Yes, we really do want to trust crypto interests with the future of money

A response to Dr. Amit Seru

As I built up a bit of a reputation for debunking MSM FUD around Bitcoin mining back in the day, people often ask me to write takedowns of crypto-critical pieces in the press. The topic du jour these days is stablecoins and their perceived imminent risk to the financial system. The trouble is there’s just such a high volume, so I have to pick them at random. I did get a few requests to look at this NYT Opinion piece by Dr. Amit Seru, a professor at Stanford. So let’s dig in, shall we?

The thrust of Dr. Seru’s argument is that stablecoins privatize what otherwise ought to be an important government function, namely issuing dollars. Stablecoins aren’t to be trusted, as Dr. Seru tells it – they aren’t “run proof”, they don’t come with privacy guarantees, they can’t transmit monetary policy, and they will be too heterogeneous. The government should step in and take responsibility for the digitization of the dollar with a CBDC, such as the ones that China and Europe are exploring.

Money isn’t always public

My first issue with the article is that it frames money as a kind of public infrastructure, with stablecoins an apparent aberration. He says: “If current legislative momentum continues, digital currency — the future of American money — will be forged by Silicon Valley, not by the state.” But this supposes that money should be the exclusive purview of the state, which it isn’t.

Today, only 25.9% of the money supply (as measured by M2) consists of “base money” which is physical currency and bank reserves at the Fed. Base money is the only directly government-issued money. The other three quarters – and we certainly consider it money or believe it has moneyness – is issued by financial institutions, albeit heavily regulated ones. The 75% of non-government issued money includes checking and savings accounts in commercial banks, money market funds, and so on.

And historically, there have been times when large portions of the money supply have been issued entirely by institutions totally unaccountable to the federal government. During the infamous “free banking” era in the US prior to the Civil War, between 1/5^th and 1/3^rd of money supply circulated in the form of privately issued banknotes produced by “free” banks, which were lightly regulated at the state level. (Free banking in the US is very unfairly maligned as I explain in a prior substack.)

And the US has always had a robust dual banking system, with states playing an important role. Today, of the 5000 or so banks that exist, 75% of them are state rather than nationally chartered. Most of these state chartered banks avoid Fed supervision entirely. A CBDC which sought to centralize money and payments and place it inside the ambit of, presumably, the Fed, would be an extremely aberrant structure within the context of our monetary history.

Money is already digital

Dr. Seru frames digital money as a kind of futuristic piece of infrastructure that the government should be building, to defend the supremacy of the dollar and keep up with other developed nations. He says “the United States would become the only major economy to voluntarily rule out government-issued digital money [while] our global rivals [rush] ahead, marketing their digital currencies as safer, state-backed alternatives to the dollar”.

And maybe I’m being pedantic, but the vast majority of US dollars are already digital. Only about 10% of M2 exists as physical cash and bills. It’s just that most of our digital dollars exist as commercial bank liabilities. There is as of yet no digital central bank liability that a retail individual or nonbank commercial institution can hold in the US, and this is what Dr. Seru is after. I do always find it funny that people stress how important it is that we digitize the dollar. The dollar is already mostly digital, it just mainly lives on commercial bank ledgers.

The US does not need a public retail settlement network

Dr. Seru commits the classic error of the critic which is to compare a real, deployed system with a theoretical yet of course superior system that exists only in his head. Stablecoins are by no means perfect, but I haven’t seen anything better so far. And it looks like Dr. Seru’s proposed alternative is either trivial or impossible. He describes his idealized system:

Done right, such a dollar could combine crypto’s efficiency — instant settlement, round-the-clock availability, low transaction costs and no need for intermediaries like banks or card networks — with the safety of public money. Imagine a service like Venmo, PayPal or Zelle, but with no risk of frozen funds or transfer reversals, no corporation mining your personal data, no fees for foreign transactions and backed by the full faith of the U.S. government.

It’s a little unclear what he’s actually arguing for here, so let’s interpret it in two ways. First, let’s say he wants to create a domestic retail push settlement network, like Brazil’s PIX or Mexico’s SPEI (as he implies by comparing his system to Venmo or PayPal).

The reason the US doesn’t have this and other countries (at a lower level of development) do is because the US built a mature payments system centuries ago via banks and private networks, with the Fed only taking responsibility for interbank settlement. The US is a much more federated system than most others, with 50 states (which also grant charters) and thousands of banks. Additionally, and this is where Dr. Seru has something of a point – the card networks, banks, and payment processors actually benefit from the latency in the system by monetizing float and interchange. So there is an economic pressure against centralizing everything.

