Skip to content

Private seigniorage, defined and estimated (includes a free Eurozone example!)

↓ Jump to responses

Download the newsletter ›

By Merijn Knibbe

  1. Introduction

Economists routinely estimate ‘public seigniorage income’, or the amount of money the government earns from its monopoly on issuing bank notes. However they do not estimate ‘private seigniorage’, or the amount of money banks make by issuing deposit money. This is remarkable as, in the Eurozone, banks create about 94% of all money.[1] Can ‘private seigniorage’, or the amount of money which deposit-taking banks make by issuing money, also be estimated? Is it useful to estimate it? Below, it will be argued that it can be defined and measured in a straightforward way using existing data. Moreover, it is also useful to estimate it, as it gives us a possibility to gauge the amount of money banks make using their ‘exorbitant privilege’. This privilege means that banks have a right to issue money at a guaranteed 1:1 exchange rate with the banknote-money issued by the government. The government also accepts this bank money for payment of taxes. Also, money created by bank A can (thanks to the 1:1 exchange rate mentioned above!) also be used to redeem debts accepted by bank B at a 1:1 exchange rate – bank lending and money creation, including the profits made by this, are clearly backed by the government in a decisive way. The concept and measurement of private seigniorage might shed some light on the magnitude of profits made. To be able to understand this we first have to analyse the definition of public seigniorage and to investigate how this definition can be extended to the private sector.

  1. Public seigniorage: definition and measurement

According to estimates of the Bank of Canada, ‘seigniorage income’ of this central bank in recent years was about 1.2 to 2 billion dollar a year, an amount which, after the deduction of some costs, was transferred to the government. The Eurosystem (the ECB plus the national central banks) earned (based on a somewhat fuzzy graph on page 20 of this publication) about 20 to 25 billion Euro seigniorage income in 2008 which, as the spread between some specific interest rates declined (see below), dwindled to about 6 billion in 2010. This money is partly (8%) transferred to the ECB and, after deduction of costs, transferred to national Eurozone governments. How do central banks estimate this income? According to the recent tract on monetary reform in Iceland by Frosti Sigurdjonsson this does not always happen in a consistent way: “The term “seigniorage” has been used to mean different things by different texts.” (Sigurdjonsson, 2014, p. 27). One can indeed find many examples of such different things in economic texts. Seigniorage is for instance often conflated with ‘the inflation tax’, but these are in reality are two entirely different animals (a government can even in a situation of mild deflation sometimes still reap a seigniorage profit). However, the way seigniorage is actually calculated is pretty clear once we take account of institutional differences.

  1. Modern definitions of seigniorage

Two examples of the calculation of seigniorage will be investigated, one by the Bank of Canada (2013) and another by the European Central Bank (ECB) (Efthymiadis et al., 2010). Note that institutional differences lead to quite large differences in the operationalization of seigniorage, while at the same time a clear common denominator exists at the definitional level. First, the statement by the Bank of Canada (emphasis added):

Seigniorage is the revenue earned from the issue of money. Historically, this revenue accrued to the “seigneur” or ruler. In Canada today, seigniorage can be calculated as the difference between the interest the Bank of Canada earns on a portfolio of Government of Canada securities—in which it invests the total value of all bank notes in circulation—and the cost of issuing, distributing, and replacing those notes…Here is a simplified example of how this works, using a $20 note, which is the most commonly used denomination. If the Bank of Canada invests the proceeds from issuing the $20 note [I.e.: the money earned when the Bank of Canada sells notes to the normal banks, M.K.] in a government security generating 2.5 per cent interest, this note will yield $0.50 per year of interest revenue. The overall production cost for the note is about 19 cents. Given an average life of about 7.5 years for a new bank note, the production cost of the note averages out to 2.5 cents per year [the Bank of Canada has promised to buy the old notes back, which means that the proceeds themselves are not a profit, M.K.] If average distribution expenses of about 2 cents per year are added to this, the total average annual cost of putting this note into circulation and replacing it when it is worn is approximately 4.5 cents. Thus, the Bank of Canada earns an annual net revenue of about 45.5 cents for each $20 note in circulation. Seigniorage varies according to prevailing interest rates and the value of notes outstanding, but has ranged from $1.4 to $2.0 billion annually over recent years. After deducting the Bank’s general operating expenses of about $446 million (of which spending on bank notes is approximately 48 per cent), the remainder is paid to the Receiver General for Canada.


