Immediate money and mediated money

                              Vlastimir Vuković                           

         Immediate money and mediated money       

 Two primary forms of money and their concealed differences 

(full PDF version)

Discussion Paper, No. 5, Central Bank Money Research, November 2024

Abstract

Money has two primary forms – immediate money and mediated money. Decisive distinction is that the former circulates off-line, while the latter in-network, via intermediaries. Immediate money accomplishes its functions without intermediaries, circulating hand-to-hand (ingots, bars, coins, banknotes) or device-to-device (in the future). Mediated money exclusively circulates account-to-account in books of payment institutions. It implies that money stays in-network, i.e. in banking system. Due to clearing and settlement technology it amy appear that payments in-banks are performed without money or with little money. All other functions of mediated money are also realized in-network only. These two forms of money have circulated in parallel, intertwined and mutually conditioned during their multimillennial evolution. Three basic channels of creation and circulation of money have been built in this process– factor incomes channel, asset channel and credit channel. 

The analysis of clearing and settlement process is the key, because it reveals concealed nature of mediated money and differences to immediate money. The quantitative analysis based on BIS-BCBS practical example for intraday liquidity of banks discovers ‘secrets’ of deposit and credit multiplication, fractional reserve banking, commercial bank money, bank run, lender of last resort and the power of clearing and money settlement. Therefore, two final suggestions of this research have resulted: RTNS* solution (sequential netting) of problems intraday/overnight liquidity in banks and balance sheet reconstruction of central bank money for public.     

Key words: money evolution, immediate money, metal money, intrinsic problems, mediated money, book money, payments, clearing, settlement, banknotes, liquidity, banks, central banks, reserves, payment intermediaries (PSPs), electronic/digital/quantum money, RTGS, accelerated sequential netting, RTNS solution, balance sheet reconstruction, central bank money.

1 Introduction

Coins and banknotes are clearly visible, tangible, widely distributed and ubiquitous in everyday life. Transaction accounts with banks also seem understandable to most of their holders. Nevertheless, economic explanations of money nature and their flows and stocks are mostly different and often completely opposite. Therefore, since the 19th century and Stanley Jevons to the present day, most definitions of money rely on the three most frequently cited functions – medium of exchange, unit of account and store of value. Occasionally, several additional functions are mentioned – means of payment, standard or measure of value, and means of deferred payments.

One of the fundamental causes of controversial definitions and opposite economics explanations is the neglect of two primary forms of money – immediate money and mediated money. The decisive distinction is that the former circulates off-line, while the latter circulates in-network, via intermediaries. This assertion is an axiom, so obvious that it is needless to prove it. At the same time, it is the premise on which the analysis presented in this study is based.

Immediate money performs its functions without intermediaries, circulating hand-to-hand (ingots, bars, coins, banknotes) or device-to-device (announced off-line CBDC). Of course, an initial issuer (authorized mint, government, central bank, and others) and distributor (money-changer, bank, ATM, money provider) is needed. The payment function of immediate money is performed instantaneously, finally, irrevocably and unconditionally. The stock of immediate money is always in the money-holder’s possession. If the holder entrusts their stock of coins and/or banknotes to a deposit institution, it transforms into mediated deposit money, which can circulate only in-network via selected intermediator.

Mediated money circulates exclusively account-to-account in the books of payment institutions (templ, trapezite, argentarius, bank, central bank, PSP). The basic issuer of modern mediated money is the central bank, and it circulates between commercial banks. Payment instructions may be given using instruments or devices in paper form (check, payment order), plastic (payment cards) or electronic (PC, smartphone). Regardless of the physical form of payment instructions, transfers are most often made digitally today. Payments are processed internally (in-bank) and via clearing houses and settlement systems. Due to clearing and settlement technology, it may appear that payments are made without money or with very little money. However, in every transaction, the payer pays a specific amount of money, and the same amount is received by the payee on their transaction account. The certainty and finality of payment depend on the liquidity of the payer and their intermediator. The stock of mediated money is always in the possession of the intermediaries. Intermediaries are not custodians, but can freely use the deposited money.

Over thousands of years of money’s evolution, long before the first Lydian coins in the 7th century BC, these two forms of money circulated side by side, intertwined, and conditioned by each other. Their joint evolution often concealed their fundamental differences, caused by their circulation mode – off-line opposite in-network.

