you overspend on infrastructure, you don’t get worse infrastructure, you get better infrastructure. The problem is, is it economically viable? Is it worth the cost? When you overspend on technology, you will get great technology, but it may bankrupt you. We’ve seen this before in the Soviet Union in the 1960s. The Soviets expended an enormous amount of resources on catching up and surpassing the US technologically, and they did in some areas. We all know about the famous Sputnik moment. And yet in the end, it didn’t really provide the productivity boost that Soviet officials were hoping would get them out of their domestic economic problems, I suspect because their great technology was never economically viable.

The same thing happened in Japan. I always tell people that the best way to understand China is to read as much as you can about Japan in the ‘70s, ‘80s, and ‘90s. I’m old enough to remember that in the late 1980s Japan easily had the best infrastructure in the world. Nobody compared to it. It also had some of the best technology in the world. Every cool consumer product in the 1990s was Japanese. And yet it still didn’t lead them out of the economic problems that the economy suffered. So I would argue that great technology is certainly a wonderful thing, but you can bankrupt yourself with great technology just as easily as you can get rich with great technology.

Nicolas: And what about scale in that case? Because China is much bigger than Japan ever was, more comparable to what the US was in the early 20th century compared to European nations. Can scale actually save you when the reckoning comes? Can China overcome the obstacles because its sheer scale makes it possible to impose new terms, a new paradigm onto how the economy works?

Michael: I think the scaling argument is one of those retrospective arguments. When an economy is growing very quickly, we look for reasons to explain that rapid growth. This is why it’s really useful to be obsessed about economic history in general, because scale isn’t as much of an advantage as you would think it is when you look at the historical precedents.

In terms of population, India is bigger than China. Those have been two huge countries basically for the last 100 years, longer of course, but in the last 100 years those are potentially the countries with the best scaling benefits, which they’ve used miserably. That hasn’t really been very successful. But more importantly, the extent of scaling isn’t a function of population, it’s a function of population times income. And Japan was 17% of the world in 1991. That’s what China is today. So when people talk about how massive China is, yes, it’s about as big as Japan was in the early 1990s, and that didn’t resolve Japanese problems. It’s hard to see why it would resolve Chinese problems.

For me, it’s a very simple arithmetic problem. If you take 50 euros of resources and combine it with 40 euros of resources and create 80 euros of value, it doesn’t matter how big you are, you’re losing money. There, I would say the poster child is the province of Guizhou in the south of China. You may know of Guizhou. It’s a beautiful mountainous province, a little bit richer than Cambodia, so it’s quite poor. And it has easily the most spectacular bridges in the world. More than 50 of the 100 highest bridges in the world are in this one little province, and approximately half of the top 10 are in this province. And yet it was the first province to go bankrupt.

The argument there is that there was this belief that infrastructure creates growth, that infrastructure is a source of growth. So because Guizhou is so poor, let’s massively spend on infrastructure and Guizhou will get rich. Well, it’s not that easy. Infrastructure is not a source of growth, it’s a cost of growth. If you could grow without spending money on infrastructure, you’d be better off. But when I say it’s a cost, that cost is only justified if the subsequent increase in productivity exceeds the value of building that infrastructure.

It turns out that more productive countries tend to benefit more from infrastructure than less productive countries, because you can basically say the value of infrastructure is the amount of man-hours it saves times the productivity per worker. In very low productivity countries you can absorb much less infrastructure productively than in high productivity countries. The reason Switzerland has more investment per capita than another mountainous landlocked economy like Bolivia is not because that infrastructure made Switzerland rich. It was because Switzerland was rich and was able to absorb that infrastructure. So the question then becomes, why is Switzerland richer than Bolivia? That’s the question that can win you a Nobel Prize if you answer it. But the simple answer is that Switzerland has the set of institutions, political institutions, legal institutions, financial institutions, that allow workers and businesses to operate at much higher levels of productivity. So what you really need is for infrastructure to keep pace with the growth in productivity, not for growth in infrastructure to replace growth in productivity. In China, you have the case where it’s really been increased spending on infrastructure that has driven much of the growth.

But then you run into a strange paradox, and Martin Wolf wrote about this in the Financial Times about a month ago. He said, how is it possible that a country that invests more than 40% of GDP is only growing at 5%? That’s, in theory, impossible. But that is what we see in China. And the only possible answer is that that investment is not real investment, it’s non-productive.

Just to give you a sense of how much of an outlier China is: investment is roughly 25% of global GDP, and that’s been a pretty consistent number. But it’s not the same in every country. Europe, the US, the capital-rich countries invest much less. They invest around 18 to 20% of their GDPs. Developing countries, poor countries, also invest much less. But there are countries that invest much more. Very rapidly growing developing economies growing at 6%, 7%, 8% might invest 32%, 33% of their GDP. That tends to be the high end. I once saw that South Korea invested as much as 39% of its GDP for one or two years, and that’s the only higher number. But China for the last 40 years, for the last 30 years, has invested between 40 and 47% of its GDP year after year after year.

And as in Japan in the 1980s, as in Brazil in the 1960s and ‘70s, all of that investment created significant economic activity that looked like growth. But we realized in retrospect that much of that investment, the famous bridges to nowhere in Japan, had no economic value. So they actually made the country poorer, not richer. Obviously, if you’re a very big country, there’s probably more room for you to invest in productive infrastructure. But there’s always a limit. And once you pass that limit, then all of this investment is driving GDP growth, but not real economic growth.

Marieke: It's really interesting, because hearing what you were saying, I'm thinking about AI, which is at the same time an infrastructure but also carries the wish that it will increase productivity. And listening to you, I was thinking, well, who is it going to bankrupt? Because when you look at the massive valuations that are floating all around the world, and the very different paths that China is taking from the US in terms of its AI investment and development, I don't know, do you have a view on the role that AI will play in the model you were describing?

Michael: If you look back at these technology bubbles, or, I shouldn’t call them a bubble, stages of the Industrial Revolution, periods when we saw very rapid growth in technology and very rapid growth in productivity, it’s interesting, because I used to track stages of global financial liquidity being driven towards developing economies. You’ve had many bubbles in developing economies, Latin America in 1925, several in the last 200 years. And there was an economic historian at Oxford, I forget his name, who listed stages of the Industrial Revolution where you saw very rapid increases in productivity and technology. I was amazed, because his stages exactly matched my stages. What it suggested to me is that a lot of this is driven by, for whatever reason, a significant increase in risk appetite, perhaps because of a massive expansion in liquidity. Several of these stages were driven by gold discoveries in the 19th century, for example. So risk appetite goes up and money pours into risky ventures, risky countries, risky everything. And then you have the bubble and the bust.

But not all technology bubbles have been bad. The railroad bubble in the 1860s in the United States collapsed in 1873, and yet the US was left with spectacular infrastructure. We’ve seen many cases of this. Remember there was a bubble in the 1990s for companies laying out the framework for the internet, the cables and all that. Most of them went bankrupt too, but they left the world with very good technology, which then presumably boosted productivity growth. So you can have good technology bubbles and bad bubbles. I would say real estate bubbles are almost always harmful for the economy. Infrastructure bubbles can be or cannot be, depending on their extent. And the same thing with technology bubbles. This is sort of a cop-out, because what I’m saying is that we don’t really know. Sometimes these are good things, sometimes these are bad things. I think we have to bear that in mind. But China really is different. The level of investment is just incredible. We’ve never seen so much investment pouring into an economy for such a long period of time.

“Whether you like it or not, your economy will adjust to my industrial policy.”

Nicolas: So the complement of that huge investment effort is low consumption. When people try to understand the macroeconomic situation, they always look for causality, and macroeconomics is difficult to understand for that reason, because it's a complex system of feedback loops, each loop feeding into the others, and we don't really know what started the whole thing. But I would say, after having read much of your work and listening to many of your interviews, that at the heart of it there's this phenomenon of very low household consumption in China. Is it fair to bring it back to that core phenomenon?

Michael: There are two arguments here. To simplify the world, let’s say there are two countries, China and the US. The US has very low savings, and China has very high savings. And globally, savings must equal investment. So by definition, if one country saves more than it invests, the other country must invest more than it saves. That’s definitionally true. It could be just an astonishing coincidence that the world always balances. It’s a little bit hard to believe. So if you don’t believe it’s a coincidence, then you must believe that there is some sort of causal relationship. Either high-saving countries force low savings elsewhere, or low-saving countries force high savings elsewhere.

Now, for many Americans, and I think for many people, it’s hard to believe in a world in which the US doesn’t have agency. As one of my professors told me many years ago, most Americans believe that either everything good in the world is caused by the United States, or everything bad in the world is caused by the United States. In fact, lots of things happen whether the US wants them or not. So for me, the question is, which way does causality flow between China and the US?

Here I’m very Joan Robinson. If I control my external imbalances, my external account, if I control my capital account and my trade account, then I’m able to create domestic imbalances, which are reflected in my external imbalances. Because normally, if I subsidize manufacturing at the expense of the household sector, I should run a trade surplus. That trade surplus should change capital flows in ways that probably will drive up my currency. As my currency goes up, households are better off and manufacturers are worse off, so I rebalance.

Now, if I don’t want to rebalance, I control my capital account, I intervene in my currency. So here’s the thing. If I control my external accounts and you don’t control your external accounts, then which way is causality likely to run? Probably from me to you. So I would argue in that case the source of the imbalances is more likely to be in countries like China that very sharply control their external accounts.

I would add to this that the decline in the household share of GDP wasn’t smooth. It began in the 1980s, when Chinese consumption was quite normal. Chinese savings were quite normal by global standards, and then it started coming down all the way until 2011. Then it went up a little bit until 2019, and since then it’s been coming down again. But the really sharp decline occurred in the 2000s, between 2002 and 2010. That’s when we saw an almost collapse in the household share of GDP.

So what explains that? That’s when China decided to clean up its banking system. In the 1990s, bad loans were such a problem that private sector estimates were that as much as 40% of the loans were bad. And the People’s Bank of China (PBOC) never really disagreed. They didn’t agree, but had they disagreed, they would have said something, and they didn’t. China, because it was initiating the IPOs of its four big banks, a really big political event in China, had to clean up the banks.

The way they cleaned up the banks was by shifting the bad loans to what were called asset management companies, which were themselves funded by the banks. So it didn’t really solve anything. But the way they solved it, if you look at real interest rates in China, they were hugely negative basically from 2000 to 2011. And when I say negative, I mean the GDP deflator in China was 8% to 10% during that decade, the nominal growth rate in China was 16% to 20% during that decade, and the deposit rate was 2.5% to 3%. So if you put your money in a savings account, much of it was confiscated through this hidden tax. The IMF estimated that as much as 5 to 8% of GDP was transferred every year from the household sector to the rest of the economy in the form of this tax. So to me, it’s not at all surprising that the income of the household sector, basically there was a huge tax on their savings, so their disposable income as a share of GDP dropped very rapidly. And as a result, we saw a huge surge in the Chinese trade surplus.

