I've written a response to an anti-stablecoin op-ed published in the WSJ by acclaimed Stanford Professor Amit Seru,
Narrow Banks Create Credit
One of my favorite things about crypto is the way its novelty challenges otherwise intelligent people. Some rise to the occasion, opening their minds, learning, and evolving their thinking. Others reveal previously unearthed gaps in their knowledge, at best, or just good old fashioned bias.
I’m fortunate to know plenty of the former in academia. For a good example of the latter, see this anti-stablecoin op-ed in the Wall Street Journal from accomplished Stanford professor Amit Seru. In it, he argues that stablecoins are a sort of sheep in wolf’s clothing (my words), promising financial innovation and safer banking but delivering neither. He is not a fan of the Genius Act.
For a good explanation of what Seru gets wrong about Genius, I recommend @intangiblecoins Thorn’s article on Twitter. My beef with Seru’s op-ed is what he gets wrong about banking and credit.
Stablecoins, as Genius dictates, are just narrow banks whose liabilities exist in a more superior form factor, the cryptographic token. Seru doesn’t seem to take issue with the token, but he doesn’t like narrow banking:
But clarity isn’t the same as safety. The act formalizes stablecoins as narrow banks—entities that collect deposits but don’t lend—in all but name. That means no maturity mismatch, yes, but also no credit intermediation. The economic engine of finance, transforming savings into investment, is bypassed. Run-proof money becomes idle money.
This is a recycling of an old canard that levered (AKA fractionally reserved) banks create credit while narrow ones don’t. This is false, and I can prove it.
Circle actually published an attestation of how much credit it has created every month. Per Deloitte, Circle’s reserves at the end of May consisted of Treasuries, repurchase agreements, and cash parked at banks.
Treasuries are a transferable loan to the US government.
Repos are secured loans to other financial institutions.
A bank deposit is money lent to a bank.
If these aren’t credit, then I don’t know what is. And I promise, all 3 types of counterparties put that money to work. The US government doesn’t just borrow money via the bond market and sit on it. Neither do firms paying over 4% interest for term loans, or banks that collect deposits. There’s no idle money here.
People who argue that only levered banks create credit ignore the inconvenient fact that in America, banks account for just 20% of credit creation. There are five trillion dollars in money market funds alone. Is that also idle money? What about the agency market for mortgage-backed securities?
Are the pension funds that buy MBS from Fannie and Freddie—as opposed to parking their money at a levered bank—doing the equivalent of putting their money in a shoe box? Of course not, they are participating in the credit creation that drives the housing market.
Credit is something that gets created anytime an economic agent lends money to another. It doesn’t matter if you do it via a stablecoin, money market fund, vanilla savings account, or direct loan to your uncle.
There are benefits to lending via an intermediary—I’d much rather invest in a bond fund that buys agency debt than lend the money directly to a stranger who wants to buy a house. But there are many kinds of intermediaries, and each has its own pluses and minuses. Levered banks like SVB practice maturity transformation, and that brings down the cost of long term debt, but they do so at the risk of catastrophic runs.
Ironically, Seru himself explains this dynamic:
This wishful thinking is fueled in part by the collapse of Silicon Valley Bank in 2023. That was no tale of subprime mortgages or exotic derivatives, but a rerun of the oldest story in banking: maturity mismatch. Depositors, in particular those uninsured, can withdraw on demand. Banks invest long-term. When interest rates jump and confidence cracks, withdrawals follow, assets are fire-sold, and the government steps in. Again.
But quixotically, he doesn’t use this flaw of traditional banking to defend narrow banking. He projects it unto stablecoins later in the op-ed:
As ever in finance, as in fables, great power often hides even greater fragility. If stablecoins become embedded in everyday transactions, their failure won’t be contained in the crypto world. It will become a problem for households, firms and, through an inevitable bailout, taxpayers.
That stablecoins will face the same kind of runs that levered banks do—and require the same kind of anti-run measures like deposit insurance and taxpayer bailouts—is a second fallacy pushed by confused skeptics.
In reality, narrow banks are safer. That safety makes depositors less likely to run in the first place, thus requiring less deposit insurance or bailouts. In theory, a Genius-compliant stablecoin issuer can unwind its entire balance sheet with minimum impact.
The liquidation of its reserves might cause temporary distortions in the bond and repo markets, but that’s just how markets work. We can’t regulate away selling. The distortion caused by a narrow bank unwind are always less than when a fractionally-reserved bank fails. Genius compliant stablecoins also enjoy bankruptcy remoteness, which is better than the bankruptcy supremacy FinTech depositors often get.
I’ve read professor Seru’s op-ed at least five times, and still don’t really understand what his point is, except to argue against change. He’s an expert on bank runs, so you’d think he’d be pro a safer form of intermediation.
Perhaps he’s just a fan of Bitcoin. How else are we to interpret this statement:
For all the hype, stablecoins haven’t transcended banking. They have replicated its tensions in new form. The real promise of blockchain was to end trust dependencies. Instead, we are doubling down on them, now under federal supervision.
Which brings me to one of my other favorite things about crypto: the more people argue against one aspect of it, the more they inadvertently end up defending another. Maybe that’s what happens when you oppose progress.