The decisive difference lies in the power of clearing and settlement in payments, although many well-known textbooks on monetary economics and banking do not even contain the words clearing and settlement. This power has relativized the importance of money stock in modern payment system to the point of unrecognizability, so that the household sector, which holds more than half the value of all transaction accounts, makes 200 times fewer payments than wholesale transactors (Table 1, source: BIS). The effects of payments in-network with mediated money are illustrated by a quantitative analysis based on BIS-BCBS practical example for intraday liquidity of banks. The presented analysis discovers the ‘secrets’ of deposit and credit multiplication, fractional reserve banking, commercial bank money, bank run, and lender of last resort. Hence, two final proposals emerge from this research: the RTNS solution (accelerated sequential netting) of problems of intraday/overnight liquidity in banks and the balance sheet reconstruction of central bank money.

A clear understanding of the two primary forms of money through their multimillennial evolution requires evolutionary and dynamic monetary analysis. Respective analysis implies explicit differing of direct exchange opposite wholesale barter, weighed metal money opposite coins, retail payments opposite wholesale payments, transaction accounts opposite deposits (time, saving), money stocks opposite money flows, money issuers opposite payment intermediaries, as well as money circulation hand-to-hand (off-line) opposite account-to-account (in-network).

The analysis includes the main channels of money creation and circulation: factor incomes channel (wages, profit, rent, interest), asset channels (real and financial assets), and credit channel. Without such differentiation, it is impossible to explain how individuals and companies who have never used credit have money in their bank accounts, how people without any income or creditworthiness also have money in their portfolios or how it is possible to withdraw $140 billion uninsured deposits from an illiquid bank in two days.

The paper after this Introduction (1) is organised as follows: Evolution of money (2), Immediate money (3), Mediated money (4), Concealed differences (5), Clearing and settlement power (6), RTGS: the way backward (7), RTNS solution: accelerated sequential netting (8), Balance sheet reconstruction (9) and Conclusion (10).      

2 Evolution of money

‘One of the chief things which monetary theory ought to explain is the evolution of money. If we can reduce the main lines of that evolution to a logical pattern, we shall not only have thrown light upon history, we shall have deepened our understanding of money, even modern money, itself’ (Hicks, p. 2).

It is indisputable that modern money cannot be understood without explaining the evolution of money. Four main lines of that evolution can be observed: metal money – book money – paper money – electronic/digital money. Separating these main lines of evolution from the apparent, i.e., material forms of money shows the unique logical pattern: immediate money and mediated money.

Both primary forms embody the same functions of money and illuminate their different creation and circulation, without being obscured by material forms (metal, paper, plastic, PC, mobile device). The term paper money is the best example of confusion (book money, cheques, banknotes, etc.).

The first archaeological evidence of the existence of money over 5,000 years ago comes from temples in Mesopotamia. They were the first to shape immediate metal money as a standard of weight and mediated (accounting) money in the third millennium BC (Keynes, 1930a, p.12–13). The original immediate money consisted of weighed metal ingots and bars, most commonly silver. ‘Silver by weight was a widely used form of currency in the Near and Middle East during the second and first millennia B.C.’ (Grierson, p. 3).

Regardless of the physical form, transactions in temple were documented. These documented transactions in temples were not bureaucratic tools, but brilliant innovations in settlement of obligations, i.e., in payments (Peacock). The misconception that payments occurred only when ‘silver changing hands’ and the identification of monetization with the amount of metal in circulation (see: Dale, p. 167, 170–1), results from the metallist view of money. Temple’s documented transactions preceded the emergence of modern book/deposit money (Vuković, 2022).

The development of seafaring enabled the spread of wholesale barter, with the use of weighed metal money for settlement. This type of international trade reached its peak with the super-commercial Phoenicia and its most powerful colony – Carthage. Weighed metal money remained dominant even after the appearance of metal coins, because coins, due to their intrinsic problems, were not suitable for wholesale trade and large-value payments. The bars of gold in modern central banks’ vaults reveal the importance and uninterrupted continuity of weighed metal money. Coined metal money is rapidly disappearing, as our wallets testify.

The minting of coins and their usage were first started by the Lydians in the seventh century BC. Herodotus testifies about this: ‘They were the first men whom we know who coined and used gold and silver currency; and they were the first to sell by retail.’ Thus, Herodotus, in one sentence, connected currency, i.e., coins, with retail. The Lydians’ role in the history of coins ended unfortunately with the military defeat of Croesus, the last king of Lydia.

