Stablecoins are private money. There’s nothing wrong with that

A response to the WSJ and a century of misreadings of the "free banking" era

WSJ: Stablecoins Are Private Money. That’s Why They’re a Risk to the Economy

Greg Ip, the chief economics commentator for the Wall Street Journal recently published an article entitled Stablecoins Are Private Money. That’s Why They’re a Risk to the Economy. The argument turns on the familiar claim often bandied about by establishment journalists and policymakers: a proper money can only be issued by the state. As a privately issued monetary medium, stablecoins cannot possibly be fit and proper. As such, their emergence poses a risk to the financial system.

This is fundamentally a definitional argument. Money is public, perhaps even a “public good” as Ip says. Stablecoins are private. Private issuance cannot work, the story goes. History demonstrates as much, if you consider the unstable “wildcat” banking era. As such, stablecoins are destined to fail, or perhaps should be regulated out of existence, so we avoid the fallout from such a failure.

But what if the history we were taught in economics class is wrong? What if competitive issuance, absent central bank authority, can create a stable and functional system? What if the American free banking era was a failure because of regulation, not due to its absence?

This would pose a significant problem for the establishment. They would no longer be able to define stablecoins out of existence. They would have to concede that private monetary media can indeed lead to a flourishing, efficient, and stable system.

Whether it’s due to an aversion to stablecoins, or a rigid commitment to a statist view of financial history, or simply ignorance of free banking episodes in the US, Scotland, or Canada, the economics establishment seems determined to deny the history of private monetary issuance, substituting instead a version that aligns with the post-1970 consensus around money: something that only the government can create.

If you actually consider the history, the conclusion is clear. Emergent private systems without any central bank can indeed mediate the competitive issuance of money, and have worked perfectly well for long periods of time in the past. Stablecoins are in many ways the spiritual successors to these “free banking” regimes, though they differ in some critical ways. Even if you abandon the free banking analogy, the very real track record of stablecoins over the last decade – not to mention regulatory improvements post-GENIUS – demonstrates how functional a private monetary system can be. Let’s consider Ip’s claims one at a time.

Claim: the history of Free Banking demonstrates the imprudence of allowing private banknote issuance

Ip says:

To proponents, stablecoins are crypto’s killer app. They will make payments faster and more efficient, especially across borders, than the legacy banking system makes possible. With that promise, though, comes the risk that this could lead to a financial crisis, much like some past experiments with private money.

Later on, he elaborates by referring to the US “free banking” era:

During the free banking era from 1837 to 1863, banks could issue their own currency. But the system was inefficient, with currency values that fluctuated against each other.

He quotes an NBER paper summarizing the problems these free banks exhibited:

All states maintained a range of requirements for banks to collateralize their notes, but many proved ineffective; fraud was widespread, and the system was fragmented—banknotes of one bank were often not accepted by other banks outside the local area; bank failures were widespread.

To summarize Ip’s argument, he seems to believe that private issuers of banknotes (or digital banknotes, in the case of stablecoins) are inherently unstable, since only the state can guarantee bank stability through oversight and centralized liquidity provisioning, alongside depository insurance. The US’ troubled experience with “free banking” is cited as evidence in his favor.

Unfortunately, Ip misreads history and makes a poor analogy with stablecoins.

“Free Banking” in the US was not unrestricted or unregulated. Rather, it was an excess of the regulation from the States – namely, forcing the banks to hold low-quality collateral, and restricting branching (geographic diversification) that led to its notorious failures.

The factors that led to the instability, deviations from par, and failures of the free banks in the US are not present in the context of GENIUS-regulated stablecoins. They simply don’t apply, so the analogy fails.

Other episodes of more genuinely “free” or laissez-faire banking, such as those that appeared in Scotland and Canada (which, as a Canadian himself, Ip should know about), were notable successes, demonstrating that banks can indeed issue banknotes with no bank failures and complete stability – absent the supervision of a central bank.