So, the ‘normal’ banks lend money to the Bank of Canada at 0% interest and receive notes (a promise that the loan will be paid back) in return. The Bank of Canada invests the funds in interest bearing financial assets. The difference is (after deduction of production costs) called ‘seigniorage’.

The ECB statement is somewhat less clear but nevertheless contains comparable elements (the Eurosystem consists of the ECB in Frankfurt plus the national central banks):

As inflation moderated in developed economies in the last decade, the average seigniorage income has decreased for many central banks over time…The monetary income of the Eurosystem mainly consists of interest income arising from the liquidity-providing operations due to the refinancing needs of the banking system. Those refinancing needs primarily stem from the issuance of euro banknotes whose volume is a function of euro banknote demand. The income earned by Eurosystem central banks is linked to the interest rate applied on liquidity providing operations, which is usually close to the marginal rate of the main refinancing operations (MROs), slightly higher than the main policy rate of the ECB [in a sense this situation is the opposite of the Canadian system. Banks have to borrow at a stipulated rate of interest ‘Reserve Euro’s’ which they have to use to buy notes, M.K.] …. The monetary income consists of this interest income minus certain expenditures. The main expenditure is the interest paid on the amount of required reserves the monetary financial institutions hold with the Eurosystem. This interest rate paid is the average MRO rate such that on average the required reserves do not provide income to the Eurosystem. Any excess reserves kept with the Eurosystem are not remunerated and therefore profitable to the Eurosystem, but are typically small (<0.5%, for that same reason, because it is costly to commercial banks). The central banks only incur a small cost for printing banknotes; hence the so-called seigniorage income represents most of the Euro system’s monetary income. The ECB receives interest on its 8% share of the total euro banknote issuance. As the ECB does not put the euro banknotes into circulation, it holds a claim on the NCBs of the Eurosystem which fulfil that task. Interest on the claims of the ECB in respect of its share of banknotes is earned at the latest marginal rate (or fixed rate) for the Euro system’s main refinancing operations… [Mind that profits are not paid to the ‘Receiver General’ for the Euro Area or the European Union, which is another difference with Canada. A part of national profits are however paid to national governments, M.K.].


Clearly, in both the Canadian and the Euro Area case seigniorage income is defined as net interest income generated by issuing banknotes, part of M-3 money in circulation plus banknotes in the vaults of banks and inside ATP machines (which are not part of M-3). Interest paid on bank reserves is deducted. The congruence with issuing deposit money is obvious. MFI-banks provide deposit-money creating loans and households and companies pay interest on these loans while they receive interest on their deposits.