Through the cities of Ionia and nearby Aegina, silver-rich Athens (mines at Laurium) embraced coins at the beginning of its era of brilliant prosperity and expansion. Thucydides testified that the primary purpose of coins at that time was to pay mercenaries (soldiers and sailors).

With the appearance of coins in the Athenian Agora, the emergence of moneychangers, i.e., trapezites, inevitably followed. By testing coins, they facilitated payment (Andreau, p. 30). In doing so, they established themselves as indispensable intermediaries in monetary transactions and ‘the first bankers to receive deposits’ (ibidem). In their everyday deposit business, the trapezites rediscovered the Babylonian secret – not shift metal money from one purse to another, but from account-to-account. Thus, mediated money was reinvented as book money.

The military (and political) power of coins motivated subsequent issuers. ‘The third people to issue coins were the Persians’ (Seaford, p. 128). They were followed by the super promoters of coins – the Macedonians, led by kings Philip II and Alexander the Great. The Macedonian conquests were launched by taking control of gold and silver mines and the expansion of coinage, supported by the systematic looting of enemy temple treasuries. Numerous mints continuously produced new coins (staters) and bullions, which mercenaries dispersed across the then-known world. This global promotion of immediate money was continued by Alexander’s successors (Diadochi), as they paid their mercenaries (Kroll, p. 14).

The first argentarii appeared in the Roman Forum at the end of the fourth century BC, but coins in Rome never gained the significance they had in the Greek polis. Two main problems were the inconsistent intrinsic value of coins and their constant scarcity, due to insufficient coinage and excessive hoardings. A long list of intrinsic problems of coins in Antique includes difficult transport, risky storage, different metals and debasement.

Unsolvable issue were large-value payments. In Rome, with chronically scarce metal money, such transactions were unfeasible only in cash. Therefore, payments were necessary ‘by a transfer within a particular bank’ (Finley, p. 141). With the acceleration of debasement of silver denarius in the 3rd and 4th centuries AD, mediated book money was inevitably disappearing.

As early as the fourth century AD, the main role in coinage with gold solidus was taken over by Byzantium, which also had major expenses for paying mercenaries. In western Europe, silver coins circulated until the 13th century, followed by large silver coins (groats or grossi), which increased another intrinsic problem – outflow of silver to Asia (Spufford, p. 259). At that time, gold coins also appeared (florin in Florence and ducat in Venice), announcing bimetallism problems.

The book money was once again reinvented as deposit money in Genoa (at the end of the 12th century) and Venice (in the 13th century). The ancient money intermediaries – trapezites and argentarii – were replaced by new ones: bankers. Similar to their antiquity predecessors, they began as moneychangers with their own secure storages (Kohn), and then expanded their business to accepting deposits and making payments on behalf of clients (Spufford). From the 14th century onward, one can already observe a pattern: mediated book money was increasingly used for payments in wholesale trade, and coins in urban retail.

Venice, the greatest financial and trading power of the 14th and 15th centuries, was the leader in the development of deposit banking and payment in books with mediated money. A few private banks (maximum 10) were grouped on the Piazza di Rialto, within easy reach of merchants arranging commercial transactions. The very names of these banks – banchi di scritta – indicated their primary mode of payment and the possibilities for client lending and investment.

Thanks to their proximity, the banks on the Rialto Square ‘[…] could effect such interbank transfers only through reciprocal accounts that they held with each other, and which they settled periodically as needed’ (Kohn, p. 3). Thus was discovered interbank clearing and the benefits of payments ‘in network’.

Despite all the advantages, these private transfer banks were fragile institutions. Times were uncertain and turbulent, and depositors’ runs on banks were frequent (Ugolini, p. 39-40). In the end, by the mid-1580s, no one was interested in the banking business in Venice anymore. ‘Only in 1587 would the public Banco della Piazza di Rialto be substituted for private banks, […] Finally Venice had the bank envisioned more than two centuries earlier, which held, at least at the outset, 100 percent reserves and provided the service of clearing debts via giro among its depositors’ (Mueller, p. 118). The biggest changes in the design of the new bank were public ownership and lending prohibition. Thus, mediated money was revitalized as a state project.