The story of the failures of American quasi-free banking, and the successes of Scottish or Canadian free banking, undermine rather than support Ip’s point.

Ip is treading on familiar ground. He is not the first to make this claim, and I already considered it at length last year:

This issue – ahistorical comparisons between stablecoins and misreadings of the “free banking” era in the US – has actually plagued me since 2021. It’s an incredibly popular line of attack among economists and central bankers. The only trouble is – none of them seem to have read the history.

I don’t have space to get into a full breakdown of the American, Canadian, and Scottish systems, so I suggest you read my prior article. But I will summarize here.

“Free banking” in the pre-Civil War era in the US only meant that you did not need a special charter to open a bank. This did not mean that banks were unregulated. Indeed, the States actually harshly regulated these “free” banks, prohibiting branching (all free banks were unit banks – as in, they were only permitted to serve one community), and requiring the banks to hold questionable state bonds as collateral. The bank’s assets – and hence banknotes – were only as good as the states’ solvency. Moreover, the inability to branch meant that banks were not geographically diversified, and horribly exposed to local crises. Banks failed when their bond portfolios depreciated – and unsurprisingly, many states’ bonds did not perform well in the lead up to the Civil War. Banknotes traded at discounts reflecting the actual cost of redeeming them for specie, which was normally a function of how far the banknote had traveled from the issuing bank.

None of these lessons really apply to stablecoins, nor do they make Ip’s point that there is anything inherently dastardly about private banknote issuance.

Instances of genuine laissez-faire or “free” banking can be found in Scotland (1716 to 1845) or Canada (1837 to 1935). These systems more closely resemble a truly private monetary system lacking a central bank, and much to the chagrin of today’s economists, they thrived. These counterexamples to the commonly-cited but misunderstood American “free banking” era are devastating to the establishment’s normative views around money and are hence seldom brought up.

For over 100 years, Scotland had multiple competing note issuers (issued against gold specie), no central bank, no state-imposed collateral, with active branching. Banks were disincentivized from over-issuing by market forces. They mutually accepted each other’s banknotes and developed a clearinghouse in Edinburgh, ensuring that the notes traded at par. Bank failures were extremely rare (I am aware of only one major one – the Ayr bank) and inflation was low. Meanwhile, England which had a central bank during the time period lurched from crisis to crisis and had hundreds of bank failures. This is a compelling argument because both countries were subject to similar fundamental economic conditions.

Canada is another great counterexample to Ip’s framework whereby only the state can guarantee financial stability. Canada had no central bank until 1935. Though charters were more restricted than in the Scottish system, the banking system was still relatively “free”. Banks issued their own notes (redeemable for specie), branched widely, and were not forced to hold government bonds (unlike their peers to the south). Private clearinghouses ensured that notes were generally traded at par, and the system gradually consolidated into a small number of large, nationally branched banks, giving them the geographic diversification their American unit bank counterparts were denied.

The result is a comparison that should be uncomfortable for anyone holding Ip’s view. During the Great Depression, the US – equipped with the Federal Reserve and centralized oversight – suffered roughly 9,000 bank failures. Canada, with no central bank and privately issued banknotes, had none. Both countries faced the same depression and the same monetary contraction. Only the US had a banking crisis. If only the state can guarantee financial stability, this outcome is inexplicable.

Verdict: the American free banking analogy fails on its own terms, and the more apt analogies – Canada and Scotland – undermine Ip’s framework.

Claim: Stablecoins might “break the buck” like Money Market Funds

Ip compares stablecoins to “private monies” like money market funds and suggests that, like MMFs, we risk fragilizing the financial system if we rely more on them while assuming they will always trade at par. He says:

Money-market funds are a type of private money, promising to redeem shares at a dollar each, on demand. But during the global financial crisis, one fund couldn’t honor that value—it “broke the buck”—because it held devalued assets. A broader panic ensued.