  1. Modern fiat money: loans create deposits

It is, alas, still necessary to stress the money creating nature of bank loans. The endogenous and largely private nature of modern fiat money creation was stressed by prominent authors such as Hayek, Keynes and Schumpeter. Money as we measure it is, to a large extent, not created by the government but by private banks (including those owned by the government, including in the UK the Royal Bank of Scotland and in the Netherlands ABN-AMRO). These banks do this by lending money which, by law, entitles these MFI’s, or Monetary Financial Institutions, to create transferable assets (deposit money) which can be used to pay back loans to the bank issuing the money but also to pay back loans to other MFI-banks at a 1:1 exchange rate and which also have a central bank guaranteed 1:1 exchange rate with ‘government money’, or banknotes and coins. These banks (or rather: the banking system) basically face no lending constraints as the assets created are also used to fund banks, at least not as long as there are increases in the value of assets used as collateral (to an amazing extent these are houses) and/or as long as the financial reputation of borrowers (non-financial companies, households) is considered good. Though this idea can be found in older, non Anglo-Saxon textbooks it seems, according to the Bank of England, for the wrong reasons to have been excluded from mainstream Anglo-Saxon textbooks, and these nowadays dominate the market. The omission applies not just to textbooks. Even the seminal work on bank lending in 17 countries between 1870 and 2011 by Òscar Jordà, Moritz Schularick and Alan M. Taylor, which as far as I’m concerned changes the nature of macroeconomics and which shows that 60% of money issuance is backed by houses, still uses a ‘loanable funds’ idea of lending (which in the end prevents them from explaining how the ‘Great Mortgaging’ could take place: they have no idea where all the money lent to home owners and enabling the financialization of our economy came from!). It therefore still seems necessary to stress the private nature of much money creation. Banks, together with borrowers, can and do create the larger amount of this thing called money and face few constraints to do this if macro prudential policies fail (as they did before 2008, except in Germany). The point to stress is that they are able to do this because of all kinds of explicit and implicit government licenses and guarantees – an exorbitant privilege! Calculating ‘private seigniorage ’enables us to gauge the amount of income generated, or at least enabled, by this exorbitant privilege.

  1. Calculating private seigniorage in the Eurozone

The statistics of the ECB (here for MFI balance sheets, here for interest rates) enable us to calculate private seigniorage in a pretty straightforward way. The non-consolidated MFI-balance sheet statistics show that the total amount of deposit money (just like, of course, the total amount of loans) is about 17 trillion Euro. Average interest rates on the amount of outstanding loans is 2.81%, down from 3.26% one year ago (non-financial corporations only, I couldn’t find the household data on the ECB site, but this statistical bulletin shows that, contrary to interest on deposits, interest on new loans to households is about the same as new loans to non-financial corporations). Interest on the outstanding amount of deposits is 1.58%, down from 1.95% one year ago for deposits with an agreed maturity (outstanding amounts) and 0.78%, down from 1.07% one year ago, for deposits redeemable at notice, so let’s take the average (1.07%). The spread between both interest rates is (2.81-1.07) = 1.74%. The outstanding amount of loans and deposits is €17 trillion, 0.0174*17,000,000,000,000 = €296 billion… (or, per Euro area inhabitant, about €1,000). That’s – a lot. It is about as much as the total government debt of a country such as Greece. However, some costs have to be subtracted, including the costs to safeguard the internet payment system etc. (as an aside – banks devote large amounts of attention to this and do a pretty good job). I do not know how high these costs are. Economic statisticians use the concept of FISIM (Financial Services Indirectly Measured) to calculate these, but this concept is calculated by using hypothetical interest rates and not by looking at actual costs. When spreads between interest rates change (as they did post 2008) FISIM becomes extremely volatile. This makes it unreliable as an indicator of true costs. Taking a well-argued percentage of the total value of the balance sheet of banks would even be better. See also Diane Coyle about this. It would take too much space to develop this argument here, but, considering i) the pre- and post-2008 events, ii) the still extremely high incomes in the banking sector and iii) the massive amounts of tax-payer money transferred to the banks, it can be stated that part of the bonuses, wages and dividends paid by the banks should not be considered ‘costs’ in the business accounting sense, but rather a rent income and consequently part of seigniorage income.