For large payments in metal money during the 16th and 17th centuries, silver bullions were used, mostly for waging wars (Kohn). If there were no more coins and bullions, nor new creditors, armies would disband, and peace would follow (Kindleberger). Therefore, the inflow of gold and silver from the Americas motivated more wars than it facilitated large and small payments. Understandably, a large portion of wholesale and retail trade satisfied military needs. The financing of wars was also supported by mediated money as early as the 13th century (Mueller), and to this day has become almost the only means of paying military expenses.

The peak of coinage chaos was recorded in the Netherlands at the beginning of the 17th century: countless types of coins, issued by numerous autonomous mints, of poor quality and dubious weight. With the establishment of the public Bank of Amsterdam (BoA) in city ownership in 1609, this problem was significantly mitigated, although the primary goal was to create a system of large international payments account-to-account. It was a triumph of mediated money: the BoA guilder as account money was the dominant currency in European and global wholesale for almost two centuries, until the 1790s (Quinn and Roberds, p. 63).

‘The initial design of the Bank borrowed heavily from Venice’s Banco di Rialto. The Bank’s charter granted it extensive legal privileges’ (Roberds and Velde). Adam Smith classified the BoA guilder into bank money, i.e. deposit money, emphasizing its ‘intrinsic superiority to currency, security, simplicity and risk-free transfer’. Thanks to these features, mediated money remained the backbone of the development of private and public deposit banks and banking in general.

Innovations and stricter coinage rules in the 17th and 18th centuries also mitigated the intrinsic problems of coins, but they could not solve the enigma of bimetallism. The problem of the fluctuating market price ratio between gold and silver was alleviated only in the 19th century, when the era of the gold standard officially began. At that time, the myth of the intrinsic value of metal money was solidified.

Paper immediate money appeared in London during the second half of the 17th century, initially in the form of goldsmith notes or deposit receipts. The development of this innovative form of money was accelerated by the founding of the Bank of England in 1694. ‘It is from these goldsmiths’ notes that the Bank of England note ultimately derives. […] The notes issued after 1725 were in mainly printed form and they began to appear in fixed denominations; the first was for £20’ (Bank of England, 1969, p. 211-2). Their high denomination – from £30 to £1,000 (ibid, p. 222), confirmed that all BoE notes and other bankers’ notes in circulation were primarily intended to serve as ‘instruments of commerce’ (Smith). ‘For context, the majority of Londoners had never seen a banknote before the 1790s’ (Vuković, 2020c, p. 5).

However, banknotes were in practice impractical for larger transactions, which as a rule had to be documented. This new immediate money also could not solve the old snags of hand-to-hand payments, unsuitable for wholesale trade: ‘the trouble of counting, or the risk of transporting it from one place to another’ (Smith, p. 448). Moreover, the banknote issue business was risky and low-profit. For that reason, London bankers gradually abandoned the issuance of their own notes, replacing them with BoE notes in interbank transactions as early as the mid-18th century (Arnon, p. 23).

The abandonment of private banknotes did not prevent private bankers from developing internal clearing and settlement of payments among its depositors. At the same time, London bankers initiated the clearing of interbank payments, which would soon revolutionize mediated money. The epicenter of that revolution was the Bankers’ Clearing House, established in London during the 1770s. Interbank clearing, with the use of metal money for final settlement, was well known to Rialto bankers, but Venice in the 15th and 16th centuries was a less secure environment than London in the second half of the 18th century.

At that time, BoE notes, as an advanced form of immediate money, significantly contributed to the promotion of mediated money in interbank payments, serving as the means of final settlement. In this way, the millennia-long intertwining of two primary forms of money continued. Pure mediated money emerged by excluding BoE notes from final interbank settlements. ‘Starting in 1854, these notes were replaced by cheques drawn on bank deposits with the Bank of England’ (Vuković, 2020c, p. 6). Already in the 1870s, ‘[…] debts to the average amount of nearly twenty million sterling per day are liquidated without the use of a single coin or bank-note’ (Jevons, p. 263).

The evolution of money accelerated dramatically in the 19th century, when mediated central bank money with unlimited deposit transactions was shapped. This unlimitedness was ensured by the Bank Charter Act of 1844 by separating banknotes issue and deposit business. The benefits of such liberalization were enjoyed by commercial banks as well: ‘We must never forget that a quarter of a century has passed since 1844, – a period singularly remarkable for its material progress, and almost marvellous in its banking development’ (Bagehot, p. 3). The phrase ‘almost marvellous banking development’ describes the eruption of mediated money in commercial banks.