This is an interesting line of attack since money market funds were indeed subject to a lot of criticism when they were first being developed, along the same lines as stablecoins are today. Critics alleged that MMFs were creating a parallel private monetary system outside of the regulated banking perimeter. There was great concern, as there is today, about deposit flight from the banks. Critics complained that money market funds were “free riding” on the financial system while not submitting themselves to the supervision and regulation or insurance of banks.

Of course, MMFs proved tremendously useful and grew to around $7.7 trillion largely without incident, and these criticisms now seem quaint in hindsight.

Of course, there were hiccups, like The Reserve Fund’s infamous depeg during the Financial Crisis. But that was hardly a catastrophe, and it doesn’t generalize to Genius-regulated stablecoins.

As Ip surely knows, the Reserve Primary Fund – the flagship and first money market fund, established in 1971 – suffered a run in September 2008 when Lehman collapsed and holders realized that the Fund was holding $758m of Lehman commercial paper. The next day, the NAV fell to $0.97 per share. This was the infamous “breaking of the buck”. The fund actually held $65b of assets so the “true loss” was 1.2% per share, but the flood of panicked redemptions drove the NAV below that level (similar to the USDC dynamic when SVB failed). In the end, Treasury stepped in and guaranteed MMFs and stopped the run, while the Fed created emergency liquidity facilities. The ultimate loss realized by holders of the Reserve Fund was less than one cent per share.

Now could this happen with stablecoins? Of course, runs and depegs have been observed with stablecoins in the past, notably USDC during the SVB collapse in March 2023, since SVB held $3.3b of USDC cash or 8 percent of backing assets. USDC traded as low as 87 cents on the dollar before SVB depositors were made whole. Could this happen again? Perhaps, but issuers these days are much more cautious about where they park their cash. Issuers are now much less inclined to leave cash in banks at all, preferring to hold short-duration US Treasuries. Banks can always experience crises, but this is hardly a stablecoin-specific issue.

And the worries about the Reserve Fund depeg just don’t apply to GENIUS-regulated stablecoins. GENIUS requires that issuers maintain 1:1 reserve backing consisting of cash, short-duration Treasuries, repo, or certain MMFs (holding those same assets). They would not be able to hold commercial paper like the Reserve Fund, or MMFs holding those types of assets. Could the 90-day Treasury market become dislocated? Perhaps, but if so, the world is basically ending, because it’s the most liquid market on the planet. And despite the scaremongering over the Reserve Fund, no money market fund holding only short-duration Treasuries has ever broken the buck.

Can I guarantee to you that stablecoins will always trade at par? No. A derivative is by definition less liquid than the underlying. But the underlying for GENIUS stablecoins – cash and short-duration Treasuries – is pretty damn liquid.

Verdict: no, GENIUS-regulated stablecoins won’t break the buck like the Reserve Fund. Bad things can always happen, but if they do, something much bigger and much worse is going on, like a full-on banking or government solvency crisis.

Claim: Stablecoins do not exhibit “singleness”

Another objection we hear all the time is that stablecoins do not trade reliably at par, and hence violate the “no questions asked” or “singleness” principle. To work properly, the thinking goes, a dollar needs to be a dollar at all times and circumstances. Though he doesn’t cite it, this objection frequently traces to Gorton and Zhang’s regrettable paper “Taming Wildcat Stablecoins”. Ip says:

An essential quality of money is “singleness,” meaning a dollar must always equal a dollar no matter when, where or with whom it is used. Bank deposits are a form of private money, but because banks can borrow from the Federal Reserve to redeem deposits, and dollars move between banks via the Fed, their dollars exhibit singleness.

By contrast stablecoins move through proprietary, fragmented infrastructures. They don’t exhibit singleness. Though coins issued by Tether and Circle are intended to stay fixed to the U.S. dollar, they often deviate from that value, albeit usually by tiny amounts.