Additionally, our payments systems were largely built pre internet and pre smartphones, so when it came time to create a universal digital payments system, incumbency created too much inertia. So there has been a path dependence in the development of the US system that meant we never got any kind of public payments infrastructure.

If you look at Brazil’s PIX (2020) or Mexico’s SPEI (2004), these were launched by countries that modernized much later, so a government-administered payments system was actually a huge upgrade over the established payments infrastructure. PIX did so well because Brazil was already highly smartphone penetrated but beset by weak banking, so a public realtime payments layer made plenty of sense. In both Brazil and Mexico there was sufficient political will and enough of an economic need for the government to nationalize a portion of the payments landscape. The US never had this acute need and has always been way too politically fragmented and suspicious of strong central government to pull this off. Not to mention the fact that the US has more banks than almost any other country, which makes them much harder to wrangle.

And the US has actually built a public real time gross settlement network. It’s called FedNow and allows banks to settle with each other instantly. It doesn’t serve clients directly, but in theory, if enough banks adopt it (and 700 have so far), your bank or fintech could offer instant payments that settle via FedNow.

Lastly, the private sector has filled in pretty well for the government here, with PayPal, Venmo, Cash App, Zelle, Visa Direct among others all being perfectly serviceable options. It’s hard to argue that the US consumer is underserved from a realtime payment app perspective.

The US could build something like this if it wanted. There’s just no immediate need.

A public global push network is a fantasy

If Dr. Seru is actually advocating for a global push network run by the US government, that’s a much more complicated proposition. As you can see in my taxonomy of payment networks that I created for this recent article, there actually are no government-run, global, push networks that exist (look in the upper right quadrant).

There are, conversely, many government-run domestic push systems (top left). There’s an obvious reason for this. Managing a real time settlement infrastructure requires strong central control, and that’s only possible domestically.

If we look again at the criteria Dr. Seru proposes for his digital dollar network – instant settlement, 24/7/365 availability, low transaction costs, no intermediaries, no recourse – these require a single administrator to enforce a specific ruleset on the entire network. For this to work globally, especially if retail users are involved, the US government would have to be responsible for onboarding and maintaining a database of hundreds of millions of individuals around the world, not just in the US. Somehow all of those people or institutions would have to agree to transact within the same ruleset and governance framework.

This stretches the bounds of imagination, which is why there are no global public push networks in existence. The international settlement system is getting more fragmented, not less, and there is no global hegemon that can impose a payments system on the entire planet.

The US has spent the last few decades taking advantage of its primacy in global payments and using this privileged position for strategic purposes. Our love of sanctions has steadily eroded the world’s trust in the US dollar system, and we now have parallel systems emerging like the China-Russia-Iran-India trade/commodities/payments axis. Our adversaries, mindful of our willingness to weaponize the dollar, are busily building their own settlement systems, selling US treasuries, and buying gold instead.

Against this backdrop of fraying trust in the dollar, even among our close allies, it’s hard to believe that a US government administered digital payment app would see much traction abroad.

And of course how can you forget that the Biden administration – and the Obama admin before it – spent its time politicizing domestic bank rails, as I discussed extensively with my reporting on Operation Chokepoint 2.0? It’s something of a habit in this country to weaponize regulators against the banks they supervise in order to debank and pressure entire industries – whether it’s crypto, payday lenders, or firearms. Given this reality, why would anyone who might be politically disfavored by a mercurial US government trust a government-administered payments network?

CBDCs have mostly failed so far

Though Dr. Seru treats CBDCs as an exciting innovation catching on globally, the CBDC mania that swept central banks over the last decade has seen mixed results and may be in retreat. He says:

China’s digital yuan is already in use. The European Union has passed a sweeping framework that treats digital assets as part of sovereign infrastructure — public systems built and controlled by governments — while the European Central Bank is testing a digital euro. Dozens of countries, from India to Brazil, are building their own sovereign digital payment systems.

It’s possible that we are dealing with a definitional issue, because according to the CBDC tracker, only 3 CBDCs have been truly deployed: in Jamaica, the Bahamas, and Nigeria. All of these launches have been failures, with very low adoption rates and virtually no organic demand. When I visited the Bahamas in 2024, I asked around about the Bahamian Sand Dollar because I was interested in getting my hands on some fresh CBDC. People were either puzzled or told me to look in the ocean. Not one person had heard of the digital currency project.