Banknotes of small denominations were launched throughout Europe only in the 19th century (with the exception of Scotland during the 18th century), convincing people that even ‘worthless bits of paper’ (Jevons, p. 194) could represent hard money. This was the continuation of the interrupted paper money alchemy started by John Law between 1716 and 1720. Nevertheless, despite the alchemy of ‘bits of paper’, banknotes reached their zenith before the end of the 1840s. ‘The total bank-note circulation of England and Wales in June 1914 was almost precisely what it had been in 1866 and in 1844, that is, between £28,000,000 and £29,000,000. Peel’s Act became a classroom theme, while with every decade the bank note became less and less important’ (Clapham, II, p. 270). The declining use of immediate money was noticeable throughout the 20th century. This negative trend dramatically accelerated in the 2020s, signaling a cashless society, in which money would circulate exclusively in-network via intermediaries, i.e. payment institutions.

The global expansion of mediated money was hindered by the international settlement in gold until the 1930s. The Bank for International Settlements (BIS), to settle the question of war reparation payments, was established in January 1930. From July 1932, ‘[…] the BIS focused its activities on the technical cooperation between central banks (including reserve management, foreign exchange transactions, international postal payments, gold deposit and swap facilities) and on providing a forum for regular meetings of central bank Governors and officials’ (www.bis.org/about/history). This general formulation about the technical cooperation between central banks describes the creation of a global clearing and settlement system within the most powerful international monetary institution—BIS—although much less known and more discreet than the IMF, World Bank, and other supranational financial organisations.

The gradual rise of cheque usage for disposing mediated money in books of banks marks the first half of the 20th century. From the 1970s, cheques were supplanted by payment cards, which, with the development of electronic communications, came to be called electronic money. In this way, the spread of mediated money accelerated. Of course, these were new instruments and devices for the circulation of the old book money, which had meanwhile been renamed into deposit money. The same applies to digital money, embodied in smartphones, as well as to the future quantum money. During the 2020s, these multifunctional mobile devices took the lead in transferring mediated money, increasingly suppressing payment cards and PCs in developed economies. Paradoxically, the same mobile devices have already become dominant in most developing countries, and even predominant – in China 99% of people.

Accelerated technological progress has multiplied the clearing and settlement power of mediated money. Consequently, total networking of retail and wholesale transactions is taking place in national and global payment systems. This announcements of totalitarian monetary future raise difficult questions that require comprehensive research. 

3 Immediate money

The original immediate metal money enabled the establishment of the standard of weight of a representative metal, most commonly silver. Therefore, the emergence and use of ancient money are inseparably linked to weighed metal money. The money weight standard was usually determined by the sovereign or the state, which remained common practice until the abolition of the gold standard in 1971. ‘The first State reform of the standard of weight, of which we have definite record, was the Babylonian reform towards the end of the third millennium B.C.’ (Keynes, 1930a, p. 12).

Non-metal commodities were used as a primitive barter standard among archaic communities in pre-coin time. After the invention of coins, it is evident that most primitive communities encountered money forcibly, through their conquerors and colonizers.‘The primitive man’s distaste for exchange’ (Simmel, p. 94) has also been proven.

Weighed metal was in every respect the embryo of money. “To weigh” was the term used for payment in metal, […] The shekel and the talent which came to be known subsequently as coins or monetary units, first originated as units of weight (Einzig, p. 212). These units of weight served as monetary units at least two thousand years before the first coins.

Ancient Mesopotamia’s millennial experience had a strong influence on Phoenician trading cities and Hellenistic poleis. In pre-coin age, for Solon, silver was weighed (Kroll, p. 16).

Weighed metal money was regularly stamped. ‘The monetary character of silver was reinforced by the stamping of the ingots to guarantee their weight and/or fineness’ (Einzig, p. 212). Modern representatives of weighed metal money, such as gold bars in modern banks’ vaults, are also stamped with markings of weight and fineness. Additionally, it was often practiced ‘[…] that the ingots were enclosed in sealed sacks’ (ibid, p. 213), which clearly facilitated large-value transactions.

The widespread use of weighed metal money enabled the development of wholesale barter, primarily as the means of settlement. This is likely the reason why ‘[…] coins should have been invented by Lydia, a predominantly pastoral community, instead of Phoenicia, one of the most highly commercialized communities of all times’ (ibid, p. 224).