So to me, when people say that that must have been the result of domestic policies in the United States, I find that really hard to reconcile with what China was doing at the time. It was very specifically cleaning up the banking system using financial repression, and that had to have an impact on income distribution, which in turn had to have an impact on its external account, et cetera, et cetera.

So that’s really my big argument with, not argument, but I meet with a lot of European policymakers. In fact, next week I’ll be going to Brussels, and this is what I tell them. Again, it’s a very Joan Robinson argument, and the argument is this: if I am able to control my external account to the point where I can externalize the costs of my internal imbalances, and you don’t control your external account, then whether you like it or not, your economy will adjust to my industrial policy. So not only do I get to decide the industrial policy for my country, I get to decide the broad outlines of the industrial policy for your country. You get to choose whether you absorb it in the form of higher debt or higher unemployment, but you are going to lose manufacturing, whether you like it or not. And that’s something that I think is incompatible with a global trading system.

Marieke: I think it's fascinating, because we talk about repression also a bit at the individual level. But I can't help but look, with a 20-year perspective, at the fact that there's been a massive uplift in terms of poverty in China, and there are fewer people that are very, very poor. There's also a major amount of billionaires and so on. So how do we reconcile both things? Because on one hand, what you were describing does not sound great, but on the other hand, if you go witness it and look at some numbers, people seem better off.

Michael: And that was the story with Japan too. Japan lost 60% of its economy after the war, and by, I think, the 1960s, it had recovered it fully. That was a real recovery. Brazil, the first country I believe to be called an economic miracle, in the late ‘50s and ‘60s and early ‘70s, was growing very quickly and very healthily. It was only in the mid to late ‘70s that we started to see the problems emerge. The Soviet Union: we forget that in the 1960s, most intelligent people, not just in the Soviet Union but also in the US, from President Kennedy down, believed that the Soviet Union was going to overtake the US economically and technologically sometime in the 1980s. You can find it in Paul Samuelson’s first edition of his book on economics. I think he said 1984. So it’s always a bit of a surprise, because there have been real advances.

What I would argue is that a high-investment model, which all of those countries followed, is the medicine that the doctor ordered when you are significantly underinvested. Japan after the war, Brazil as a typical developing country, China, remember, when it began its reforms, had gone through five decades of first the anti-Japanese War, then Civil War, then Maoism. So China’s infrastructure, housing stock, and manufacturing stock was a total disaster. On the social basis, it was not so bad. It was relatively educated, very healthy, with political stability. So China, I would argue, entered the reform period hugely underinvested, in which case very high levels of investment are exactly what you need.

The problem is that when you have a very rapid growth in investment, and in China it was the highest growth rate in history, then at some point you close the gap between what you have and what you need. And that’s when you should shift your growth model. No country, by the way, has ever shifted its growth model in time. One of my favorite economists, Albert Hirschman, this was a problem that really puzzled him in the ‘70s. His conclusion was that a very successful growth model creates a very powerful constituency that benefits from that growth model, and they tend to block efforts to change the model, which is why you always overshoot.

Now, how can you tell if you overshoot? In China, as in Japan, there’s a very easy way of telling. Remember, I said that in China almost all lending goes to investment. So if you are investing productively, your debt can grow very quickly, but your debt-to-GDP won’t grow at all, because presumably that productive investment is creating real growth in the economy. If you look at China before 2008, 2009, debt grew very rapidly, but the debt-to-GDP ratio grew slowly, bouncing around. It really wasn’t a problem.

But it’s precisely around 2008, 2009, probably not a coincidence that that’s the global financial crisis, that you start to see an acceleration in debt and a deceleration in GDP growth. That should really puzzle us. If you saw that in the US or Europe, fine, because a lot of lending in the US and Europe goes to consumption. But in China, almost none went to consumption, it all went to investment. So that’s a real puzzle. And the only explanation I can come up with is that an increasing amount of this investment was non-productive.

Now, that should show up in the productivity data, and it does. Productivity growth has slowed very sharply since then, and even that’s overstated if you believe the GDP growth is overstated. So the story we see in China is a very good model that reached the end of its useful life, but they were unable to shift the model. Every country has had that problem. And they kept it going for far too long.

Nicolas: Before we discuss the deindustrialization and maybe the role of the dollar as a reserve currency, I'd like to embrace the perspective of a single household. If I'm a Chinese household, what prevents me from consuming more? What constraints do I feel that force this behavior on me? And if I'm an American household, what prevents me from saving more? I think it would help people understand better how the macroeconomic situation translates into individual behavior on the ground, both in China and the US.

Michael: Sure. The problem in China is quite simple. The household share of GDP is very low. So if you produce €100 worth of goods and I pay you €80, your consumption will probably be €70 to €75. If I pay you €60, obviously your consumption isn’t going to be that high, it’ll be lower than that. And China has the lowest household share of GDP in the world. So to me that’s pretty straightforward. There are variations in it, because if I’m more optimistic about the future, my saving rate might go down, and so my consumption rate will go up. If I’m pessimistic, which is where people are now, my saving rate will go up. But basically the long-term constraint on my consumption has got to be my income.

Now, why do Americans consume so much more than they should? There, I think it’s easier to look at Europe than at the US, because in Europe you saw a very clear example of what happened. Remember that when the euro was created in 2000, 2001, when the currency came out, a country like Spain, I was born and grew up in Spain so I focus mostly on Spain, but it could be Greece, it could be Italy, it could be Portugal, it could even be France to some extent, a country like Spain actually had a fiscal surplus and a tiny trade imbalance. And then something happened. Spain, Portugal, Greece, even France, Italy, suddenly they all went crazy and started spending like crazy.

If you read the German press in 2009, 2010, the problem with Spain is that they always go to the beach and have siestas and they don’t work. Well, that only started after 2003. And it started in all these other European countries at the same time. So that’s a remarkable coincidence until, and I’m sure you know this, you come to what was probably the real driver of all of this. And that was that in 2003, 2005, Germany implemented the Hartz reforms and the labor reforms.

You can call them labor reforms, but it’s a euphemism for wage suppression. Because if you look at the relationship between wages and GDP growth, before 2003, wages kept pace with GDP growth. And after 2003, GDP growth continued and wages stopped growing. So if the household income share of GDP goes down, something else must go up. And what went up? It was business profits, which surged. And of course, that means consumption will go down. So German consumption went down, not because the Germans became thriftier, but because they received a smaller share of what they produced. And the German savings rate soared.

Now, the reason Europe is so useful is because it’s a small experiment, in the sense that the creation of the euro and the rules governing the single market meant that most of the excess saving in Germany didn’t go to the US or anywhere else, it went to other countries within Europe. Interestingly enough, the countries that entered the euro with low inflation ended up exporting capital to the countries that entered the euro with high inflation, which makes sense if interest rates converge, because they become negative in the high-inflation countries and positive in the low-inflation countries.

So in a country like Spain, money poured into the country, into the banking system. And again, from a personal point of view, as late as the 1990s, getting a credit card in Spain was really prestigious. If you had a credit card, you would brag about it. By the end of the 2000s, if your children didn’t have multiple credit cards, then you were a total loser. Everyone had credit cards. So what really happened? Why did the Spanish suddenly go crazy? I would argue that it isn’t that the Spanish went crazy.

If you assume that Spain, like every country in the world, has a normal distribution of risk-taking, there are people who are too prudent, people who are not prudent enough, people who are overly optimistic, people who are foolish. You can always get the consumption rate to go up, not because the whole culture of the country changes, but because if you loosen liquidity conditions in the banking system, the banks are eager to expand lending, and the way you can always expand lending is by lowering your credit standards. So as money poured into the Spanish banking system, banks were under increasing pressure to do something with it. Some of it they used to fund pretty good infrastructure. Spain has decent infrastructure. Some of it went to fund the worst real estate bubble after Ireland’s. It also helped the Irish real estate bubble. And a lot of it went to fund the surge in consumption. I would argue that that’s the mechanism. When money pours into your country, when there’s a net inflow, by definition you need a gap between investment and savings. And that can happen in many ways. If you are a developing country whose investment is constrained by scarce capital, then money inflows could drive up the investment rate.

But in Europe, there’s no scarcity of capital. A recent ECB study asking businesses in Europe why they aren’t investing, none of them said, or very few of them said, because we can’t access capital. They mostly said because we don’t have demand, or the number two reason, which is the same reason in my opinion, because it’s not profitable. So if you go to these businesses that are not investing because they are not competitive with Asian countries, and you say, here’s money, build more industrial capacity, they’ll say, no, we don’t need more capacity, we can barely sell what we’re producing.

So if investment doesn’t go up, then something must happen to cause saving to go down. It’s got to balance. So what could happen? In the world of Keynes and Joan Robinson, it was very clear what would happen. Unemployment would go up, because as German manufacturers expanded their share of the Spanish market, Spanish manufacturers would close down and unemployment would go up. And remember, conveniently enough, unemployed workers have a negative saving rate. So that’s how you balance it. In the modern world, we don’t like unemployment to go up. So we have two ways of dealing with it. We can expand household borrowing through monetary policy, or we can expand the fiscal deficit. And Spain did both. The fiscal deficit went up and household debt went up. And that’s the way you balance without unemployment.

The problem is that in 2009, when Spain was no longer able to increase debt to balance its trade deficit, how did it respond? It responded in the Keynesian way, with a surge in unemployment. So I would argue that that’s really what drives up consumption. When people say that Americans are addicted to consuming, I would say no. If you have a banking system with very loose lending standards, like we did with real estate in 2003, 2004, 2005, you will get very risky, foolish behavior. So that’s really what matters. Sorry for that very long answer.

“Foreign inflows into the US are bigger than ever”

Nicolas: No, it's very clear, and also very interesting for us as Europeans to reflect on the role of Germany in all of that. So maybe coming to deindustrialization, I think it's pretty clear from everything you've said so far that the excess investment in one country, in that case China, or maybe Germany from when the euro was set up in 2002, triggers deindustrialization in the rest of the world, in countries that don't control their capital accounts. So is that the origin of that trauma, all the factories closing?

Read more

In twelve months, we’ve published 40+ deep-dives, built a readership of 1,000+ subscribers with a 60% open rate — three times the industry average — and reached millions across LinkedIn, Substack, X and major media outlets.

Our work has been quoted in outlets such as Bloomberg, The Observer, Hidden Forces, Manager Magazin, L’ADN

Read more

Pax means peace. The Roman Senate coined the phrase, the historians burnished it, and every hegemon since has borrowed the concept. Yet if you survey the world in 2026 — trade wars, export controls, chip embargoes, rare earth weaponisation, drone-contested borders, stablecoin legislation passed in the shadow of geopolitical anxiety — peace is the last word that comes to mind. Even the ‘Pax Silica Initiative’, launched in Washington in December 2025 with ambitions to be “what the G7 was to the industrial age,” carries the name of a peace that does not yet exist and may never arrive. Such a discrepancy is deliberate: every Pax in history has been imposed rather than negotiated. The question is always who does the imposing — and with what coin.