As of right now, according to CBDC tracker, there have been more cancelled CBDC projects than there are active ones.

In this 2024 article, the economist Kevin Dowd reviews every major CBDC experiment to date and comes to this grim conclusion:

I find that every single CBDC experiment has been a failure, in the sense that it failed to make the population any better off. In every case where data exist to assess the case, the public have been reluctant to adopt them. It also appears that CBDCs failed to offer any benefits that could not already be obtained using existing payments media. At best, these findings suggest that CBDCs are much more difficult to establish than their advocates realise

Mirroring my sentiments, Dowd concludes that “they will all fail and that the Great CBDC Project, the ‘current thing’ in central banking, will eventually be abandoned and will be looked back upon as an embarrassment to those currently promoting it.”

The reason being, I think, that CBDCs are stuck. If we’re talking about a retail CBDC, we are envisioning a system where the government endeavors to sign up as many citizens as possible to use its proprietary payments app and allows them to become direct holders of central bank liabilities. This would transform central banks into retail-facing institutions, something none is equipped for.

If we are talking about the more conservative hybrid or “two-tier” CBDC, that refers to a system where the government gives people access to the central bank ledger via financial intermediaries like fintechs or superapps who handle the onboarding and the wallets. This is of course very similar to how stablecoins work today, and hence redundant. You could quibble and say that the tokens are liabilities of the private issuer, and the CBDCs are (intermediated) liabilities of the central bank, but this is a minor distinction. The real difference between the two is that in the government case, the state retains visibility and control over user funds, whereas in the stablecoin case, the government is held at arm’s length.

Why someone would prefer a two-tier CBDC over a stablecoin arrangement is beyond me. A stablecoin is almost identical from a risk perspective, yet far more insulated from government interference than a CBDC. With a stablecoin, if the state wants to freeze your account, they need to establish probable cause and persuade a judge to send a letter to the issuer. With a CBDC, it’s unilateral and immediate.

Nonetheless, 19 of 20 G20 nations have somewhat advanced CBDC schemes and 130 countries around the world have at least considered the idea. And whenever I talk to central bankers about stablecoins, they are convinced that once they unleash their glorious CBDCs, stablecoins will become irrelevant and die off.

I would wager instead that central banks will seek to ban stablecoins as they release their CBDCs, as the success of stables is a constant source of embarrassment to them. With the government, it’s never a fair fight. I think this is why the US breaking from the pack and passing GENIUS is causing so much belly-aching around the world, both among our allies and our adversaries. Our allies know that the success of dollar stablecoins risks undermining their CBDC pilots; and our adversaries know that they will have a tough time persuading their citizens to drop the strong property rights of tokenized dollars and adopt the e-CNY instead.

Stablecoins do extend the dollar

Dr. Seru says “[Legislatively embracing stablecoins] also hands our global rivals a perfect opportunity to lure global trade and investment away from the dollar, which could have major consequences for America’s standing and its economic future.”

This simply ignores the track record of stablecoins so far. Stablecoins are already trusted by hundreds of millions of individuals around the world. Tether alone estimates that they have 493m users worldwide. As I’ve pointed out, over 99.8% of stablecoins reference the dollar. No other major FX is even close or appears able to compete. The EU, coming in a distant second at 16 bps of stablecoin supply, has passed onerous stablecoin regs ensuring that they’re basically uneconomical to produce. The Bank of England rags on stablecoins every chance it gets. China is apparently uninterested.

Stablecoins have surged from a single digit billion supply to around $300b in just five years, and have just been embraced by Congress. Treasury Secretary Bessent thinks they could grow into the trillions by 2030. And over this entire time, the market share of the dollar within stablecoins has remained constant at over 99 percent. Meanwhile, the offshore-focused Tether remains by far the largest issuer. Hotspots of adoption for USDT include Latin America, West Africa, South East Asia, and the Middle East. That stablecoins are reaching people who otherwise would have never had convenient access to the dollar is undeniable (see our research on this topic).

