COMMERCIAL BANK MONEY AND SEIGNIORAGE POWER

Don’t banks extract seigniorage (as governments and central banks do) when they create money via lending?

COMMERCIAL BANK MONEY AND SEIGNIORAGE POWER

by Biagio Bossone and Massimo Costa

What is "seigniorage"?

Derived from the old French for "seigneur", which stands for "feudal lord" (the lord of a manor), the word "seigniorage" according to the Oxford Dictionary etymologically means the "right of the lord to mint money" and economically denotes the difference between the value of money and the cost of producing it, where the former equals the real resources that money can buy or the economic benefits that lending it can earn.

Seigniorage has historically belonged to governments, which over the ages have appropriated real resources through the monopoly of the coinage. It is thus a form of taxation, which the state imposes on the economy and implies a net transfer of real resources from the economy to the state. In contemporary economies, the monopoly of the issue of legal tender is assigned to an agency – the central bank – which may operate under varying degrees of independence from government. The seigniorage on the money issued by the central bank accrues to the latter, and part of it is returned to the state.

Seigniorage differs from the profits extracted by intermediating pre-existing resources, and ultimately rests -like any rent – on the existence of a (quasi) monopolistic power. However, whereas a (quasi) monopolist typically holds exclusive control over an existing (scarce or even irreproducible) resource, the creator of money creates the resource itself by fiat.

The question of interest is: given the intimate relationship between money creation and seigniorage, isn’t it the case that commercial banks, too, extract seigniorage from the economy (much as government and central banks do), once it is recognized that they create by far the largest share of money used in the economy?

One of us started investigating this question many years ago as part of a World Bank research program on financial sector issues (see here), while the other approached the same question from the standpoint of the nature and classification of financial statements items, being dissatisfied with mainstream views especially on banks’ balance sheets (see here). Recently, we have joined forces and further enquired into the issue of commercial bank seigniorage in the context of the Accounting View of Money that we are developing together (see also our Economonitor commentaries here and here).

Below, we share some of our reflections on the topic, noting for clarity purposes we use the general definition of "commercial banks" (henceforth "banks") institutions that accept demand deposits from the public and lend money to borrowers in the form of demand deposits.

Do banks extract seigniorage?

As banks create money and their money is only partially backed by (costly) higher-powered money (central bank reserves), they extract seigniorage from the economy much as the state does. By creating money, banks provide the economy with the ²public good²that is needed to finance economic activities, thus affording greater elasticity to the economy and ensuring decentralized money supply services to dispersed demand. Yet, while these value-adding services could be rewarded by fees and commissions, and the credit risk involved in lending it could be covered by collateral, the interest income that banks earn on the money they create and lend does not reflect any rewards for savers’ thrift (as in neoclassical economics) or for agents parting temporarily with liquidity (as in Keynes’ liquidity preference theory). In fact, one could conceive of decentralized money supply regimes where the provision of money is separated from lending and seigniorage is socialized, that is, redistributed to society (see here).

In fractional reserve regimes, commercial banks create the money they lend in the form of newly issued demand deposits. Commercial banks create their own demand deposit liabilities by crediting borrowers’ accounts with the funds loaned to them; as a result, new liabilities and assets are simultaneously recorded on the bank’s balance sheet. In fact, banks create money not only when they lenddeposits but also when theysellthem, for instance to (re)purchase securities or in exchange for cash – although in the latter case there is no new netmoney creation in the economy. Lending deposits features close analogies to selling deposits. As banks issue deposits to customers in exchange for cash or funds, banks become owners of the cash and funds received and acquire the rights to use them as they wish (subject to existing laws and regulations). Even if the banks are constrained in the use they can make of the cash and funds received, such as, for instance, in the case of regulation prescribing the types of assets to be held, they (not the depositors) are the owners of the purchased assets and they (not the depositors) are the owners of the income generated through the assets purchased.

Only commercial banks, not pure financial intermediaries, may create money since they only are allowed by regulation i) to hold customers’ demand deposits andto issue loans to customers in the form of demand deposits and ii) to carry both operations on customer accounts held in their own books. Pure intermediaries, on the other hand, may only lend pre-existing funds (savings) and can perform lending only through transfers of funds held in bank accounts. That is: when a pure intermediary intermediate funds, say, by accepting funds from customers against the issuance of term deposits and by extending mortgages to homebuyers, the intermediary receives the funds from customers in the form of demand deposits (which it holds in a bank account) and transfers those funds to the homebuyers in the form of demand deposits (drawn from the same account). In such an example of pure financial intermediation, there is no money creation and the funds lent to the homebuyers are the same funds that were originally saved by the intermediary’s customers. Notice that all fund transfers in the example (i.e., from the customers to the intermediary, and from the intermediary to the homebuyers) take place across accounts held with banks.

Importantly, when extending loans banks need to avail themselves of the resources needed (by interbank conventions, payment system rules, or central bank regulation) to settle the obligations that are triggered when the loaned deposits are mobilized by customers or when they are redeemed in exchange for cash. The settlement resources needed by commercial banks consist of

  1. cash reserves and reserves deposited with the central bank
  2. reserves from settlement of incoming payments from other banks
  3. borrowings from the interbank market
  4. borrowings from the central bank
  5. immediate liquidation of unencumbered assets in the balance sheet
  6. new deposits of cash from old and new customers (notice that new noncash deposits may consist only of deposits transferred from other banks, which as such fall under item 2 above).

