Stablecoins, narrow banking and the demand for safe money
Legislators should consider using the crypto regulation bill to end the Fed's arbitrary war on narrow banking
Earlier this summer, Senators Gillibrand (D-NY) and Lummis (R-WY) introduced a crypto regulation bill that would, among other things, attempt to bring crypto currency under the US regulatory umbrella. Most bills don’t become law, so I wouldn’t hold my breath that the exact language in the bill becomes law. But even if it’s not this bill, the sector is too large to forever evade the oversight of the CFTC, SEC or some other body.
One major provision that caught my eye is the requirement that stablecoins post a 100% reserve requirement on deposits in the form of specified “high-quality liquid assets”. Not a stablecoin backed by an algorithm (like Terra), or by commercial paper (like Tether). By actual reserves, like US currency, short-dated government debt, and the like.
Such a coin, if operated honestly and brought under the oversight of US regulators, would thus be highly similar to the long-proposed narrow banks. The Fed has historically been hostile to such organizations–and even recently rejected an application by one for a master account at the Federal Reserve–and it’s worth examining why this is the case. Overall, I think the case against allowing narrow banks is incredibly weak, and I hope that as this bill gets revised that legislators considering extending the logic of responsible stablecoin regulation to the realm of narrow banking.
What is a narrow bank?
There’s a stylized version of the banking system that looks something like this:
Banks accept deposits from regular people like you and me.
They issue bank loans backed by these deposits. Those loans then become someone else’s deposits, which get loaned out as well. As a result, $1 in deposits can become, for example, $10 in actual money throughout the system
While you can withdraw your money at any time, the bank does not actually have 100% of all the deposit claims it holds in its vaults at any given time. They have some fraction of that.
If everyone really does demand to withdraw all their money at once (a bank run), then the FDIC will make sure you can get your money back (up to $250,000)
Theoretically, since you know your money is insured, then there’s no reason to demand all your money back when times get tough, so the FDIC never actually has to act on its insurance promises
A narrow bank operates differently. They wouldn’t actually lend out any of their money. Instead, they back 100% of their deposits with actual reserves. There’s no need for FDIC insurance or anything else. If everyone wanted their money back, they could just all get their money back. In its ideal state, a narrow bank would just set up a master account at the Federal Reserve, collect the interest on reserves from the money stored in that account, and pass through all of it to their customers (minus some small profit).
There’s an obvious parallel with money market mutual funds, which are just funds that invest your money in highly money-like assets (like short-dated treasuries). The subtle difference is that you can use your deposits in a narrow bank for transactions (e.g. you can have a debit card linked to a narrow bank) so there’s a modest transactional benefit.
There’s also an obvious parallel with stablecoins. In their most basic form, a stablecoin is a cryptocurrency that is 100% backed by cash-like assets such that it always equals a dollar (sometimes it pays you a modest interest rate). The benefit is purely transactional, in that it allows you to easily transfer between different cryptocurrencies without interfacing with the regular US financial system. A narrow bank is just like a stablecoin, but for the regular world.
The case for narrow banking
The case for narrow banking is fairly straightforward. For consumers, a well-regulated narrow bank offers first and foremost guaranteed safety against runs, especially for customers with deposits over the FDIC-insurance limit of $250,000. But more significantly, they can also offer higher returns. The current rate of interest the Fed pays on excess reserves held at the Fed is 1.65%, while the average checking account pays less than 0.05%. A narrow bank–assuming it has meaningful competitors–functionally becomes a “pass-through interest entity” that takes the interest the Fed pays and transfers it to retail customers. Ideally, competition with the existing big players would also force them to raise the interest rates that they offer on deposit accounts as well, to the benefit of savers.
This does beg the question of: “aren’t there already safe, money-like assets?” The answer is of course yes. Buying ten-year treasury bonds is pretty close to safe, as are money market mutual funds. But those both are fairly hard to use as transaction mechanisms, as you first have to convert your money into regular bank dollars before using them for spending. It’s admittedly a modest change, but it’s worth flagging when we get to the potential drawbacks that the smaller the change from other safe assets, the smaller those drawbacks are as well.
For the financial system overall, the benefits are clearer. A narrow bank is run-proof, so the Fed and the Treasury would never need to worry about having to bail it out in a financial crisis. Indeed, it can even access as a counter-balancing force by preventing a debt-deflationary spiral. Let’s revisit the stylized story of the banking system from before, where loans create money in the system (e.g. person A puts $100 in a bank, which lends it put to person B, who deposits it in their bank; while assets minus liabilities is still only $100, there’s $200 in the total system; the loan created an extra $100). In this system, a loan default destroys money. At a large enough scale, vast defaults should thus be deflationary. Since deflation raises the real value of debt (since debt is nominally determined, a $100,000 debt in a deflationary environment is harder to pay off), then even more people default causing even more deflation, causing even more loan defaults. This is a decently accurate (though highly simplified) story of the Great Depression, and it simply isn’t possible with a narrow banking. By decoupling money creation from bank lending, there’s no risk of financial contagion through this channel. One job market paper from Princeton’s Sebastian Merkel suggests that while a 100% narrow banking system could lower economic growth by around 1% (more on this later), the benefits from reduced financial instability are enough to boost welfare by around 6.7%.