Every Pax Has a Coin

Strip away the military history, the diplomatic treaties and the cultural mythology, and what sits at the centre of every hegemonic order is a single monetary truth. What binds the order together is a coin, around which coalitions form and through which shared values acquire force.

Pax Romana. The denarius was Rome’s instrument of integration: it paid soldiers, standardised trade across three continents, turned conquered territories into economic participants, and made Roman power legible at every marketplace from Britannia to the Levant. When the emperors of the third century began debasing the silver content to fund frontier wars, the markets registered the end of Roman hegemony well before any decree acknowledged it. The denarius lost credibility before the legions did, and the monetary collapse of the third century preceded political fragmentation by decades. By 476, when Rome fell to Odoacer, the currency had long since died in the money changers’ hands.

Pax Sinica. China’s historical hegemony operated on a different logic but reached the same conclusion. The tributary system of the Han, Tang, and Ming dynasties was a gravitational monetary order: silver flowed inward to the Middle Kingdom as tribute, as payment for Chinese goods, and as proof of a peripheral state’s acknowledgement of the centre. The Ming-era decision to fiscalise the tribute system, demanding payment in silver rather than symbolic goods, turned an ideological hierarchy into a monetary one. Participation in the Pax Sinica turned on silver rather than ethnicity: it meant trading with China, which meant accepting China’s terms. Here too, the currency functioned as the architecture of the order.

Pax Americana. After the collapse of the Bretton Woods system in 1971-1973, the petrodollar became the most elegant version of this logic in modern history. America’s 1974 deal with Saudi Arabia is well known by now: price oil in dollars, recycle the surplus into US Treasury bonds, and you have created permanent structural demand for American currency from every nation on earth that needs energy — which is every nation on earth. Even without the Bretton Woods system, the dollar became the air the global economy breathed, invisible and omnipresent, controlled by one central bank. Dependence on it was universal and unchosen.

Pax Silica’s Monetary Answer

The Pax Silica Initiative, launched in December 2025, has been described as a supply chain alliance, a semiconductor coalition, a critical minerals compact. All of that is true but, to an extent, all of it obscures the point. In reality, what Washington is building, layer by layer, is the infrastructure for the next version of the same monetary hegemony that the petrodollar created. The AI economy needs compute the way the industrial economy needed oil. Whoever ensures that compute is priced, settled, and financed in dollars wins the same structural game that Henry Kissinger won in Riyadh fifty years ago.

As we argued in Barrels to Bytes, the most plausible vehicle for this new phase is the dollar-denominated stablecoin, ahead of a central bank digital currency, a new Bretton Woods architecture or a reformed IMF special drawing right. USDC, USDT and whatever comes next are the denarius of the AI age. The system has an underlying elegance: AI systems transact at machine speed, across borders, in volumes and frequencies that existing banking rails cannot handle. An autonomous agent paying for API access, a robot settling a logistics invoice, a data pipeline micropaying for training data: none of these fit inside a system built for human-speed, branch-based correspondent banking, whereas stablecoins do. They are programmable, instantaneous, borderless and denominated in dollars. The Genius Act, passed by the US Senate in 2025, did more than regulate stablecoins. It consecrated them as an instrument of American monetary foreign policy, making explicit what the market had already decided.

The strategy follows from this premise. Pax Silica secures the physical stack — the mines, the plants, the data centres, the energy infrastructure — while dollars and dollar stablecoins settle the transactions that flow through it. The coalition of trusted partners that Washington is assembling around compute and critical minerals will, by design, conduct their AI-economy commerce in dollars. The petrodollar recycled oil revenues into Treasury demand; the AI-era equivalent will recycle compute revenues into stablecoin dollar demand. The mechanism differs but the outcome is identical: structural, self-reinforcing dollar hegemony, this time built into the payment rails of artificial intelligence rather than the oil futures market.

Can the World Be Cut in Two?

Read more

Neha Narula started her career building databases — fast, scalable, multi-core systems designed to handle the cloud infrastructure of companies like Google, Microsoft, and Meta. The work was intellectually interesting. It was also, as she eventually concluded, being used mostly to keep people scrolling and clicking on ads. That dissatisfaction sent her looking for harder problems.

She found one in Bitcoin. Not the ideology, and not the investment thesis, but the engineering puzzle: why is a payment system so difficult to scale? The question pulled her into the mechanics of how money actually works — central banks, fractional reserve banking, the basic plumbing of the global economy — and she emerged with a view that has shaped her work ever since. Money is not a natural fact. It is a system built on design choices, and those choices encode political decisions about who gets access, on what terms, and at whose expense.

That view now sits at the centre of some of the most consequential debates in finance. As Director of the Digital Currency Initiative at MIT’s Media Lab, Neha led Project Hamilton, the research collaboration with the Federal Reserve Bank of Boston that laid the technical groundwork for a potential digital dollar. She has written on the hidden risks in stablecoin infrastructure and on what it would actually take for Bitcoin to become quantum-resistant. She sits on the board of Block and advises the Federal Reserve Bank of New York.

In this conversation, Neha explains why CBDCs and stablecoins are less different than most people assume, what Project Hamilton set out to do and what it left unresolved, why zero-knowledge proofs are not a privacy solution on their own, and what a decentralized community like Bitcoin can realistically do to prepare for quantum computing. She also takes the geopolitical temperature: what Iran’s reported willingness to accept Bitcoin payments actually signals, and why, in her view, the dollar remains the game to lose.

Her website is at nehanarula.org. The Digital Currency Initiative is at dci.mit.edu. She is on X at @neha and on LinkedIn.

Marieke: Hi! I’m Marieke Flament.

Nicolas: And I’m Nicolas Colin.

Marieke: And this is the Currency of Power podcast, the companion to the Currency of Power newsletter, where we dig into currencies, power, and the people connecting the two.

Today’s guest is uniquely positioned to bridge deep technical engineering, public policy, and mainstream finance when it comes to crypto and digital assets. Neha Narula started as a senior software engineer at Google and helped relaunch Dig before completing her PhD at MIT, where she built Doppel, a fast, scalable, multi-core database. She currently serves as Director of the Digital Currency Initiative at the MIT Media Lab, a board member at Block — formerly Square — and a member of the Federal Reserve Bank of New York’s Innovation Advisory Council.

Today’s conversation runs in four parts. The first two are for all listeners; the last two for paying members. Thank you to all members for following us, and to those paying, thank you. Part one covers Neha’s career and how she entered the digital asset space. Part two covers stablecoins, Project Hamilton, and why the US does not want a CBDC — or does it — and what that means for other countries. Part three covers privacy and quantum. Part four covers the geopolitical outlook.

Neha, you and I recently met at an ING Orange event in Amsterdam. We shared a ride and couldn’t stop talking about stablecoins and digital assets. I’m delighted to have you here. Thank you for making the time.

Neha: Thank you. It’s wonderful to be with you both, and I’m excited to continue the conversation we started there.

“The success of platforms like Bitcoin has done a great deal to rethink how to build decentralized systems.”

Marieke: Let’s get started. We love understanding people’s journeys — what got them into what they’re doing. Walk us through yours, and in particular the moment when you realized that distributed systems could do something important for how money works.

Neha: I was doing my PhD at MIT in computer science, working on databases and distributed systems — nothing to do with cryptocurrency. I finished around 2015. Bitcoin was launched in 2009. I’d heard of it around 2010 or 2011 but hadn’t paid much attention. A couple of friends were very enthusiastic about it, but I wasn’t convinced. The proof-of-work mechanism seemed odd.

By 2015, I had finished my PhD and reached a moment of real doubt about what I was doing. I had been working on very fast, scalable databases — systems that run the cloud for large tech companies like Microsoft, Meta, Oracle, and Google. But at the time, that work was being used mostly for advertising and for products married to advertising. There was a lot of effort going into keeping people scrolling and clicking, showing them more ads. It was deeply unsatisfying. I wasn’t sure the work I found intellectually interesting was being used to solve real problems.

I took some time off and finally had the chance to read the news. I kept seeing articles about Bitcoin having trouble scaling. 2015 was the height of what we call the “Blocksize wars” — a large debate about how to get Bitcoin to scale, reduce fees, and increase users. I read those articles and thought: what is the problem here? Bitcoin is a payment system. Payment systems are easy to scale. What could possibly be hard? That is how I got — we like to say — nerd-sniped into the cryptocurrency world.

I fell down the rabbit hole. I learned how Bitcoin worked, but more importantly, I learned how money worked, and I had no idea. I learned about central banks, commercial banking, fractional reserve banking, interest rates, the basic infrastructure of the economy. And I thought: this is so different from anything I had imagined. There are a lot of people left behind — people who pay very high fees to access financial infrastructure, people who carry debt their entire lives while others have theirs forgiven. This happens not just at the individual level but at the level of nation states.

These parameters are built into our financial system, but they don’t have to be this way. Cryptocurrencies seemed like the best tool for change, because changing things from within the financial system seemed far too hard. Cryptocurrencies offered a whole new way of thinking about structure — very much parallel to the architecture of the internet, with packets of information and protocols layered on top of one another. And the underlying technology — distributed consensus, peer-to-peer networks — was directly related to the problems I had been working on. It felt genuinely exciting. That is how I got into it.

Nicolas: Would you say that software architecture has already been changed by the lessons learned in crypto?

Neha: Yes, I think that is fair to say. The idea of decentralisation existed before crypto, but the success of platforms like Bitcoin, Ethereum, and stablecoins has done a great deal to rethink how to build decentralized systems. Before Bitcoin and Ethereum, the most well-known decentralized system was BitTorrent — a way of sharing files over the internet without a trusted third party. In 2006, BitTorrent was the majority of traffic on the internet, used primarily for music and films, but also for distributing software and other files.

BitTorrent was a precursor to cryptocurrencies, but crypto has taken that technology well beyond it. And some ideas from cryptocurrency have been filtering back into traditional finance — the idea of digital wallets, more open access, real-time 24/7 payments and settlement.

Marieke: You gave a TED Talk almost exactly ten years ago that now has 2.6 million views. You outlined the programmability of money and what it was going to change. It feels like it took ten years for everyone to arrive at the point you described. What is your view for the next ten years? And when it comes to programmable money, what changes in power are you starting to see in the financial system?

Neha: I stand by that talk, which is remarkable considering it was ten years ago. Back then, everyone was confused about cryptocurrencies. People would start with SHA-256 and miners and electricity and proof of work, and just confuse their audience entirely. The framework wasn’t there. I was trying to separate what would be true in five to ten years from the implementation details that, while important, were not the critical thing to understand.

