Most Recent Stories

Potential Problems with Narrow Banking

Here’s an interesting post from John Cochrane on “narrow banking”. Narrow banking is the idea that we could create banks that take a “narrow” type of risk by investing only in very high quality assets like Central Bank reserves or government bonds. This would back the bank’s balance sheet with virtually risk free assets. The argument is that this would be ideal compared to modern banks because there would be no risk of runs and no need for bailouts.

As the Cochrane post discusses, the Federal Reserve has apparently rejected an application for a narrow bank (TNB) which has subsequently resulted in a lawsuit (see here). You might wonder why the Fed would reject an application for an entity that just wants to be a super safe bank, but I think there are strong arguments against this form of institution. I’ll explain some of my own concerns below, but I also reached out to a good friend of mine who works at a regional Federal Reserve bank and he explained to me why he finds the idea misguided. He asked to remain anonymous so I will pass on what he said:

“Narrow banking does not ensure low risk banking. A pure commercial bank is as risky as the loans it makes. Requiring a bank to fully reserve its deposits does not mean the bank’s total assets are safer. It just means their deposit liabilities have a safety net. For instance, if XYZ Bank issues bad loans during the financial crisis then the value of their assets relative to their liabilities will decline regardless of how safe the deposit liability side of the balance sheet is. In other words, narrow banking does not equate to narrow lending.”

Cullen here. That’s pretty clean. It’s oversimplified for emphasis, but you get the basic gist. The TNB bank discussed in Cochrane’s post also isn’t a real “narrow bank” so my friend is speaking of a true narrow bank which extends loans as opposed to TNB which just takes deposits and operates more like a payment processing money market fund. But the same basic points apply.

This “narrow bank” idea has always reminded me of the idea of a full reserve system. We know that banks issue loans and find reserves after the fact (see, The Basics of Banking). So, if a system is fully reserved then that just means the Central Bank must supply reserves when banks create new deposits. This doesn’t mean the loans themselves are safer. It just means the Central Bank is fully reserving the deposits.

The question of the safety of our banking system really comes down to how much risk we want banks to take. We need them to take some level of risk in the process of supplying liquidity to the economy, but we also need them to do that with certain constraints. Those constraints are a regulatory measure on the asset side of the bank’s balance sheet and so supporting the deposit liability side of the balance sheet is essentially an ex-post solution to a problem that arises before reserves and “safe” assets ever come into play.

I have two other thoughts on this topic that I think the Fed could be considering in the scope of their rejection of such an application:

  1. The Fed can’t risk relinquishing control of monetary policy.
  2. The Fed doesn’t actually want all banks to be safe and wants the distribution of safe assets to be somewhat diverse.

If the Fed were to approve a Narrow Bank like TNB, which is just a savings institution and not a lender, then the Fed would be approving a de facto reserve quantity setting institution. For instance, if this Narrow Bank became extremely popular and took in $5 trillion in deposits then the Fed would have to supply those reserves since they are the monopoly supplier of reserves to the banking system. This would handcuff the Fed’s balance sheet size to this narrow bank.

In other words, if the Fed ever wanted to reduce their balance sheet size then they would be unable to do so since they would need to accommodate the narrow bank’s reserve facility. This makes no sense from the perspective of the Federal Reserve. Why would they relinquish potential control of their balance sheet to a private bank by essentially agreeing to fully reserve this one bank?

The other potential concern here has less to do with TNB specifically and more to do with how “Narrow Banks” would operate more broadly. The Fed might not want fully reserved banks for worry that they would be too reliant on the Federal Reserve instead of broader risk management measures. For instance, a large narrow bank could monopolize cheap funding sources across the banking system thereby benefiting from Fed support without taking any risk. So, if TNB collected huge quantities of deposits then that’s that much less cheap funding for other banks. This would be making a small part of the banking system more dependent on the Fed and therefore safer while potentially destabilizing the rest of the banking system even though it was their loan creation risk that created all those deposits in the first place! Or alternatively, you end up with a narrow bank that is very safe, but can’t compete so it goes out of business because it doesn’t take enough risk.

Think of how this might play out in reality – TNB would collect huge amounts of deposits in times of danger when they’re viewed as being safer and then they’d struggle to earn a profit as the business cycle ramped up and other banks made riskier bets. This could result in more runs when the Fed most wants to stop runs and make the narrow bank less safe during times when all other banks appear safer. I don’t see how this helps the Fed’s role in the banking system. Narrow banking, like full reserve banking, sounds good in theory, but doesn’t solve all the problems its proponents seem to think and in fact could create unforeseen risks.

I know it’s popular to hate on banks and I have plenty of hate for certain bank activities, but we need risk taking liquidity providers in the economy. Banks are entities that compete over risky assets (mostly loans) in the process of supplying liquidity to the economy. Yes, at times they make mistakes, but for the most part this system of competitive private banking works pretty well. The alternative system is to have politicized bureaucrats creating new loans in a non-competitive system. Would that work out any better? I doubt it.

Now, could the existing system work better? You bet your ass it could. And yes, requiring banks to hold higher quality assets is a big part of making improvements. But again, that’s a regulatory issue that begins at the time of loan creation and risk taking as opposed to supplying a safety net after the fact.