Acquiring and holding such resources are costly activities for banks and detract from their seigniorage. Thus, while banks may in principle create all the money that the economy is willing to absorb (in the absence of institutional limitations or constraints), in practice their money creation is a function of the price at which money is lent (the interest rate), the credit risk inherent in lending it, and the cost of funding (that is, the terms and conditions at which central bank reserves are made available to banks). As James Tobin argued in his famous Commercial Banks as Creators of 'Money', lending will be extended until "…the marginal returns on lending and investing, account take of the risks and administrative costs involved, will not exceed the marginal cost to the banks of attracting and holding deposits" (p.9).

Yet, four factors allow banks to economize on the use and cost of reserves vis-à-vis the deposits created, thus enhancing banks’ money creation capacity and incentives and increasing the associated seigniorage. These factors are:

  • The public’s desire to hold commercial bank deposits that pay less interest than bonds. This is due to the liquidity of demand deposits, as supported by the whole complex of institutional, technical and policy solutions that support public trust in their use as money (which includes inter alia payment infrastructures, banking supervision, central bank liquidity support, deposit insurance, etc.). This factor allows banks to save on the cost of liabilities vis-à-vis other types of financial intermediaries.
  • *The operating scale at which individual banks operate.*In fractional reserve regimes, banks hold only a fraction of reserves against their total deposit liabilities. Similarly, the volume of reserves they actually use for settling interbank obligations and for cash withdrawal from customers are only a fraction of the total transactions settled. Economies of scale allow banks to further economize on the use of reserves. They relate to i) the share of total payment transactions that a bank intermediates: the larger the share of incoming payments (and, hence, of settlement reserves received) the bank receives from the others, the less it needs to raise and hold reserves to settle outgoing payments, and ii) the number of depositors: this allows the bank to settle a larger share of payments in its own books (“on us” payments), to expand its sources of reserves through incoming transfers and payment, and to benefit from asynchronous withdrawals of deposits from depositors. Consolidation generates increasing returns for banks, enabling them to create liabilities (by lending or selling deposits) with decreasing reserve margins needed for coverage. Thus, the larger the deposit base (scale) of a bank, the higher the seigniorage the bank can extract from deposit creation. Economies of scale in the use of reserves combine with scale economies in bank production, which empirical research show to be significant (see, for instance, here, here, and here).
  • The market power of each bank. If free entry were allowed into commercial banking, seigniorage would be reduced accordingly, as banks would compete with each other on the liability and asset side of their balance sheet and seigniorage would decline. In fact, policy regulations restrict the number of bank entries into the market, preserving their financial soundness as well as their franchise. Commercial banks do extract seigniorage as a consequence of such regulatory restrictions. All else equal, seigniorage is higher in more consolidated banking systems where banks exercise their quasi-monopolistic power and extract larger rents from higher interest rate margins.
  • *Payment system settlement rules and technologies.*While the operating scale of individual banks, as discussed above, affects each bank’s optimal demand for reserves at the margin, payment settlement rules and technologies determine the structural need for reserves of the whole system. Payment system rules affect the demand for reserves via two channels: the settlement modality (typically netting or gross) and the technology adopted. Each of the two modalities has a drastically different effect on the system’s use of reserves, with netting requiring much less reserves than the gross modality (at the cost of higher credit risk from settlement). Similarly, by re-introducing elements of netting into the gross modality, modern technologies (such as liquidity-saving mechanisms) increase the velocity of reserves circulation in payment systems and allow banks to economize on the use of reserves for any given volume and value of payments to be settled. Both factors interact with payment system scale economiesand, all else being equal, affect the overall level of seigniorage that commercial banks may extract from the economy. Significant scale economies exists can be observed in payment systems (see, for instance, here and here).

As an exemplification of the above factors, take the hypothetical case of a cashless economy with a fully consolidated banking system where all agents hold accounts with only one bank. In this case, all payments and money transfers would be "on us" for the single bank. The bank would need no reserves for settling transactions, it would be under no debt obligation towards its customers, and it might in fact create all the money the economy would be willing to absorb at a given interest rate without having to hold (costly) reserves. The single bank's deposit liabilities would become the accepted instrument to settle debts within the economy and would de facto be irredeemable, much as legal tender is any economy.

Obviously, in reality systems are populated by a plurality of banks. Yet, in a highly consolidated system the largest banks (and the system as a whole) feature significant economies of scale in reserves use and extract the extra seigniorage that such economies make possible vis-à-vis less consolidated systems.

What next?

In fractional reserve regimes, commercial banks create money and thus extract seigniorage.

The point is how relevant this seigniorage is and what its impact is on the economy – questions whose answers, of course, may vary from country to country. Both are questions that require additional theoretical and empirical research to identify carefully all sources of bank seigniorage, and those that detract from it, to determine their quantitative relevance (in particular in relation to various degrees of banking sector concentration and competition across different jurisdictions), and to understand the channels through which seigniorage affects the economy.

In our work referred to earlier, using international accounting principles, we show that bank seigniorage in the UK during the period 2007-2016 averaged more than 10% of the country´s tax revenues, and similarly high estimates are obtained by Macfarlane et al (2017). While more granular analysis would be required for gauging better estimates, these preliminary attempts suggest that bank seigniorage is not empirically irrelevant.

Attention should then be devoted to the incentives effect of bank seigniorage on the economy’s output and resource distribution. Very preliminary research suggests that bank seigniorage may affect, inter alia, such variables as labor productivity growth, inflation, the relative price dynamics of wages and capital goods prices, and capital intensity (i.e., capital endowment per worker).

While more precise statements on these relationships will have to await the results from the ongoing analysis, the purpose of this commentary was shed some light on a topic for research, which to date has been almost totally neglected by the economics literature.