Going forward, we are still capitalising on programmable money. That is still in progress, but we are seeing a great deal happen. And you have to make AI a central part of any ten-year outlook. People are very excited about AI agents — autonomous software that needs to transact economically, make payments sometimes on behalf of users, but sometimes at a scale and speed at which the user will not be in the loop. Are stablecoins or cryptocurrencies the best tool for that? There are interesting arguments on both sides. One of the actors doing the most in this area is Visa, a very traditional company, which released a command-line interface for AI agent payments. So there is movement there.

But it is not just how AI agents will make payments. It is how AI as a tool will change the way we think about money and how money flows. One example: monetary policy today involves a great deal of data and modelling, but ultimately it is humans deciding. I think it is inevitable that this will become more and more algorithmic — more automated, with perhaps a human in the loop but not at the centre. Another example: most money today sits on the balance sheets of commercial banks. What is the next step? Will stablecoins or new types of fintechs create money? What will the creation of money look like?

Marieke: It sounds as though, if anyone at TED is listening, they should invite you to give the next ten-year talk.

Neha: I would love that.

“Looking back, I think I underestimated how political money really is”

Nicolas: Let’s move to Project Hamilton. I imagine you worked on it before Trump returned to power, because CBDCs are now out of favor with the current administration. And the name — is it Hamilton because Alexander Hamilton was the founding father of the US financial system, and the project was about reinventing or experimenting with that?

Neha: Yes, and it is actually a lovely story. The name did come in part from Alexander Hamilton, but there was another Hamilton the project was named for as well — Margaret Hamilton, a software engineer during the Apollo missions who was critical in writing much of the software for those missions. There is a famous photograph of her standing next to a stack of books representing all the code her team wrote. Jim Cunha, who was a leader at the Boston Fed at the time and pushed for the collaboration, chose the name for both reasons — technology and the person who helped architect the financial system. It is a wonderful coincidence that it is the same name.

At no point had a decision been made to actually build a digital dollar and launch it. This was about doing the research necessary to form a foundation and understand what a digital dollar could be, how it would work, before even deciding whether to explore it as a serious option. Technology sets the rules of any system — what it can do, what is even possible — so we wanted to build that understanding.

Our initial goal was around scalability. Something at the scale of a digital dollar, operating in an economy the size of the United States and potentially globally, would need to handle hundreds of millions or billions of users with very high throughput and very low latency. That was a good match for my expertise. The Boston Fed knew about our work in cryptocurrencies and cybersecurity, and so we collaborated. It was a wonderful partnership — we continued to work with some of those people even after the project ended.

But the work was just the tip of the iceberg. We started with scalability and security. We did some work on programmability. We did not get to touch privacy fully, though we did design the initial system so that it did not store personally identifying information, did not store balances in clear text, and did not store user addresses or pseudonyms in clear text. That was rooted in the idea that central banks do not want a honeypot of sensitive information. They want to focus on building a currency that fulfils their mandates — in the United States, employment and price stability. They did not want to know all the payments users were making.

We kept the work very technical — distributed systems, software architecture, published papers. We tried not to get into policy because that was not our domain. That was the job of the Federal Reserve Board of Governors, Congress, and the executive branch. But looking back, I think I underestimated how political money really is. Even clearly technical work that is, I believe, quite neutral is very hard to divorce from people’s feelings about money — their worries, their fears, their incentives, political and otherwise.

There are real fears around surveillance. People are worried about the government, their employer, friends, family, and coworkers potentially seeing their financial transactions, which reveal a great deal about who we are. But people are not just worried about data leaking. They are worried about control.

There is a conversation happening around the world about what I want to call restricted money — as distinct from programmable money. The idea that government-issued money could restrict how it is used. SNAP benefits in the United States are a good example: payments for food that can only be used on certain things. But we do not call SNAP money. We think of it as a coupon or a voucher. When you start to mix the idea of restricting money with the concept of money itself, that becomes very worrisome. There were fears that a CBDC would prevent you from buying certain things, donating to certain political parties, or that spending data would be used to deny visas. I never want to build a system that can be used in that manner. The system we built cannot, I believe, be used that way. But I understand the concerns.

The 2024 election was highly charged. The idea of a CBDC immediately fell along partisan lines. We now have an executive order from Trump rescinding Biden’s executive order on digital assets, and an anti-CBDC Surveillance Act that has passed the House. There is currently a moratorium on this type of research. For what it is worth, I do not know whether it is a good idea for the United States to launch a digital currency. I was clear about that throughout — I was not advocating for it. But I think it is a serious mistake to suppress research. That starts to encroach on questions of free speech and academic freedom. It is important to be able to study things.

Marieke: Is your reading of the Anti-CBDC Surveillance Act that it blocks both wholesale and retail CBDC? And what is the actual US position, given that other countries are still testing and embracing CBDCs?

Neha: I have not read the act recently, so I may be slightly wrong, but I believe it is primarily focused on retail. Wholesale CBDC concerns them far less. The Federal Reserve Bank of New York is still working with the BIS on some wholesale CBDC concepts. A wholesale CBDC would only be used by large financial institutions — and the question then becomes why it is better than FedNow or Fedwire. You could add programmability, atomic settlement, and so on. But then why can you not do that with APIs on top of FedNow or Fedwire? These are genuinely interesting questions.

To answer your direct question: it was primarily about retail, and some wholesale CBDC research is still ongoing. However, the act was written quite broadly. It essentially said the Federal Reserve could not issue a digital version of the dollar, and I think even existing central bank reserves could fall under that definition. It needs to be sorted out before anything is implemented.

Nicolas: Because money is largely digital already. The concern seems to be about preventing a permanent connection between the dollar you use in daily life and the central bank. How are you looking at the geopolitical landscape, with different countries going in different directions?

Neha: We work across all those categories — CBDC, stablecoins, tokenized deposits — and I think they are more connected than people realize. A lot of people see CBDC versus stablecoins as a binary. I do not think that is true at all. One way of viewing a stablecoin is as a wrapped CBDC. Stablecoins currently wrap Treasuries in the United States, and to some extent commercial bank accounts and money market funds, which ultimately wrap Treasuries.

A lot of stablecoin providers would love access to central bank balance sheets. In the UK, they are actually thinking about that — stablecoins backed by the central bank balance sheet. That is a wholesale CBDC. You are broadening access to the central bank balance sheet, and once you wrap that in a stablecoin, it becomes a retail instrument issued by the private sector instead of the central bank. That is essentially a two-tier CBDC, which is exactly the CBDC architecture where banks issue the tokens.

The whole point is that we are tokenising money — turning it into this packetized thing. Whether it is issued by a central bank or the private sector, what matters are the design choices: what is it backed by, what are the legal agreements behind that backing, what happens in bankruptcy, can users access something of value when institutions fail? From a technology standpoint, a CBDC and a stablecoin are the same idea with different choices for those parameters.

I do not want to minimise the differences between something issued by a central bank and something issued by a private actor — the legal frameworks are quite different. But at a high level, they are the same thing. People do not quite understand how connected they are, or how design choices can blend them together.

Nicolas: Can you give an accessible explanation of the difference between wrapping Treasuries, as stablecoins do under the GENIUS Act, and having access to the central bank balance sheet? Is it fundamentally about power — who governs the issuance and use of money?

Neha: Yes, exactly. If it is backed by Treasuries, it is issued by the Treasury, operating in concert with the Federal Reserve. Treasuries are managed through the New York Fed’s Market Desk. But the key question is: who decides how much is issued into the system? And most importantly, what happens when things fail? That is what gives money its value — the trust that you will have something of worth at the end of the day if institutions fail. That is the bankruptcy question. Can consumers and users get at something still valuable when that happens?

Nicolas: So we are prisoners of current categories. If you had to reinvent the wording, how would you describe the space?

Neha: CBDC has become a dirty word, so I would throw that one out. I was annoyed for a while by the conversation about tokenized versus account-based systems, but I am coming around on it — tokenized money versus non-tokenized money is a meaningful distinction. I think the more fundamental question is: how decentralized is the substrate? Does the system have the property of censorship resistance? Cash has that property. Tokenized money is supposed to have it, but if it is issued by Circle or Tether, it does not. Tether just froze 500 million dollars’ worth of USDT. So that is the real question. I am not sure I have the right vocabulary yet. It would take months, if not years, to really sort that out properly.

Marieke: This is very interesting, because Nicolas and I have been working with a concept of a tokenized money stack. I love the word you use — wrap — rather than stack. In a stack it feels layered. But it is actually that you wrap one thing within another. And ultimately, as you say, it comes back to bankruptcy: who is the lender of last resort?

You recently wrote a paper on the hidden plumbing of stablecoins. Walk us through what it found, and what could go wrong — especially in an international system where parties are operating under different rules.

Neha: The interoperability question is central. Within one blockchain ecosystem — say, the Ethereum ecosystem, using the EVM and ERC-20 tokens — you get interoperability almost for free. That is the whole point. You can use any ERC-20 token in Aave or Compound. It is the concept of money Legos.

It becomes much harder across different blockchain systems. Ethereum to Solana, for example, is traditionally done with bridges. Bridges have been a major source of hacks because they rely on multisig or trusted actors managing the movement of money. Anyone can start a bridge, and there have been many. But interoperability between chains is not solved.

In the traditional finance world, large institutions — the BIS, CLS, Swift — come together to manage interoperability through messaging standards and FX holdings. They have projects they are starting to productionize. It will be interesting to see how that develops.

As for the hidden plumbing paper: everyone was focused on reserve backing. The GENIUS Act is primarily focused on backing — are the reserves there, how are they accessed? But we thought there were many other risks that people were not thinking about, starting with the underlying blockchain infrastructure. Circle and Tether issue on dozens of different chains. The largest chains by volume are Tron and Binance Smart Chain, which most people have not heard of — not what you would call blue-chip chains like Ethereum. These are chains with fewer validators, less decentralisation, shorter track records, and weaker security. Blockchains are not a universal, well-understood substrate. They are all quite different.

Then there is the question of how stablecoin issuance actually works. Not everyone can redeem directly with a stablecoin issuer. Most people have to sell through a cryptocurrency exchange, which introduces additional risk for maintaining par value. When Silicon Valley Bank failed and Circle had three billion dollars there, Circle resolved all redemptions at one dollar — for the institutions allowed to redeem directly with Circle. But ordinary users had to sell USDC on a cryptocurrency exchange, and on exchanges it broke par. That is an important structural risk that is not widely understood.

The third element is the structure of the Treasury market itself. Even if you are redeeming with a stablecoin issuer, can they actually sell the underlying Treasuries quickly enough to give you commercial bank account money? There are weaknesses there. I am not sure these risks are understood by the people writing or voting on the laws.

Nicolas: What is the systemic reason why ordinary users cannot redeem directly with issuers?

Neha: From the issuer’s perspective, it is overhead. They have to KYC everyone, maintain business relationships, manage all those accounts. Circle typically only redeems with a set of institutional customers; Tether with an even smaller set. Interestingly, under MiCA in Europe, stablecoin issuers are required to offer redemption to all customers. I am not sure how that is playing out in practice, but it is a significant development.