  1. Conclusion and discussion

As Minsky stated, every economic institution can create money but the point is to get it accepted. One example of a privately issued type of money are stamps: they can serve as a store of value, a means of a specific kind of exchange, and nowadays even as their own unit of account. Also stamps are accepted as money because they are convenient. Deposit money is also accepted because the government guarantees a 1:1 exchange rate between deposit money and bank notes, as well as a 1:1 exchange rate between money issued by bank A and money issued by bank B (provided that these banks have the necessary government license), while it also accepts this money as a way to redeem tax debt – an exorbitant privilege and even more so as, in reality, governments also transfer tax payer money to banks (as well as, in the Eurozone, newly printed ‘Emergency Liquidity Assistance’ money) whenever balance sheet problems arise. This means that bank profits contain an element of rents and monopoly profits, even more so as the government also guarantees and often also subsidizes property, especially the ownership of, among other assets, the houses which are used as collateral for money creating lending by banks. Using the concept of ‘private seigniorage’, an upper limit can be calculated for the Euro Area of (at this moment) about €300 billion, or €1,000 per Euro Area inhabitant, for these rents. Central banks transfer their seigniorage income to the government. Considering the government guarantees which are the backbone of the modern monetary system as well as all kind of implicit guarantees, an argument can be made that at least part of this seigniorage income should be transferred to the government (on top of taxes), a policy which has to include macro-prudential checks for the banking sector, aimed at restraining bonuses and wage income for the senior managers – as these are, in effect, not private entrepreneurs but civil servants.

Bank of Canada (2013), Backgrounders. Seigniorage.

Coyle, D. (2014), GDP: a brief but affectionate history. Princeton.

Efthymiadis et al. (2010), ‘Main drivers of the ECB financial accounts and ECB financial strength over the first 11 years’, ECB occasional studies no. 111. Frankfurt.

Jordà, O., M Schularick and A. M. Taylor (2014), “The Great Mortgaging: Housing Finance, Crises, and Business Cycles“, NBER Working Papers 20501.

McLeay, M., A. Radia and R. Thomas (2014), ‘Money creation in the modern economy’, Bank of England Quarterly Bulletin 2014-I. pp. 1-14.

Sigurdjonsson, F. (2015), Monetary reform. A better monetary system for Iceland. Reykjavik.


[1] In this article money is defined not just as M-3 money, or deposit money on short-term time deposits plus banknotes and coins but also includes deposit money stacked away in longer-term savings accounts and the like.

From: pp.8-10 of World Economics Association Newsletter 5(4), August 2015

Download WEA newsletter Volume 5, Issue No. 4, August 2015 ›

2 responses

  • Marco Saba says:

    I am sorry but I can’t understand why you don’t calculate the seigniorage in the old straight (and right) way: the difference between the nominal value of the currency and his intrinsic cost. You cannot make a real difference between coins, banknotes and electronic funds other than for confusing the public. I.e. this article miss the point of the seigniorage on capital creation. If a coin cost 10 cents but the nominal value is 2 euro then the seigniorage is 2 euro minus 10 cents: 1.90 euro.
    If a banknote of say 500 euro cost 20 cents, then the seigniorage is 4.80 euro. If a loan of 200,000 euros cost 2,000 euro of ‘general bank expenses’, then the seigniorage is 198,000 euro.
    While the bank use accounting tricks to avoid registration of this value (the bank didn’t write the sum just created in the asset side of the cash flow before using the sum for lending AND open market operations, like buying a new skyscraper for their headquarters, but it register the sum as a fake overdraft on herself), the economists should not be confused and the bank accounting should be audited in the right way.
    As seigniorage didn’t change as a concept if you use paper, metal or any other material to issue the tangible symbol of the money (coins, or banknotes), it didn’t change also if you use digit on a computer – which are less expensive than other raw material. I know that the central bank apply black magic wishful thinking to hide the seigniorage from customers view – else every one would want to take a share of the gorgeous money pie (or maybe pretend to have a sort of universal basic income just to stay alive) – but we as economists should not be hypnotized by their weird and inappropriate seigniorage definition. We may be corrupt, yes, but not stupid.

  • Victor says:

    nice paper … and a well protected secret.
    The exact same calculations can be found in the book of Helmut Creutz “The Money Syndrome”. A good read is also Frederick Soddy from the 30’s, “The Role of Money”, and many, many more.

Respond to this article

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Please note that your email address will not be published.