The Fed’s arbitrary and capricious war against narrow banking
Now, a 100% narrow banking system isn’t on the table. Rather, would-be entrants simply want the option to compete with the existing banks. But the Fed has curiously put a stop to this, rejecting the application of one narrow banking applicant (“The Narrow Bank”) to have a master account at the Federal Reserve. Economic John Cochrane has a good rebuttal of all the flaws in their reasoning (available here), but I wanted to highlight a few common arguments.
First, the Fed argues a narrow bank could compromise the ability for the Fed to safely conduct monetary policy. Specifically, the introduction of a narrow bank would increase the secular demand for reserves, which would limit the Fed’s capacity to control its balance sheet (reserves are a liability on the Fed’s balance sheet). In a recession, people will flee their existing banks for the narrow bank’s reserves. In good times, the higher interest rates they can get might mean people want to park their money in a narrow bank which is 100% backed by reserves.
This is completely backwards. If I take my money out of my bank account at Bank A and put it in a narrow bank, I’m not showing up with a briefcase full of cash to deposit at the Fed. Instead, at the end of the day, Bank A is transferring some of the reserves it holds at the Fed to the narrow bank. The total amount in the system stays constant.
Also, the Fed has complete fiat control over the size of its balance sheet. If they want to reduce it, they could just sell the assets they have on the balance sheet to other banks (which would raise interest rates) until the size is what they want.
But at a deeper level, why is the Fed targeting the size of the balance sheet? The size of the balance sheet is a means by which to communicate the stance of monetary policy. It doesn’t actually mean anything by itself. It may be that under a narrow banking regime, you might need a larger balance sheet to maintain rates at the previous level. But that’s not actually a problem? There may be some political aversion to having a $6 trillion balance sheet, but frankly that squeamishness is not a great reason to oppose them.
Second, a narrow banking system could lower the supply of loanable funds or shrink the money supply. If loans create money, and now there’s less money by which to create loans, that could be fairly deflationary (though as above, the Fed can counteract that by expanding their purchases and growing their balance sheets until the interest rates return to where they were prior). But there’s a deeper argument. The higher interest rates that banks would have to give to depositors to compete with the narrow banks would probably raise the amount they charge on loans. And if they fear that people might flee to the safe haven of the narrow bank during choppy economic waters, the banks might voluntarily hold greater excess reserves, further lowering the amount of money available for lending.
Economist Scott Sumner thinks that the above is the real reason the Fed opposes a narrow bank. In short, the Fed is blocking competition for banks and securing them excess profits out of the reasonable belief that these banks are engaging in socially significant behavior. It’s worth flagging of course that this proposal does not end all lending activity. Investment banks, for instance, engage in vast amounts of lending activity despite accepting no deposits: they back themselves through equity financing.
Merkel also has a clever argument for why total supply of loanable funds should be equal under both systems in the long-run (though flagging that “long run” does a lot of work in that sentence). The short version is that the risk premium offered by the loans is the same in both systems. But since the supply has declined under a narrow banking system, the remaining lenders could extract a greater excess return. Now flush with extra cash from that excess return, the supply of loanable funds ticks up a bit and the excess return shrinks in the next period. Over time, the excess return shrinks bit by bit as the total supply of loanable funds makes the interest rate banks charge on loans converge closer towards the risk premium.
Third, some worry about flight risk. During a crisis, people might start mass-withdrawing their money from normal banks and putting it in a narrow bank, causing a bank run. This risk feels overstated to me: in a crisis, people are clearly highly privileging safety over transaction value. So there’s no real marginal benefit of transferring to a narrow bank versus e.g. treasury bills or a money market mutual fund, which they can do already.
But admittedly, there are still some growth costs in the short-run. Is that worth prohibiting what appears to be a completely legal bank from offering safe assets to consumers? I don’t really think so.
If I can be uncharitable for a second, I think some of the reticence is political. If people held their money in narrow banks, the pass-through from the Fed interest rate to their personal deposit account would be near one-to-one. It does put the Fed in an awkward situation when they need to cut rates and depositors immediately feel the pain in the pocketbook as the money they earn on their deposits drops. The existing intermediation obscures that relationship sufficiently in the status quo that it might give the Fed more leeway to act decisively to stave off recessions in the future.
Next steps
The Lummis-Gillibrand bill is encouraging by recognizing the important role that money-like assets like stablecoins play in the crypto ecosystem. They also recognize that the most important feature of those stablecoins is that they actually have to be stable. There can’t be some algorithm that is stable only until someone tries to short it. They can’t mostly back their deposits with kinda safe assets. For all the byzantine nature of the US financial regulatory apparatus, it exists to address a real problem that retail bank customers have a hard time credibly monitoring their banks’ finances and therefore it’s important to make sure they are not going to get scammed out of their deposits when the market turns. When stablecoins start hiding what’s on their balance sheets or putting it in decidedly non-stable assets (and then lying about it), that’s a genuine threat to consumer safety.
It would be nice to extend that same logic to narrow banks. People have a real demand for safe, money-like assets that offer reasonable returns. Why should those privileges be limited to the crypto world? This bill offers a great opportunity to bring narrow banks into the legal fold and I hope during mark-up sessions that legislators seriously consider their merit.