Marieke: On the Clarity Act and the question of yield on stablecoins — within this stack there are many actors participating in the stablecoin ecosystem, and the yield can be distributed by different companies at different levels. Are banks right to be worried about stablecoins?

Neha: Banks are right to be worried broadly. Coinbase would love to have JP Morgan’s business — that is very clear. But should they specifically be worried about yield? I am not 100% sure the arguments they are presenting are entirely valid. I am not sure stablecoins take money out of the banking system in a material way. They may take customers from the banking system, which is different. This is really a question about distribution and who owns the relationship with the end customer. Smaller and mid-tier banks are most concerned about interchange fees. If a large share of payments move to stablecoins, they lose on interchange. Yield is the first thing people have identified, but there are other aspects of their business that will be eaten away. Yield is essentially a customer acquisition tool — the incentive that draws people in and determines where they go.

“Encryption can be used to increase freedom or to reduce it.”

Read more

Sir Howard Davies has spent fifty years at the intersection of finance, regulation, and public policy — long enough to have helped build the architecture he later had to defend, and to have watched it buckle.

He started as a diplomat, posted to Paris in the early 1970s, where an encounter with the French administrative elite convinced him that the British habit of treating a university degree as a lifetime qualification was not going to serve him well. He left the Foreign Office for the Treasury, went to Stanford for an MBA, and returned to Whitehall just in time to work on both the nationalisation and the privatisation of the same industries — a coincidence that turned out, as he explains, to require exactly the same skills.

What followed was one of the more varied careers in British public life: McKinsey, the CBI, the Bank of England as deputy governor, and then the FSA, which he chaired from its creation in 1997 until 2003. He was there, in other words, when the regulatory framework that would be blamed for the 2008 crisis was being built — and has had to live with the accusations ever since. His view on the “light-touch regulation” charge is direct and, on the evidence, largely correct.

Since then he has led the LSE, chaired NatWest through the post-pandemic recovery, taught financial regulation at Sciences Po Paris, written several books on economics and central banking, and taken on a number of board roles — including, most recently, the chair of Qivalis, a consortium of European banks building a regulated euro stablecoin.

In this conversation, Howard walks through the origins and failures of the Basel capital framework, explains why the 2008 crisis was fundamentally about inadequate capital rather than regulatory culture, and makes the case — carefully — for what regulators got wrong and what they did not. He discusses Brexit as a strategic miscalculation, assesses what two decades of advising Chinese financial regulators has taught him about how Beijing thinks about global rules, and sets out why he believes Europe needs a bank-issued euro stablecoin before dollar-denominated alternatives make the question moot.

His new book, Global Financial Regulation: An Opinionated Guide, is out from Polity Press in the summer. You can find his Project Syndicate columns here, and he is active on X at @HowardJDavies.

Marieke: Hi, I’m Marieke Flament. This is the Currency of Power podcast, the companion to the Currency of Power newsletter, where we dig into currencies, power, and the people connecting the two. Today’s guest needs little introduction, but the breadth and depth of his career demands one anyway.

Sir Howard Davies has spent 50 years at the intersection of finance, regulation, and public policy — from the Foreign Office and HM Treasury to McKinsey, the Confederation of British Industry, the Bank of England, and the Financial Services Authority. He led the London School of Economics through a turbulent decade, chaired NatWest Group through the post-pandemic recovery, and today, among several other things, chairs Qivalis, one of Europe’s most serious efforts to build a regulated euro stablecoin infrastructure.

Sir Howard is also a professor of practice at Sciences Po Paris, where he teaches master’s courses on financial regulation and central banking, and is the author of several books on economics, finance, and regulation, with a new book coming shortly.

Howard and I go back some years. We worked together when he chaired NatWest and I was building Mettle, and we are working together again at Qivalis. Few people have seen more of the machinery of global finance from the inside. Today’s conversation runs approximately two hours. The first hour is open to all listeners; the second — going deeper on China, digital assets, and what comes next — is for paid subscribers. We will cover four areas: Howard’s career and what it teaches us about how institutions actually work; the current geopolitical moment and what it means for markets; his unusually direct perspective on China; and finally stablecoins, CBDCs, and the future of money. Howard, welcome.

Howard: Thank you. That introduction makes me rather exhausted just thinking about all those jobs.

“It was a deliberate, French-driven strategy”

Marieke: Me too! Before getting into the substance — 50 years across so many institutions and geographies, and a career that looks less like a straight line than a series of reinventions. What has been the through line? How did you navigate all of that?

Howard: It all started, interestingly, in France. I began as a classic British amateur. In the UK at the time, a university degree was regarded as a kind of vaccination against needing to learn anything further for the rest of your life. So most people did a degree at the best university they could get into — in my case Oxford — and then entered whatever career they chose as a complete amateur.

I chose the Foreign Office, which seemed an interesting option. Like most young people, I thought travel would be fun. When I was sent to my first overseas posting in Paris, I found myself dealing at length with people in the French system who were, at that time, undoubtedly better trained and better educated than we were. You did your undergraduate degree, or one of the Grandes Écoles, then ENA, and you came out knowing a great deal about public administration, economics, even philosophy. In the UK, that was regarded as of no interest whatsoever — you picked it up as you went along. I found myself somewhat intimidated. Not on language, because I had done a year as an assistant in a French lycée and could hold my own at the Quai d’Orsay. It was simply that they had so much more educational background.

That observation combined with what I could see happening in Britain at the time. We had a Labour government re-elected in 1974, and it seemed the government would integrate forward into more and more areas of economic life through a gradual process of nationalisation — sometimes ideologically driven, sometimes simply because companies were failing and needed rescuing, as with British Aerospace and Rolls-Royce.

I thought the French had the right idea. What would be interesting for a future career was to work at the interface between government and the private sector — someone who understood how government worked but also understood something about business. So I decided to leave the Foreign Office for the Treasury, which was surprisingly prepared to accept me despite my lack of economics training. I also told the Treasury I wanted to go to business school in the US, get a proper business qualification, and then work largely in the public sector. It seemed like a good niche.

The problem was that by the time I finished business school, Mrs Thatcher had been elected. Far from the state integrating forward, the opposite was happening — more and more areas were being privatised. What turned out, though, was that although my personal career strategy was built on a completely wrong assumption, it remained perfectly viable for the opposite reason. There were now more and more areas of the private sector that depended quite heavily on government policy. It is fine to privatise a formerly nationalised industry, but the idea that it then operates entirely without government involvement is wrong. So the strategy proved usable in different ways. I went back to the Treasury after business school, working first on nationalising companies and then on privatising them — and actually the same skills are required.

It was a deliberate, French-driven strategy, motivated by a feeling of embarrassment at being less well educated than the French officials I met. There is nothing that motivates an English person more than feeling the French have got one over on them. That was quite a useful emotional dynamic.

Nicolas: For the record, I went to ENA, and the students there enjoy looking down on the school and explaining to everyone that you learn nothing there. So we may have a different perception — but perhaps things were different then.

Howard: They certainly had a good way of talking about it.

Marieke: That does sound very French. Now, looking at everything you do today — Sciences Po, boards, books — you are incredibly active. What is your filter? How do you prioritize?

Howard: It looks complicated when you are running a portfolio — I chair three companies plus an arts organisation, an orchestra, do some teaching and a few advisory things. But reflecting on it, it is not actually that different from being chief executive of a large organisation. If you are running a university, or chairing a bank, or running a big regulator, you always have to be very selective about where you intervene. There were all kinds of parts of NatWest I could have involved myself in, but would I have added value? In most cases, probably not. The mortgage business was well managed; the people knew what they were doing. A new venture with a complicated regulatory dimension like Mettle — that was somewhere I could contribute.

Even in a single job, you are constantly making choices about where to intervene and where not to. A portfolio is not so different. I do think about it rather like a children’s train set — you have a series of trains and you want to push them all forward. I usually take stock on a Thursday: have I done anything on each of these this week? Is there a call I should make, something I should read, someone I should chase up? It is not so different from running one large organisation. I was chief executive of large organisations for about 25 years, and you make those choices constantly.

“There are no prizes for running a light-touch regime”

Marieke: That is interesting — you can apply OKR-type thinking to yourself, checking each week whether you have moved things forward. Should we go into the substance? The first area is the 2008 crisis. You were the FSA’s first executive chairman from 1997 to 2003, and then the crisis happened. The post-mortems were quite critical of regulators. Is that fair?

Howard: Broadly, it is not fair in most senses, though there are one or two respects where it is. The subsequent analysis in the UK tended to focus on the accusation that the regulator had operated a light-touch regime and that this was a mistake. I can get quite exercised about this, but I will try to stay calm.

I actually confirmed via ChatGPT yesterday that there is no speech, no document, nothing in which I referred to a light-touch regime while at the FSA.

What happened was that when the government changed in 2010, the Conservatives under George Osborne had a strong motive to blame the financial crisis on the previous Labour government. Osborne argued it was all caused by Gordon Brown, who had changed the regulatory system and introduced a single regulator operating a light-touch regime. The fact that the regulator had never mentioned a light-touch regime was not regarded as a relevant detail. It is true that one or two Labour ministers did use the phrase, which made us very nervous at the FSA — there are no prizes for running a light-touch regime, because the market will always disappoint you if you do.

So I reject the notion that it was our strategy, and there is no evidence that it was. But even setting the words aside, you have to disentangle what a light-touch regime could actually mean. Regulation falls into two broad boxes: prudential — the capital regime, how much capital banks were required to hold — and conduct of business.

On the first, the FSA rigorously applied the globally agreed capital regime. There is no evidence that UK banks held less capital than Basel rules required, or less than comparable institutions elsewhere. The problem was that because our economy is heavily dependent on financial services, when financial markets catch a cold, the UK gets influenza — it is simply a larger share of the economy. Some of our banks were badly affected, but the same was true of other large financial centres. In the Netherlands, practically the whole banking system went bust: ABN AMRO, Fortis — more or less everything. So on the substance of light touch in capital terms, no, absolutely not.

On conduct of business, I think with some reflection that we were probably not as aggressive on enforcement as we might have been. I would not say that caused the financial crisis — I do not think mis-selling of interest rate swaps to small businesses, or LIBOR manipulation, brought the system down. But some of those practices did not look pretty after the event.

Insofar as I would accept any guilty charge: not at all on light touch, not at all on capital. The global capital rules were too weak — we know that now — but that was not something the FSA could unilaterally change. Had we said in 2001 that an 8% minimum was wrong and we were going to impose 15%, the reaction would have been hostile. You could not have done it unilaterally because of the level playing field. The global regime was too weak, but it was applied in the UK just as it was everywhere else.

On policing conduct in the wholesale markets — yes, in retrospect, we probably should have been tougher. Some of the practices in securitisation markets were pretty dubious, and the regulators should have been harder on them.

Nicolas: What is your view on how regulation was handled in the US, and the extent to which the US was influencing the approach in other countries?

