Narrow Banking with Modern Depository Institutions

Since the financial crisis of 2007-2009, much has been studied regarding its impact on the economy and what could be done to impair its recurrence in the future or, at least, to reduce the effects of similar periods of financial instability. In particular, one idea is to develop efficient schemes designed to protect deposit holders since, in the context of a financial crisis, individuals can have the incentive to withdraw all their savings from banks creating a cascade effect that is known in the financial literature as a bank run. These runs can literally drive banks into insolvency, potentially taking a fragile economy to a further level of instability.

Over the last years, narrow banking has been attracting attention as a possible scheme to reform our financial system. In general, narrow banking is a term that includes different proposals designed to limit the ability of banks to extend credit. However, all these arrangements share at least one policy item: the obligation for depository institutions to hold 100 percent of their deposits in reserves.

Regrettably, the economics profession has not reached yet a consensus about the likely effects of imposing a 100 percent reserve requirement. In fact, positions seem completely confronted. To see why this is the case, we can review the theoretical model both sides of the discussion use to support their conclusions. In this view of the banking industry, depository institutions are seen as accepting deposits on the liability side of their balance sheet in order to channel these funds to borrowers who, in turn, use these loans to finance their expenditure decisions. In this banking model, the amount of intermediation banks can perform is consequently constrained by the total assets depositors have accumulated so far. Importantly, within this view, liquid but unproductive reserves are competing with productive but illiquid loans on the asset side of bank’s balance sheets. Thus, a corollary of narrow banking results from the analysis of both proponents and detractors of the proposal, namely, that fully backed deposits leads to an automatic separation of financial firms: one class being depository institutions, fully liquid and not able to provide loans; the other investment banks, able to lend but prone to instability.

For narrow banking advocates this separation between pure depository institutions, exclusively dedicated to the payment system, and pure investment banks, exclusively dedicated to lending, is a necessary condition to avoid bank runs and to ensure financial stability. On the other hand, narrow banking detractors contend this measure will lead to unbearable efficiency losses as forcing banks to maintain deposits in the form of idle reserves will prevent them from funding loans with the corresponding reduction in the financing of productive investment.

BSE Working Paper No. 955, “Narrow Banking with Modern Depository Institutions: Is there a Reason to Panic?” by Hugo Rodríguez Mendizábal investigates (i) what would be the likely effects on lending of imposing narrow banking to our depository institutions, (ii) whether that policy could be implemented, and (iii) the extent to which it will be distortionary.

The novel feature of the analysis is to abandon the view that banks channel existing deposits into loans. This view is at odds with current practice by depository institutions. Instead, in a modern financial system, deposits are themselves created whenever a bank provides a loan. The payments associated with this loan provision imply that those deposits the loan creates are to be transferred to someone else’s account, possibly at a different bank. Banks settle these transfer orders by means of reserves which are current accounts they have in the central bank. Generally speaking, these reserves are created through a loan from the central bank.

This description of a modern banking system highlights several ideas. First, deposits are not a constraint but an outcome of loan provision. Once deposits are initially created from a loan, they start circulating in the economy as agents make payments among themselves and will keep circulating as long as the loan that originated them does not mature.

Second, there is no sense in which reserves are competing with loans on the asset side of the bank’s balance sheets. Instead, reserve demand is the consequence of deposit creation and, therefore, of providing loans. Reserves and deposits (or the loans that created these deposits) are two layers of liquidity over imposed to each other and used by different agents.

In principle, a bank could go on providing loans, creating the corresponding deposits and demanding whatever level of reserves deemed necessary without the reserve demand curtailing loan provision directly. Thus, as a third conclusion, narrow banking in modern monetary systems does not necessarily call for a separation of financial institutions.

Finally, because reserves should be borrowed from a central bank at a price, higher reserve requirements may indirectly affect loan and deposit creation as they influence the costs to run the main business of banks – namely, loan provision and payment services. These costs can be seen as imposing a tax on the bank business model and should affect the loan-deposit rate spread, which in turn could potentially introduce inefficiencies in the intermediation done by banks. However, these costs could be driven to zero if, as the Eurosystem does, required reserves are remunerated at the same rate as reserves are loaned out in the first place. In such a case, required reserves should not have any influence on liquidity provision by banks through this cost channel.

In the author’s setup, the only way forcing banks to hold more reserves could affect their loan provision is through the collateral constraints of central banks. When borrowing reserves from the central bank, the commercial banks must pledge assets as a guarantee for that loan. Because the overall supply of assets accepted as collateral is exogenous to the banking system, current collateral regulations could effectively constrain the implementation of narrow banking. The paper discusses possible ways to go around this situation.

In short, Rodríguez’s contribution is quite relevant for the banking literature, particularly in the present moment of recovering from the last financial crisis and with the ongoing discussions about the creation of a banking union in Europe. Showing how narrow banking could be implemented without affecting intermediation by depository institutions, this work provides policy-makers with an additional tool to increase the robustness of our monetary systems.