Stablecoins give ordinary folks outside the US access to something that resembles dollar banking, without needing a bank account. Almost no one aside from the very affluent ex-US can obtain a dollar bank account. This is why I have long thought that stablecoins would drive periodic waves of “cryptodollarization”, which we have already seen in Venezuela, and to lesser degrees in Argentina and Nigeria. I expect that US dollar stablecoins will be blamed for the collapse of a dozen currencies before the decade is out. (Cryptodollarization requires only the internet, smartphones, and someone willing to make a market. This is much more efficient and harder to stop than literally importing physical dollars, which is how dollarization worked in the past.)

And contrary to what Dr. Seru is saying, we are starting to see trade invoiced and settled in stablecoin terms. They are not just a retail payments app anymore. Our research shows that they are being used for payroll, vendor payments, and b2b especially cross border when payments systems are expensive or slow.

Granted, a big portion of the supply of stablecoins derives from the simple conversion of USD bank deposits into dollars on chain. But there is a meaningful, if hard to quantify, portion of stablecoin supply which is net new dollar exposure resulting from currency substitution in emerging markets.

Stablecoins will be folded into the monetary toolkit eventually

Dr. Seru worries that stablecoins cannot transmit monetary policy because they are issued by the private sector. He says “The private sector is stuck with the responsibility for keeping the dollar relevant without the flexibility or authority needed to do so.”

But the Fed will have plenty of ways to transmit monetary policy through stablecoins should they want to. If stablecoin issuers get access to Fed master accounts, as I expect they will, the Fed could influence the rate stablecoin issuers pay (assuming they do indeed end up passing on interest indirectly) with the interest rate on reserves. Stablecoin issuers are and will continue to be straightforwardly part of the interest rate channel. There’s no reason that stablecoin issuers couldn’t eventually be included in Fed liquidity facilities either, especially if banks become issuers.

At the end of the day I see GENIUS-regulated domestic stables as about as relevant to monetary policy as money market funds. To me, stablecoins are “synthetic CBDCs,” because if issuers are holding reserves at the Fed, they’re basically private-sector CBDC issuers – without the privacy and control issues a direct CBDC would introduce. (This is why I sometimes joke that Circle’s token should be called USDCBDC.) Truthfully, I don’t think there’s that much of a gap between what Dr. Seru is asking for and the way stablecoins behave today.

And the Fed has historically shown an ability to adapt to changing conditions when the market embraced a new modality. Recall that Eurodollars (non-US bank dollar liabilities) also went through a similar cycle of being feared and misunderstood (in the 50s and 60s), until the Fed realized that they were essential to maintain the stability of the dollar system, post Nixon and oil shock. The Fed belatedly realized the importance of Eurodollars in internationalizing dollar liquidity (sound familiar?) and easing a supply crunch by allowing dollar funding to expand without loosening conditions domestically. They created swap lines with foreign central banks after the oil shock to recognize this, implicitly supporting the banks that were issuing dollar liabilities (without direct access to the Fed) abroad.

The Fed similarly came to implicitly back money market funds, if you look at the Reserve Fund collapse in 2008 or the MMMF liquidity crisis in 2020. I expect that they will eventually embrace stablecoins by guaranteeing Treasury liquidity in times of stress. Note that no purely UST-backed MMMF has ever had real issues, as they are derivatives for the most liquid market on the planet. The Reserve Fund (which is what people refer to when they worry about MMMFs “breaking the buck”) was stuffed full of Lehman commercial paper which is obviously prohibited under GENIUS.

On this point, it’s worth touching on Dr. Seru’s complaint about a “lack of a government guarantee” if a stablecoin fails. The thing is, it’s extremely hard to imagine a situation in which a GENIUS-regulated stablecoin could actually experience an abrupt collapse, unless Treasury markets themselves completely seize up (in which case we have bigger problems). Why would you need a guarantee for an instrument which is simply a tokenized version of the most liquid market on the planet?

The government will not be able to offer privacy through a CBDC

One of Dr. Seru’s points – and something I’ve heard more than a few times from CBDC enthusiasts – is that the US government will be able to offer privacy, in a way that private sector issuers wouldn’t. He says, of stablecoins:

Access to these currencies will surely be limited to those Americans deemed sufficiently profitable, and privacy will be governed by whatever policies companies adopt rather than by federal law

He implies that only the government, not the private sector, can offer transactional privacy. This sounds kind of elegant, but it falls apart when you consider recent political reality. Simply looking at the 55 years, financial privacy has been all but destroyed in the US via the Bank Secrecy Act, the Third-Party Doctrine established by US v Miller, expansion of AML in 80s and 90s, the PATRIOT Act, and much more besides.