Howard: For most commercial banks in the US, the rules were being applied very similarly to the UK and the rest of Europe. The exception was the large broker-dealers. Goldman Sachs, Merrill Lynch, Morgan Stanley, Bear Stearns and the rest were regulated by the SEC, which did not have the same approach to capital or the same understanding of it. Those institutions were not regulated appropriately.

Very quickly during the crisis, the US authorities acted to make them banks. Several went bust in the process — Merrill Lynch was effectively rescued by Bank of America, Bear Stearns went bust, Lehman Brothers went bust; Morgan Stanley and Goldman survived. The SEC regulated five major broker-dealers and three went bankrupt. A 60% failure rate for a regulator is quite high. That part of US regulation was definitely flawed, and it was corrected rapidly in the autumn of 2008 when the survivors were forced to become banks regulated by the Fed, putting them broadly on a level playing field with everyone else.

Marieke: On Basel — my understanding is that it emerged post-war as something like a global framework for what bank capital should look like. But it has been loosely followed by the US and more stringently implemented elsewhere. Did that play a role? What is the history of Basel and how consistently is it actually followed?

Howard: The Basel Committee has only existed since the 1980s. It was essentially an Anglo-Dutch creation. At the time, the assumption was that an 8% capital ratio would guard against any but the worst financial downturn.

Basel I was written on a postcard, effectively — you must hold at least 8% capital. The problem was that it was not risk-sensitive. You applied the same 8% whether you were lending to the US government by buying Treasuries or lending to a speculative high-risk mortgage insurer. That was not good enough, so Basel II attempted to correct it.

Interestingly, Basel II was constructed on the assumption that 8% was the right number — it just needed to be redistributed, with lower capital against sovereign lending and higher capital against speculative mortgages. The absolute level was not questioned. In the run-up to the crisis, markets had actually become more volatile, so arguably you needed more than 8% as a baseline — but that was never discussed. Basel II was about redistribution, not about the level.

When the crisis hit, it was obvious the absolute level was too low. The downturn was enough to wipe out the capital of some very large banks — Citibank, RBS, Fortis, Commerzbank — all of which had to be rescued in various ways. Basel was a sound intellectual framework, but the numbers were too small. When Basel III came in, the three-pillar framework was retained because it is intellectually coherent, but the numbers were ratcheted up significantly. A bank like NatWest, on a comparable basis, now holds four times as much capital as it did under Basel I. That is the scale of the additional capital that was needed.

“The numbers were simply too small”

Nicolas: Can we turn to what you wrote in your book Who is to blame about the crisis — your assessment of what went wrong and who bore responsibility? Was it the system, the behaviour of certain players, or something else?

Howard: The number one point is the one I have just made: the inadequacy of the capital base. The amount of capital banks were required to hold was no longer sufficient given the volatility in markets.

Second, regulators had not been good at policing what I would call the regulatory frontier. There were many ways in which banks could effectively increase their financial leverage — through special purpose vehicles, off-balance-sheet structures of various kinds — where the losses, when they crystallised, bounced back onto the core bank’s balance sheet. That was true of much of the securitisation in the subprime market and elsewhere. So the absolute rules were too low, and there were also technically permissible ways of circumventing them which exaggerated the degree to which capital was inadequate when things went wrong.

My own view is that the conduct-of-business failings — the mis-selling, the LIBOR manipulation — were not core to the financial crisis. I do not think they would in themselves have brought the system down. It was the capital. The assumption was that banks would have sufficient capital to absorb market losses and that, if necessary, central banks could step in.

This links to a debate that has gone on for 20 years between what are called leaners and cleaners. Until the crisis, the dominant view — articulated most persuasively by Alan Greenspan — was that central banks should not interfere too closely in markets, because that would damage their dynamism and wealth-creating properties. Instead, they should stand ready to clean up when things went wrong, providing liquidity to stabilise markets during a panic.

There was another current of opinion, mainly from the BIS, arguing that central banks should lean into these cycles — that when bubbles are inflating in asset markets, you should be prepared to say so and take action: tightening loan-to-value requirements on mortgages, restricting margin lending, requiring higher capital buffers. This is now called counter-cyclical regulation.

The crisis settled that debate largely in favour of the leaners. It became clear that even the Fed announcing it would provide liquidity was not enough to stop a panic. The ECB provided liquidity and banks still went bust. So the notion that you should be prepared to intervene when a bubble is inflating is one that most central bankers would now accept, at least in theory. You can see it today in Andrew Bailey warning about a private credit bubble in his role as chairman of the Financial Stability Board. That kind of forward-leaning statement would not have happened under Greenspan.

Nicolas: Are there recent examples of leaners actually acting and reigning in a bubble, or is this still largely theoretical?

Howard: There are practical examples. In the UK, for instance, a few years ago the authorities imposed penal capital requirements on mortgages above 70–75% of property value, and they adjusted those requirements up and down as conditions changed. The Chinese have always done something similar, and quite sensibly.

Counter-cyclical capital buffers are another example — where regulators require banks to hold an additional buffer, say 2%, across the board when markets are running ahead, not based on analysis of individual loans but as a general precautionary measure. Several countries have used these, moving them up and down with conditions. In the UK we had them at 2%, then removed them during COVID to avoid constraining lending during the recovery. So there are plenty of practical examples of what we now call macro-prudential regulation — a term that was barely used in the run-up to the last crisis.

Marieke: Is China the ultimate leaner? It sounds as though they have always done this in a more proactive way.

Howard: You have to distinguish between the broad concept of leaning and the micro-regulation of many individual markets, which the Chinese have gone into extensively. Some of that is for financial stability reasons; some is simply because they want to control property prices in coastal cities that they think are getting out of hand. I do not think that is quite the same as an overall judgment that credit expansion is too rapid and asset prices are generally too high. Some of the Chinese intervention is very micro in focus.

Marieke: You have written about this period, and in your book The Chancellors on Alistair Darling you describe him as a man who dealt a genuinely terrible hand who played it extraordinarily well. What does that tell us about leadership in a crisis?

Howard: The book was the second in a series — whether I live long enough to produce volume three, I am not quite sure. When I was at the LSE, I edited a book called Chancellor’s Tales. I persuaded all the then-living chancellors, going back to Denis Healey, to come to the school and give a lecture not on the politics of each individual budget, but on what it was like to run the Treasury and the British economy during their time. To my surprise, they all agreed, and most produced quite interesting lectures. I edited those into a book with a long thematic introduction covering 1964 to 1979.

During COVID, I thought it a good use of time to repeat the exercise for the subsequent generation of chancellors. I interviewed them all — with one exception — and used those interviews as the basis for a narrative about managing the British economy through that period. I wanted them to tell their own stories about why they made the key decisions they made, and also how it felt to be in charge.

With Alistair Darling, the particular interest was that he had come in as Chancellor when Blair left Number 10 in the summer of 2007 and Brown took over — having previously been Secretary of State for Transport, so with no responsibility for any of what followed — and then the world fell apart. What struck me about his approach was his composure. At that moment, everybody was losing their heads. Brilliant financiers being paid millions of pounds a year were panicking and demanding to be bailed out in every conceivable way. To steer through that required personal solidity more than mountains of analysis. The analytical models were essentially useless — the then-CRO of Goldman Sachs famously remarked that their models told them the market movements of the previous week should occur once in a million years, and they had just happened again. In those circumstances, what you needed was someone calm, prepared to listen, and able to communicate that there was a grown-up in the room trying to maintain control. It was a case where personality mattered as much as policy. Those who worked directly for Darling at the time would validate that assessment.

Marieke: The skill of not making a crisis worse by panicking yourself.

Howard: Exactly. If you can at least do that, you have made a start.

“London was on its way to becoming the onshore financial market of Europe”

Nicolas: Shall we move to the current geopolitical context? A significant thing happened in Britain a few years ago — Brexit, the decision to leave the European Union. It was widely assumed this would give Britain more flexibility to reposition in a changing world. After several years, it is fair to say that has not clearly happened, and it is still not obvious what Britain’s strategic goals are. You have called Brexit a significant mistake. Can you elaborate? Why was it a mistake, and was there anything that could have been done to make more of it?

Howard: The effects of Brexit have been less bad in some respects and less good in others than was widely expected.

In the period running up to Brexit, London had been the great beneficiary of the growth and deepening of the single market, including the single financial market — however incomplete that remains. London was moving from being the offshore financial market for Europe to being effectively the onshore financial market as well. A striking feature was that you would find Italian investment bankers selling Italian corporate debt to Italian investment managers — in London. All these people had families in Milan and worked for Milan-based institutions, but the business was actually being done in London. There is what economists call an agglomeration effect in financial markets, which is somewhat circular: people are there because other people are there, and that reinforces itself. Brexit was a blow to that, and there has been some repatriation of business to domestic markets. EY has tracked job losses, which reached about 7,000 and seem to have stabilised there. That is not enormous — BNP Paribas, for instance, says it employs more people in London today than before Brexit. But there has been a cost.

And looking forward, the push towards a savings and investment union and deeper capital market integration in Europe is exactly the kind of development London would have been well placed to benefit from. As things stand, insofar as a functioning capital markets union exists in Europe, it is in London — and we are no longer positioned to capture that growth in the same way.

On the other side of the ledger, is there anything Britain has been able to do with its regulatory freedom that it could not have done inside the EU? At one level, any pan-European regulation is a compromise among 28 countries, so any individual country, given a free choice, would probably design some rules differently. There have been some examples — most visibly in insurance, where changes to the Solvency II regime in London have been beneficial for the wholesale markets, and which would have been harder to achieve within the EU.

But as a practical matter, when you ask the large American firms whether they want UK regulation to diverge from European regulation, most say not really. The benefits of a slightly more market-friendly regime may be offset by the complications of running two different regulatory regimes within the same institution. Most would prefer consistency.

There is also a deeper point. The EU capital regime starts in Basel — the rules are devised there and then converted into binding legislation in Europe, whereas the US treats them more as best-endeavours guidelines. Why does the EU legislate them rigidly? Because in a single financial market, the French regulator is very interested in what the Latvian regulator does. A Latvian bank can set up in the Faubourg Saint-Honoré and sell into the French market without further restriction if it holds a Latvian licence. So you cannot simply rely on mutual recognition — you need the same capital definition, the same capital levels, because otherwise you are being undercut and you cannot stop it. The rigidity is the price of the open market.

So the degrees of freedom Brexit provided have been fairly limited, and most of the apparent deregulation in the UK is not actually Brexit-related. A lot of it involves reducing requirements that were self-imposed and went beyond anything the EU required — like the Senior Managers and Certification Regime, which codifies individual accountability in detail that has no equivalent in any European directive. We introduced that after the crisis because when bankers were summoned to parliament and asked who was responsible for a bank going bust, they all pointed at someone else. So we codified responsibilities. That regime has since been lightened — but that was something we did to ourselves, and now we are partly undoing it. It has nothing to do with Brexit.