Narrow Banks Create Credit
One of my favorite things about crypto is the way its novelty challenges otherwise intelligent people. Some rise to the occasion, opening their minds, learning, and evolving their thinking. Others reveal previously unearthed gaps in their knowledge, at best, or just good old fashioned bias.
I’m fortunate to know plenty of the former in academia. For a good example of the latter, see this anti-stablecoin op-ed in the Wall Street Journal from accomplished Stanford professor Amit Seru. In it, he argues that stablecoins are a sort of sheep in wolf’s clothing (my words), promising financial innovation and safer banking but delivering neither. He is not a fan of the Genius Act.
For a good explanation of what Seru gets wrong about Genius, I recommend @intangiblecoins Thorn’s article on Twitter. My beef with Seru’s op-ed is what he gets wrong about banking and credit.
Stablecoins, as Genius dictates, are just narrow banks whose liabilities exist in a more superior form factor, the cryptographic token. Seru doesn’t seem to take issue with the token, but he doesn’t like narrow banking:
But clarity isn’t the same as safety. The act formalizes stablecoins as narrow banks—entities that collect deposits but don’t lend—in all but name. That means no maturity mismatch, yes, but also no credit intermediation. The economic engine of finance, transforming savings into investment, is bypassed. Run-proof money becomes idle money.
This is a recycling of an old canard that levered (AKA fractionally reserved) banks create credit while narrow ones don’t. This is false, and I can prove it.
Circle actually published an attestation of how much credit it has created every month. Per Deloitte, Circle’s reserves at the end of May consisted of Treasuries, repurchase agreements, and cash parked at banks.
Treasuries are a transferable loan to the US government.
Repos are secured loans to other financial institutions.
A bank deposit is money lent to a bank.
If these aren’t credit, then I don’t know what is. And I promise, all 3 types of counterparties put that money to work. The US government doesn’t just borrow money via the bond market and sit on it. Neither do firms paying over 4% interest for term loans, or banks that collect deposits. There’s no idle money here.
People who argue that only levered banks create credit ignore the inconvenient fact that in America, banks account for just 20% of credit creation. There are five trillion dollars in money market funds alone. Is that also idle money? What about the agency market for mortgage-backed securities?
Are the pension funds that buy MBS from Fannie and Freddie—as opposed to parking their money at a levered bank—doing the equivalent of putting their money in a shoe box? Of course not, they are participating in the credit creation that drives the housing market.
Credit is something that gets created anytime an economic agent lends money to another. It doesn’t matter if you do it via a stablecoin, money market fund, vanilla savings account, or direct loan to your uncle.
There are benefits to lending via an intermediary—I’d much rather invest in a bond fund that buys agency debt than lend the money directly to a stranger who wants to buy a house. But there are many kinds of intermediaries, and each has its own pluses and minuses. Levered banks like SVB practice maturity transformation, and that brings down the cost of long term debt, but they do so at the risk of catastrophic runs.
Ironically, Seru himself explains this dynamic:
This wishful thinking is fueled in part by the collapse of Silicon Valley Bank in 2023. That was no tale of subprime mortgages or exotic derivatives, but a rerun of the oldest story in banking: maturity mismatch. Depositors, in particular those uninsured, can withdraw on demand. Banks invest long-term. When interest rates jump and confidence cracks, withdrawals follow, assets are fire-sold, and the government steps in. Again.
But quixotically, he doesn’t use this flaw of traditional banking to defend narrow banking. He projects it unto stablecoins later in the op-ed:
As ever in finance, as in fables, great power often hides even greater fragility. If stablecoins become embedded in everyday transactions, their failure won’t be contained in the crypto world. It will become a problem for households, firms and, through an inevitable bailout, taxpayers.
That stablecoins will face the same kind of runs that levered banks do—and require the same kind of anti-run measures like deposit insurance and taxpayer bailouts—is a second fallacy pushed by confused skeptics.
In reality, narrow banks are safer. That safety makes depositors less likely to run in the first place, thus requiring less deposit insurance or bailouts. In theory, a Genius-compliant stablecoin issuer can unwind its entire balance sheet with minimum impact.
The liquidation of its reserves might cause temporary distortions in the bond and repo markets, but that’s just how markets work. We can’t regulate away selling. The distortion caused by a narrow bank unwind are always less than when a fractionally-reserved bank fails. Genius compliant stablecoins also enjoy bankruptcy remoteness, which is better than the bankruptcy supremacy FinTech depositors often get.
I’ve read professor Seru’s op-ed at least five times, and still don’t really understand what his point is, except to argue against change. He’s an expert on bank runs, so you’d think he’d be pro a safer form of intermediation.
Perhaps he’s just a fan of Bitcoin. How else are we to interpret this statement:
For all the hype, stablecoins haven’t transcended banking. They have replicated its tensions in new form. The real promise of blockchain was to end trust dependencies. Instead, we are doubling down on them, now under federal supervision.
Which brings me to one of my other favorite things about crypto: the more people argue against one aspect of it, the more they inadvertently end up defending another. Maybe that’s what happens when you oppose progress.
Here's the op-ed: wsj.com/opinion/can-ma…
• • •
Missing some Tweet in this thread? You can try to
force a refresh