There is absolutely no political desire, it seems, to reduce the burden of financial surveillance in this country, even if our AML regime is broken and extremely costly relative to its perceived benefits. In fact, the ratification of the stablecoin status quo under GENIUS is the first major step back towards financial privacy that we’ve seen in which some time. (Stablecoins operate on the permissioned pseudonymity model, meaning that most transactions within the network are unsurveilled.)

Dr. Seru and other CBDC advocates would try to have us believe that the same Congress and courts that created a ravenous financial surveillance engine since the 70s would throw it all away for the sake of a government-run payments network which strenuously upholds user privacy. This I don’t see as remotely plausible. Any likely CBDC in the US would include, if I had to guess, KYC, transaction reporting (SARs/CTRs etc), OFAC screening, travel rule compliance, and so on. And frankly, if a newly-principled US government decided to abruptly create a fully private digital token, it would fall afoul of 50 years of hard-won judicial precedent!

As Cato’s Nick Anthony points out, even if a CBDC is not used for abusive purposes on day 1, it gives the government massive theoretical power that could be weaponized later. In my view, only private sector-issued stablecoins can provide their users with meaningful privacy. The way stablecoins work today is that if issuers get a letter from law enforcement (who periodically connect on-chain addresses to real-world identities), they freeze or seize the coins. This happens enough to placate law enforcement, but not so much that ordinary users are risking anything. And different issuers have different policies, giving users a degree of choice. Tether is more trigger-happy with freezes and seizes than Circle, for instance. Stablecoin transactions are not proactively disclosed to the government. In this respect, they are similar to physical cash. If a major stablecoin changed its ToS and started demanding that users provide KYC information with every transaction, users could presumably switch pretty frictionlessly into a more private coin.

The fact that GENIUS passed while upholding this de facto privacy model (which is wildly out of step with US financial system precedent) was surprising to me, but encouraging.

The argument that only CBDCs can offer privacy might hold if we were living in a Jeffersonian republic with extremely robust individual liberties and small central government, but that’s not the world we live in today. Until the government rediscovers its love of financial privacy, I have an extremely hard time believing that any putative CBDC would grant users meaningful privacy assurances.

Summing up

If you’ve followed my Twitter drama over the years, some of you will know that my dad, Laurence, spent his career at the World Bank. What he actually did was pioneer and develop the concept of a public-private partnership (PPP). This is basically a libertarian’s dream job, because it involves taking wasteful or inefficient government services and privatizing them so the capitalist motive kicks in and they work better. The World Bank’s job was to manage the bid process, derisk the project for external capital, and make sure the process of divesting a government asset to the private sector wasn’t corrupt (as it otherwise would tend be in those frontier markets).

So growing up my dad would tell me about so and so government service that he was taking private – typically a port, toll road, airport, water utility, or something like that. Invariably, if things went well (and they didn’t always!) the KPIs would skyrocket. This kind of thing is controversial in the developed world (you may have heard Brits yammer about “privatizing the NHS”) but in frontier markets in Africa, bringing in an established firm to build or run a wasteful government project is the difference between having water/power/roads and not. I was always impressed listening to my dad’s stories of how performance was vastly ameliorated once a private firm was put in charge. This was quite formative for me, in fact, and probably explains why I’m a VC and not a diplomat.

And I think this is the crux of the debate around stablecoins and CBDCs. Dr. Seru and his tribe tend to believe that money is a public good, like “clean air or national defense” and only the state can safely create it. There’s certain qualities of money that CBDC advocates think only the government can produce – privacy, singleness of money (the ‘NQA principle’), interoperability, financial inclusion, settlement finality to name a few. But I think when you look empirically at the track record of stablecoins, they do a perfectly good job at providing users with sufficient privacy, trading at par, serving the least privileged, providing final settlement, interoperating, and so on. In fact, when it comes to privacy and interoperability, structurally I think they can vastly exceed anything a government can do.

Perhaps I’m biased, because I’m generally inclined to believe that anything the private sector creates will be more efficient and creative than the government alternative. And the thing with stablecoins is they just keep going from strength to strength, and now post-GENIUS are safer than ever. Unless the CBDC advocates can prove that CBDCs can match stablecoins where it matters, and that there’s some quality of money that private issuers cannot produce and only the government can, I’ll remain in the stablecoin camp.