Marieke: I lived what you are describing when I worked at Circle, based in London leading the Europe operations. When Brexit happened, the FCA was actually quite advanced in its approach to crypto regulation — genuinely world-leading. Brexit put the brakes on that because the authority had to redirect its attention. And the loss of passporting rights was significant: previously we could passport a licence across the EU from London. After Brexit, we had to obtain licences on the continent and move teams accordingly. Exactly the consequences you describe, experienced at a small scale.

“All their wives were in Milan and the business was being done in London”

Nicolas: You described the agglomeration effect and London’s role as the hub that made the concept of a single European financial market a practical reality. Does losing London mean Europe has lost its ability to deliver on that promise to investors? It is very hard for Frankfurt or Paris or Milan to replicate what London offered.

Read more

Dollar stablecoins are expanding global dollar dominance outside the traditional banking system. Existing monetary infrastructure wasn’t built to handle such an extension, and every jurisdiction is now being forced to respond to it.

This week offers an illustration of the tensions that stablecoins are producing across the global monetary order. America is pushing ahead its stablecoin agenda: the Clarity Act is moving through Congress, with a fight over yield at its centre, while USDC is being embedded into the infrastructure of the emerging agentic economy. Meanwhile Brazil has banned regulated fintechs and eFX providers from using stablecoins or any cryptocurrency to settle cross-border payments. And in Europe, Christine Lagarde argued on Friday that stablecoins are not the path to a stronger euro. In her view, the right response is to build public infrastructure and wait for it to be ready.

Three jurisdictions, one underlying problem, three entirely different responses. And as this edition will argue, only one of them is playing offense.

The Yield Question

The Clarity Act is doing exactly what it says. It is bringing regulatory clarity — to the US. The problem is that in a world of borderless digital money, a domestic clarification can have consequences that are anything but domestic.

The provision drawing the most attention is the yield question: within the stablecoin supply chain, who is permitted to offer yield to end users?

The debate has been framed, predictably, as a fight between the crypto industry and the banks. And the banks have a coherent argument. In the traditional system, yield is the mechanism central banks use to direct the flow of capital. Low rates push money into markets. High rates pull it back into deposits. That mechanism is how inflation gets managed, how growth gets throttled — and how banks keep their balance sheets funded. If regulated stablecoins begin offering competitive yield to retail customers, the concern is that bank deposits drain toward a new and less-regulated alternative.

Several analyses have complicated that argument. In particular, the evidence for mass deposit flight is thinner than the lobbying suggests. That said, the underlying anxiety is undeniable, suggesting it maps to something deeper than deposit economics. The real fight, in reality, is about who gets to own the customer relationship in an era when money has gone digital. Banks are worried about losing both deposits and relevance from a customer’s perspective.

To understand what is at stake, it helps to get back to the basics. The short version goes like this: yield is a customer acquisition cost dressed up as monetary policy.

More specifically, in any competitive system, yield is what you offer someone to say: stay with me. It is the reward for holding. And in the stablecoin ecosystem, the economics of that reward are simply different. Issuers like Circle and Tether hold Treasury T-bills at around 5%. They pass little (in the case of Circle) to nothing (in the case of Tether) back to the user, because they need to retain most or all of the yield for themselves: the margin is the business, and the Genius Act was drafted based on this premise.

That said, the debate now happening in Washington is whether that should change — whether yield should flow downstream, and if so, through whom. In DeFi, stablecoins can already generate returns of 3–12% APY depending on platform, protocol, and risk profile. Platforms like Kraken, Aave, and Coinbase all offer variations of this. The higher the yield, the higher the risk embedded in the structure underneath it. This is not new. What is new is that the same question is now being clarified at the level of statute — and the answer will not stay within US borders.

What Washington Decides, the World Inherits

Here is the scenario that the Clarity Act debate has largely failed to reckon with — at least outside of the US. Whatever gets legislated in Washington applies to US-regulated entities serving US customers. It does not automatically apply to a European holding USDC in a self-custodied wallet. If, as a result of the Clarity Act, a US citizen cannot earn yield on their stablecoin holdings, but a European citizen holding the same asset through a different interface can, the resulting incentive structure can prove problematic.

For a European holder, the decision calculus becomes: convert euros into USDC, earn a materially higher yield, absorb the FX risk. Sometimes that trade makes no sense. Sometimes it does. But in aggregate, even partial adoption of that logic creates structural pressure toward dollarisation operating inside European borders, driven not by policy or coercion but by individual wallet decisions. Capital does not wait for political permission to move to where it is treated best.

Currencies are now competing for retail holders at a scale and with an accessibility they never had before. You can download a wallet, receive USDT, and transact internationally without touching a bank or a regulated institution. The dollar’s deep liquidity and global trust make USD stablecoins the dominant form of that competition. And yield is the marketing tool that can convert a passive holder into an active advocate.

In other words, we are witnessing a global competition for wallet share — where yield is the incentive layer that converts a passive holder into an active advocate.

Brazil Draws the Line

The most instructive defensive response to this dynamic is in Brasília.

Last week, the Banco Central do Brasil published Resolution No. 561. The rule bans regulated fintechs and eFX providers from settling cross-border payments using stablecoins or any other cryptocurrency. From October 1, companies like Nomad and Braza Bank — which had built USDT and XRP-ledger settlement into their cross-border rails — must route everything through traditional FX transactions.

The numbers help explain the decision. Brazil’s traditional remittance market moves roughly $25B per year. But crypto had quietly assembled something considerably larger alongside the regulated system. Brazil’s crypto market now processes $6–8B per month, with stablecoins accounting for approximately 90% of that volume. Annualised, that is $70–90B in cross-border flows operating through channels the central bank cannot observe — a parallel financial infrastructure, built without permission, larger than the regulated market it nominally complemented, and almost entirely invisible to the BCB. At that scale, the regulatory blind spot had become untenable.

The BCB’s rule operates cleanly at the institutional layer. Licensed fintechs have authorisations to protect, thus compliance will happen. What the rule cannot reach, however, is the peer-to-peer layer. Any Brazilian with a self-custodied wallet can still settle in USDT on Tron or USDC on Solana, entirely outside the eFX framework. The plumbing the regulation controls is not the same as the plumbing that actually moves the money.

Europe’s Answer: Build the Plumbing, Skip the Token

Read more

Everyone has been watching the ports and the railways. That was the wrong thing to watch.

China has quietly become the world’s largest official creditor. According to AidData’s dataset published in November last year — the most comprehensive analysis of Chinese overseas lending ever assembled — China’s lending portfolio has reached $2.1 trillion across 217 countries and territories. For every dollar it donates, it lends 35. To developed countries, the ratio is 761 to one. The share of its overseas portfolio that qualifies as development aid has fallen from 22% in 2000 to 1% in 2023. China’s overseas lending is the most aggressive return-seeking international lending operation in history — run by a nominally communist state using the capitalist playbook of investment, and then some.

That reframing is refreshing for everyone interested in China’s global soft power. Until recently, the debate about China’s overseas ambitions has been almost entirely focused on the Belt and Road Initiative (BRI) — the infrastructure programme launched under Xi Jinping in 2013, which built ports, roads, and railways across the developing world.

For years, critics have been calling the BRI debt-trap diplomacy: fragile states lured into loans they could not service, then forced to hand over strategic assets. The Hambantota port in Sri Lanka, leased to China for 99 years after Colombo could not service its debt, became the emblem of this critique. But the debt-trap framing, always contested, has been superseded by events. Infrastructure lending has fallen from 75% of China’s overseas portfolio in 2010 to less than 25% today. The BRI, in its original form, was phase one. Something more sophisticated seems to have replaced it by now.

To understand what China is actually building, it helps to start not with the BRI but with the dollar order it is trying to displace.

The Dollar Order as Alliance Coordination

The post-war dollar system is usually described in monetary terms: reserve currency, recycled surpluses, exorbitant privilege. But Izabella Kaminska, who joined us recently on Currency of Power, offers a different reading — one that clarifies what China is doing by contrast.

In Izabella’s view, the dollar system was designed to solve a structural problem in Western defense. Democracies are poor at sustained military coordination. They answer to electoral cycles and electorates that resist permanent sacrifice. The Soviet Union, as a command economy, faced none of those constraints. The dollar system solved this problem indirectly: by anchoring allied economies to American financial infrastructure, it allowed Washington to fund a global military posture without imposing the full cost on domestic taxpayers. Japan and Germany were rebuilt as export-led manufacturing powers. Their savings, recycled into US Treasuries, funded American security. The monetary arrangements served a strategic objective.

The dollar order was, in other words, an alliance coordination mechanism. The currency was the operating system. The goal was to implement “allied scale” — to pool and concentrate Western industrial and military capacity under a single umbrella, so that the democratic bloc could match the sustained resource commitment of a command economy.

Sinéad O’Sullivan, who spoke to us in an earlier edition, argues that China is itself a coordination system. Its central bank pursues social, employment, and political stability, deploying liquidity as a strategic instrument. The state planning apparatus decides which sectors grow and which absorb the costs of slowdown. That consistency — the ability to direct resources toward strategic objectives across decades — is something democratic market economies cannot match.

The US ran its system through a currency: countries that wanted access to American markets, American security, and American capital had to operate in dollars, which drew them into American-led institutions almost automatically. China has no currency with that kind of global reach, so it is building its network differently — through loans, infrastructure contracts, and increasingly energy systems that create direct dependencies on Chinese suppliers, Chinese financing, and Chinese technical standards. The monetary layer is coming, but it is being built on top of a physical and financial presence that is already extensive.

The New Playbook

AidData’s research makes clear that the BRI was never the whole story. For every four dollars China lends for infrastructure, it lends another six for activities that have nothing to do with infrastructure: corporate acquisitions, liquidity facilities, technology assets, critical minerals. By 2023, 76% of China’s overseas lending supported upper-middle and high-income countries. The single largest recipient of Chinese state credit is… America — which has received over $200 billion.

China’s playbook has several distinct components.

The first is the shift from aid to return-seeking capital. China has abandoned the role of a traditional donor. It lends on commercial or near-commercial terms, uses variable interest rates, requires collateral, and increasingly routes transactions through shell companies and offshore vehicles to limit scrutiny. Its cross-border acquisition lending — directed at technology assets, critical minerals, and strategic infrastructure — has an 81% success rate. Since the adoption of its Made in China 2025 industrial policy in 2015, the share of that acquisition lending targeting sectors that host countries have designated as sensitive on national security grounds has risen from 46% to 88%.

The second is the construction of a critical minerals supply chain. China has invested heavily across Africa, Southeast Asia, and Latin America to secure upstream extraction and midstream processing of the materials that clean energy and advanced technology require: copper, cobalt, nickel, lithium, rare earths. North American and European mining operations face a structural disadvantage: capital is available, but engineers and skilled workers are scarce. China built its position over decades. The West is now trying to replicate that in years.

The third — and the most consequential for the long run — is what we call the electrostate.