Ip cites BIS data supporting his thesis, showing tiny peg deviations in Circle and Tether’s history. This is fairly uncharitable, as they have faced virtually no deviations since GENIUS was passed, and the law instilled stringent frameworks that are intended to ensure that they generally trade at par.

It always amuses me that establishment thinkers like to claim that stablecoins don’t work in theory, so they can’t work in practice. These theoretical shortcomings might surprise the tens of millions of people who use stablecoins every month as money.

More generally, no monetary good satisfies Ip’s requirement of absolute singleness at all times. Though they were holding domestic-banking-system-dollars, depositors at SVB, First Republic, and Signature certainly did not feel the warmth of “singleness” that weekend in March 2023. Even today, uninsured deposits at banks arguably do not satisfy Ip’s singleness or Gorton and Zhang’s “NQA” policy. The quality of bank dollars is mostly a function of the Fed’s desire to intervene in a crisis. So if the banking crisis of 2023 doesn’t eliminate the “singleness” of bank dollars, a Tether depeg in 2018 shouldn’t discredit stablecoins either.

As many have pointed out, money is more of a quality than a binary. Different media have varying amounts of moneyness. Stablecoins have clearly graduated from ersatz crypto casino chips to globally trusted money inside the US regulatory ambit.

Does a medium have to be absolutely perfect to meet some economist’s definition of money? It would be completely ahistorical and empirically blind to presume so.

Verdict: if stablecoins don’t have singleness, bank dollars don’t either. Millions of people worldwide behave as if stablecoins do have singleness.

Claim: Stablecoins could back their reserves with Bitcoin

In one of his more tenuous claims, Ip says that even GENIUS-regulated stablecoins might be backed with exotic assets such as Bitcoin. He writes:

Last year’s Genius Act requires stablecoins that cater to Americans be backed with safe, liquid assets such as treasury bills and bank deposits. But Fed governor Michael Barr noted last year the law has loopholes. The bank deposits may be uninsured. The law allows stablecoins to receive money, including foreign money, through “repo” loans, and that could include bitcoin, which El Salvador recognizes as money, Barr noted.

What Ip is envisioning, via Barr, is a situation where a stablecoin issuer holds as a reserve asset a repo receivable (loan) secured against Bitcoin, in the case that it is considered money. If Bitcoin were to crash hard overnight in this extreme hypothetical, the issuer might end up with Bitcoin, or face a disorderly liquidation.

Why any stablecoin issuer would do this is completely beyond me, but luckily it’s not something we have to worry about, because this entire hypothetical rests on a faulty assumption. When Barr gave the speech that Ip is referring to in October 2025, El Salvador – the only country treating Bitcoin as legal tender – had already rescinded Bitcoin’s status nine months before. Not only is this a theoretical loophole that would likely be closed under the regulatory implementation of GENIUS – exactly what Barr was asking for in his speech – it was already outdated. El Salvador was no longer treating Bitcoin as legal tender.

Verdict: not a real concern.

Claim: Stablecoins aren’t used for “real world” payments

Ip says:

Trading crypto remains the primary use of stablecoins. Today, less than 1% of stablecoin usage is for real-economy payments, a Kansas City Fed study concluded.

Before we get into the data, I want to point out that this isn’t discrediting at all. Were Ip to consider the conventional dollar system, he would realize that the vast, vast majority of dollar transaction volume – at least 80-90% – has to do with financial markets. Securities settlement, repos, clearing, FX, and so on. The volume of “real world” payments, though large in absolute terms, is completely dwarfed by the amount of cash that sloshes around daily in financial markets. So even if that 1% statistic were accurate, it wouldn’t discredit stablecoins at all. Of course they are used to settle financial market transactions. That’s how money is overwhelmingly used everywhere.

Secondly, the data doesn’t support that “less than 1% real-economy usage” assumption. That statistic from the KC Fed study is based on a piece of research we, Castle Island, published with Dragonfly and Artemis.