Coined metal money never held such importance for wholesale trade. ‘It is a practice which has never caught on in some important commercial areas. Egypt never coined money before the Ptolemy, and China (broadly speaking) has never coined silver, which is its standard of value, until the most recent times. The Carthaginians were reluctant coiners, and perhaps never coined except for foreign activities’ (Keynes, 1930a, p. 12).

Nonetheless, coins played an epochal role in introducing people across the then-known ancient world to coined money and its military, commercial, and sacral powers. The usual domains of coin usage included mercenary salaries, urban retail, rural reserves, hoarding, and taxes. Coins were the most visible representatives of immediate metal money, as they circulated hand-to-hand much more frequently than ingots or bars. This explains why, until the end of the 19th century, coins were synonymous with money. ‘Standard legal tender coin of full weight is undoubtedly money, and as convertible legal tender bank-notes are exactly equivalent to the coined money […]’ (Jevons, p. 248).

The coins are usually minted in royal mints, which today are generally under the authority of ministries of finance, i.e. treasury, and less often within central banks. Due to their metal content and non-mediated circulation, coins were often used in neighboring countries until the end of the 19th century. Similarly, coins from long-defunct original mints (e.g. Roman coins in England until the 15th century) were accepted for payment. From this, one may conclude that coins were the most mobile form of immediate money.

Coined metal money remained on the pedestal of true or proper money for more than two and a half millennia. It served as the benchmark for representative money until the 1930s. Despite that, the reputation of coins has not been compromised. In the 21st century, numerous private cryptocoins have been launched, starting with bitcoin. Many have proposed the abolition of banknotes, but no one has suggested abolishing metal coins! Could they be the last bastion of immediate money?

Banknotes are the youngest and most developed shape of non-mediated money, but they have never been the dominant means of payment (Vuković, 2020b). Nevertheless, from their inception, they have attracted great public attention (bits of paper), magnetically drawn researchers (printing money), been loudly contested (central bank monopoly) and accused (curse of cash). Exclusive issuers of banknotes over the past hundred years are central banks, yet even that has not spared them from marginalization in modern payments.

Like coins, banknotes circulate hand-to-hand and are always in the possession of their holder. Thanks to off-line transference all payments are final. Hence, it is understandable that together with coins they are classified as cash money. Circulating without intermediaries means that this immediate money is out of networks, i.e. it can only be used by its owner. From the perspective of intermediators, this money is insufficiently mobilized, while for depositors it is the only way to step out of the network.

Due to technological neglect and systematic suppression, cash has become impractical even for retail payments. This has eliminated the threat of old-fashioned run on banks (the last one being the run on Northern Rock in September 2007). The possibility to withdraw own money from the banking system today is reduced to a minimum — almost all money is locked in network. Thus, immediate money is condemned to extinction, left to monetary reservations which unbanked people are inhabited (Vuković, 2021).

According to issuers, coins and banknotes are classified as central bank money. Consequently, with their disappearance, immediate central bank money will also vanish, and retail payments will be conducted exclusively in-network. In the end, weight metal money remains the only shape of immediate money with a somewhat certain future, but only in bank vaults.

A potential opportunity for the reanimation of cash could be provided by announced off-line CBDC, which would circulate device-to-device. This technical solution has been explored by central banks since the second half of the 2010s, when they realized that the general public had been almost completely deprived of access to central bank money.

4 Mediated money 

The rudimentary pattern of ancient money-on-account was represented by Babylonian temples’ clay records on claims and transactions. This temple money was the early book money. The extraordinary role of temples in shaping money and its functions is largely neglected or misinterpreted, despite countless archaeological findings. The contested formulation of a “marketless Babylon” is profoundly unjust to a civilization that, besides the weighed metal money and accounting money, conceived credit, interest, and interest rates, and gave the world the Code of Hammurabi (c. 1755–1750 BC). Equally unacceptable is the claim about ‘significant periods of unfettered market activity’ (see: Dale, p. 173). In any case, ancient Mesopotamia is the birthplace of mediated money.

The trapezites in the Athenian Agora, who appeared during the fifth BC (Andreau, p. 3), were the first to develop mediated money based on deposits of coins. In this way, coins enabled the rediscovering of the book money and the Babylonian secret – shifting ingots and coins from account-to-account.

‘The argentarii – money-changers, assayers and deposit bankers, installed in the Roman Forum between 318 and 310 BC’ (Andreau, p. 30). The advantage of the trapezites and argentarii was that they could provide not only coin exchange and testing, but also safekeeping and payments ‘in bank’ with documentary evidence.