Read more

Regular readers will know that China’s monetary strategy has become something of an obsession at Currency of Power. Over the past few months, we’ve traced how the petrodollar recycling loop is breaking, mapped the great bifurcation between dollar and yuan financial infrastructure, argued that the dollar has more lives than its critics admit, and tried to understand what it actually means when a country is wide open to the world and locked into a single system at the same time. We’ve also had the pleasure of going deep on the People’s Bank of China with the brilliant Sinéad O’Sullivan, whose four-part series on Chinese central banking is essential reading for anyone trying to understand how China’s monetary system actually works.

This piece grew out of something more immediate. Marieke just returned from a trip across China and Japan — and came home with a head full of observations. One in particular stuck: in Hong Kong, mainland Chinese shoppers were flooding the malls. Not because of some new retail trend, but because the maths suddenly made sense. A weak dollar means a weak HKD — which is pegged to it — which means everything priced in Hong Kong dollars is cheap for anyone earning or saving in yuan. A Chinese city, running on American monetary policy, accidentally on sale to its own citizens.

That small observation is the thread we pulled. And it leads somewhere interesting.

A City In Between

Hong Kong has always been a city in between. Between East and West, between colony and sovereign territory, between market freedom and party control. Its monetary system is, in many ways, the purest expression of that in-betweenness: a Chinese city whose entire financial existence is anchored to the United States dollar.

That arrangement has made Hong Kong extraordinarily prosperous. It has also made it a strategic puzzle for Beijing. As China embarks on one of the most deliberate dedollarization campaigns in modern financial history — expanding yuan swap lines, promoting its Cross-Border Interbank Payment System (CIPS), rolling out the digital yuan, accepting yuan payments for oil — Hong Kong sits at the centre of a paradox. It is China’s most important gateway to global capital markets, and it runs on the currency of China’s primary geopolitical rival: US dollars.

How can China use its most valuable financial bridge, Hong Kong, to escape the very currency that keeps that bridge standing? Answering that question sits at the intersection of monetary history, geopolitical competition, and the coming war over who controls the digital infrastructure of global finance.

Is Hong Kong China’s ultimate joker’s card when it comes to full sovereignty and de-dollarization or its weakest link?

From Crown Colony to Currency Board

To understand Hong Kong‘s monetary situation today requires understanding how it came to be tied to British sterling, then to the US dollar, and the political tremors that drove each transition.

Hong Kong was ceded to Britain in 1842 following the First Opium War, and for most of its colonial history its currency drifted between various anchors. After moving off sterling when the pound began its long post-war decline, the territory briefly floated its dollar in 1974. The experiment went badly. Inflation surged, speculative pressure mounted, and the economy lurched.

The decisive moment came in 1983. The Sino-British negotiations over Hong Kong’s post-1997 fate — the year Britain’s lease on the New Territories was due to expire and sovereignty over the whole territory would revert to China — had begun in earnest, and confidence collapsed spectacularly. On September 24 of that year, later dubbed “Black Saturday,” the Hong Kong dollar was devalued by 15% over two days to a historical low of HK$9.60 to the US dollar. Shops began quoting prices in US dollars and refused to accept local banknotes. There were runs on food staples. It was a currency crisis born entirely of political fear: investors had no idea what Hong Kong would look like under Chinese rule, and they voted with their capital.

The response was the Linked Exchange Rate System (LERS), announced on October 17, 1983, fixing the Hong Kong dollar at HK$7.80 to the US dollar through a currency board mechanism. The architect was economist John Greenwood, who proposed a simple, rule-based system in which every Hong Kong dollar in circulation would be fully backed by US dollar reserves held at the Exchange Fund. There would be no discretion, no central bank rate-setting, no monetary policy independent of the Federal Reserve. Credibility through constraint.

The peg survived the handover to China in 1997, the Asian financial crisis the same year, the SARS outbreak of 2003, the global financial crisis of 2008, the social unrest of 2019, and the COVID-19 pandemic. Since 2005, the HKMA has maintained the rate within a band of HK$7.75–7.85 per US dollar. The system has been a remarkable feat of institutional endurance — and one of the most transparent monetary frameworks in the world. Today, Hong Kong’s entire monetary base remains fully backed by US dollar assets held in the Exchange Fund.

The Post-1997 Paradox

Read more

The Trump administration has made a simple story the centrepiece of its economic worldview: America has been cheated. Foreign countries sell into American markets, accumulate dollars, and use those dollars to lend back to Washington at interest. Trade deficits, in this telling, are a measure of exploitation. The rest of the world gets factories, jobs, and growth. America gets debt.

It is a story with genuine emotional and political force. It is also, in important ways, backwards.

For decades, the US has occupied a position in the global economy that no other country has ever held for so long: issuer of the world’s reserve currency. That status, critics argue, allows Washington to consume more than it produces, borrow without limit, and export its deficits to everyone else. Far from being the victim of a rigged system, the US built the system, and has been its primary beneficiary. This is what Valéry Giscard d’Estaing, then France’s finance minister, meant in the 1960s when he coined the phrase “exorbitant privilege”.

The situation, however, is more complicated. The economists who have studied this most carefully — Paul Krugman, Hélène Rey, and others — reach a more qualified conclusion: the privilege exists, but its economic value is probably smaller than either its champions or its critics claim. In a recent post on X, Izabella Kaminska of The Blind Spot goes even further: the dollar system was never primarily designed to let America live beyond its means. It was designed to keep the West’s military and industrial capacity coordinated under American leadership.

So why is America now stating that it’s being taken advantage of? And why is the rest of the world under the impression that the US is abusing its exorbitant privilege? This, as can be expected, is a long story—and a sign that something has gone seriously wrong.

The Architecture of Dollar Dominance

To understand any of this, it helps to be clear about what reserve currency status actually means — and why it is so durable.

The dollar is not dominant because of a law or a treaty. There is no world government requiring Brazilian importers to pay for Malay goods in US currency. The dollar dominates for the same reason English dominates international commerce: because everyone else uses it. Paul Krugman, drawing on a 1967 paper by Charles Kindleberger, has argued that this analogy is more than illustrative. It captures the deep structural logic of currency dominance. Just as it invoices in dollars, a Brazilian firm negotiating with a Malaysian counterpart conducts the conversation in English, not because either party chose America, but because those are the languages — linguistic and monetary — that minimise friction when strangers trade.

Data shows that this self-reinforcing network effect is remarkably stable. As pointed out by Paul Krugman, the dollar is involved in 88 per cent of all foreign exchange transactions, dominates global trade invoicing outside Europe, accounts for nearly 60 per cent of central bank reserves, and underpins some $20 trillion in foreign-held dollar assets. The euro is a distant second. The yuan barely registers.

The roots of this dominance lie in events that happened at the end of World War II. The Bretton Woods conference of 1944 placed the dollar at the centre of a new international monetary order, backed by American gold reserves and the overwhelming productive capacity of an economy that had emerged from the war intact while Europe lay in ruins. When that gold-backed system collapsed in 1971, the dollar’s dominance survived through a different mechanism: the petrodollar agreement of 1974, under which Saudi Arabia and OPEC agreed to price oil exclusively in dollars in exchange for American security guarantees. Any nation that needed energy — which meant every nation — needed dollars. Those dollars were recycled into US Treasury bonds, creating a permanent captive demand for American debt.

That petrodollar recycling loop, however, is now fading. Izabella argued, presciently, as early as 2015 — when it was unfashionable to say so — that the petrodollar effectively died the moment America became energy-independent through shale. As for Brad Setser, he recently showed that petrodollars now play a marginal role in funding America’s national debt. In other words, the structural demand for dollars anchored to oil has been in decline for a decade. What has replaced it, at least in part, is what Izabella calls the “sweat dollar”: the dollars that flow outward to pay for imports from countries whose comparative advantage is cheap labor, most notably China. And beyond that, our theory at Currency of Power is that a new digital recycling loop is being engineered, anchored not in oil but in artificial intelligence infrastructure (GPU dollars) and dollar-backed stablecoins (cryptodollars).

The architecture changes. The dominance persists.

Exorbitant Privilege: Real, but Qualified

So does this dominance let America live at the world’s expense?

The intuitive case for “yes” runs as follows. To supply the world with dollar liquidity, the US must run persistent trade deficits — it must import more than it exports, pushing dollars outward while absorbing foreign goods. The world, in turn, recycles those dollars into US Treasuries, lending Washington the money to fund its deficits at low rates. America consumes; the world produces; and the bill is deferred indefinitely onto a growing stack of sovereign debt now approaching 122 per cent of GDP—even more when you account for private debt as well.

This is a phenomenon known as the Triffin dilemma: the very mechanism that makes the dollar indispensable to the world hollows out American manufacturing, as a persistently strong currency makes US exports uncompetitive. The US gets cheap imports, cheap borrowing, and a bloated financial sector. It loses factories, and the communities that depended on them.

Hélène Rey and Pierre-Olivier Gourinchas have documented what they call the “hedge fund nation” dimension of this privilege: the US issues low-yielding safe assets (Treasuries) and invests the proceeds in higher-yielding foreign equities and direct investments, earning a structural excess return of roughly 1.5 per cent per year in real terms. America, in other words, borrows cheap and invests dear — a carry trade run at national scale.

But Paul Krugman urges caution about overstating all this. He points out that the notion of a “unique” American ability to run sustained trade deficits is simply false: Australia ran massive current account deficits for three decades without holding reserve currency status; the UK has done the same since 2000. The ability to borrow from global markets is not exclusive to the dollar’s issuer. Paul also notes that Ben Bernanke, examining whether dollar status lowers US borrowing costs, found no clear evidence of the effect. The economic value of the privilege, Paul concludes, is probably not large relative to the size of the American economy.

Where the privilege is unambiguous, he argues, is in power rather than economics. Because so much of world trade and finance runs through the US banking system, Washington can observe and block transactions across borders. This is the weapon behind sanctions such as those imposed upon Russia and Iran — and it is a genuinely formidable one. When the US froze Russia out of dollar clearing after the invasion of Ukraine, it demonstrated that exclusion from the dollar system is close to exclusion from the global economy. Iran’s alleged demand that tolls through the Strait of Hormuz be paid in yuan or cryptocurrency is, in this light, less a sign of dollar weakness than a sign of how badly a sanctioned state wants out.

So the picture that emerges is this: the exorbitant privilege is real, but its economic dimension is modest and comes with a structural cost (deindustrialization). Its power dimension — the ability to weaponize dollar clearing — is large, and is not obviously a form of living beyond one’s means. It is more like owning the world’s only bridge and being able to decide who crosses.

Paul Krugman

The Dollar’s Special Status: Sources and Threats

A recent news analysis by Al-Jazeera stated…

Read more

6 months ago · 1296 likes · 214 comments · Paul Krugman

The System Was Not Built for This

This is where we need to discuss Izabella’s argument that the dollar is not about allowing America to live beyond its means.

Read more