Source: Stablecoin Payments from the Ground Up

The way this study worked was we went to 31 stablecoin payments companies and asked them for their own data regarding transaction volumes between known individuals or firms. The largest category was “b2b”, though card payments were growing extremely rapidly, and are far, far larger today. This was the first “bottom up” estimate of its kind. And it was very much a lower bound estimate. There are not just 31 stablecoin PSPs. There are hundreds of such firms.

Looking at the chart, you’d be hard-pressed to think “real world stablecoin payments aren’t material”. They are growing extraordinarily rapidly, as we (Castle Island) can directly attest. Rain, one of the issuers of stablecoin-linked cards profiled in the study, has been a massive success story. The WSJ need only look at the evidence around them to see that stablecoin payment volumes are growing at a rapid clip.

What the KC Fed did was take the (outdated, lower bound) findings from our study, make additional assumptions around payment velocity, and determine that less than 1% of stablecoin supply was necessary to support them. While I appreciate them using our data, this estimate (and hence the WSJ’s) is insufficiently caveated. Our data was a snapshot of a subset of the market, not an attempt at a full estimate of stablecoin payments volume. If I wanted to do that, I would refer Ip to Fireblocks’ estimate published in Fortune. They are a massive infrastructure provider and many firms build their transaction systems on top of them. As such, they have a more comprehensive sample with which to make an estimate. Based on data published in early 2025, ten percent of their client’s volumes related to payments with the rest trading. You can see it growing rapidly. It is assuredly higher today.

Source: Fortune

So as it turns out, the best data we have suggests that stablecoins are actually quite similar to ordinary cash in the financial system. The bulk of settlement value pertains to financial market activity, but that doesn’t discredit the payment volume.

Verdict: stablecoins mirror ordinary dollars, in that the bulk of transaction value has to do with securities settlement, but payments are still growing and material. Ip simply hasn’t considered the totality of the data.

Summing up

Despite what the critics would have you believe, stablecoins are not doomed to repeat the failures of the antebellum “free banking” era. They have none of the features that fragilized the American free banks:

• The free banks suffered because they could not branch; stablecoins do not face this problem as they can scale indefinitely and serve anyone

• Banknotes traded at discounts based on the geographic cost of redeeming them for specie; stablecoins are not costly to redeem

• Free banks failed because they were forced to hold depreciating State bonds that collapsed as the Civil War loomed; GENIUS-regulated stablecoins hold short duration US Treasuries, the most liquid security in existence

If you look at the track records of more competitive, less regulated systems like Canada or Scotland where private banknote issuance took place absent a central bank, you see remarkable track records of stability. This should encourage you if there is any analogy at all to be found between stablecoins and free banks.

But of course, stablecoins are not free banks. The free banks issued loans and held partial reserves in the form of specie (gold coins). Stablecoins, by contrast, especially GENIUS-regulated ones, can hold only high quality liquid assets like short-duration Treasuries. They are more like money market funds in this regard. Architecturally, they are closer to a narrow bank than either a money market fund or anything in the historical record of private money. Whatever the right analogy is, it isn’t a unit bank in Wisconsin in the 1850s holding depreciating state bonds, nor is it the Reserve Fund holding Lehman debt in 2008. It’s something new, and safer than both.

Stripping away the analogies, the empirical record for stablecoins is compelling enough. Tether and Circle have suffered countless shocks and confidence crises, political harassment, debanking, a monetary tightening cycle, and failures of the very banks they relied on. They’ve emerged from these crises with intact pegs and growing global userbases. It is sophistry to assert that they can’t serve as money because they don’t suit an economists’ definition of what money is.

The deeper issue is that stablecoins expose the ideological failures of the establishment over the last decade. Over the past decade, the dollar’s most successful expansion came not from the Fed, not from a CBDC, but from the crypto industry – warts and all. Acknowledging the successes of decentralized systems like Free Banking or stablecoins is painful for an establishment that defines money as the exclusive purview of the state. But neither the history nor the present-day data cooperate with the establishment’s interpretation.