Here the ancient roots become visible of ‘one of the most misleading words in the English language – the word “deposit”, when used to refer to a claim against a bank’ (Milton Friedman). The same problem exists in all other languages that have adopted this word from Latin. Circulating deposits, current deposits or transaction deposits are oxymoron. A deposit by legal definition cannot circulate and cannot flow – its essential law feature is stationarity.

Two types are known: regular deposits and irregular deposits. Etymologically, regular deposits reflect the original meaning in the sense of ‘nomen est omen’. The original meaning is also confirmed by their everyday name – sealed deposits. ‘These deposits, in the form sometimes of coins, sometimes of objects or documents, had to be restored to their owners untouched by whoever accepted them as deposits. […] But such sealed deposits were clearly not the major feature of a bank, for that was constituted by the non-sealed deposits, which the banker had the right to invest, provided he would subsequently restore an equivalent sum to its owner. However, a study of bank deposits in Rome is complicated by the legal problems to which they gave rise’ (Andreau, p. 40).

One of those legal problems was the failure to distinguish cash and non-cash transactions. Documentary transactions ‘in bank’ referred to both cash and non-cash payments (Andreau, p. 44). This still confuses archaeologists and historians when interpreting textual evidence. All payment transactions via the argentarii, cash and non-cash (numerare, in saculo dare, persolvere), were documented in the register chronologically. ‘All these operations put together constituted the deposit account of his client, his ratio’ (ibidem).

Historians provide detailed descriptions of payments based on textual evidence, but fail to see the ‘multiplier effect’ because ‘no clearing centres’ (see: Harris, p. 201) and ‘negotiable instruments’ (Finley, p. 141). Many economists share the same conclusion. The reason is that payments without coins are linked exclusively to ‘modern institutions’ such as ‘clearing banks’ (Harris). However, for non-cash payments, licensed, credible, and liquid payment intermediaries such as trapezites and argentarii were sufficient. Many cases are known where ‘[a] payment had been made but no coins had changed hands’ (Harris, p. 198).

Let us apply the mengerian methodology (Menger, p. 257–285). The economizing individual deposited coins with his argentarius, with whom he had an account. Many other Romans did the same with the same argentarius. It was certainly common for them to have mutual payments. It is hard to imagine a payment transaction executed purse-to-purse in the presence of three parties – the banker and his two clients, payer and payee. Even harder to imagine is a payment purse-to-hand and hand-to-purse. It would be absurd to shift coins from one purse to another at the same banker. These obvious peculiar difficulties due to double handling were resolved by bankers through the use of mediated book money. Mutual payments at the same banker were documentary transactions without coins, i.e. account-to-account. All coins remained in the purse of bankers! Without double handling.

Trapezites, argentarii, and other payment intermediaries represented the basic level of clearing (internal) on which mediated money is based. They were intermediaries for mutual payments of their depositors in own network. It is not difficult to conclude that the banker and his clients, thanks to overdraft in network, were able to execute many more mutual payments than there were coins in purse of banker. They could do this without connecting with other bankers, although that was inevitably the case. Numerous documentary transaction payments of enormous amounts, which would require tons and tens of tons of silver coins and ingots, cannot be explained otherwise.

Evidence of the lack of metal money in circulation indicates that Rome was under-monetized. It seems that mediated money hides one of the secrets of ancient monetization. ‘[T]he Roman monetary system was far indeed from relying entirely on coinage. Romans, especially those whose credit was good, frequently made payments without coinage’ (Harris, p. 207).

The Renaissance of book money, which began in Genoa at the end of the 12th century, gained its greatest momentum in Venice from the 13th to the 16th century. ‘[T]he more holders of current accounts in Venice used those accounts as means of payment, the more credit was created and the more specie was freed for export to the Levant, which suffered from a chronic bullion famine and which attracted Venetian coin as an exchange commodity’ (Mueller, p. 25). It is clear that this hub of wholesale trade with East would not have achieved global maritime power and commercial domination without the use of mediated money for transfers account-to-account.

A succinct description of the job of deposit bankers was given by the famous mathematician Fra Luca Pacioli in 1494, as a direct witness: ‘It is common practice to deal directly with a transfer bank, where you can deposit your money for greater security or for the purpose of making your daily payments […]’ (Mueller, p. 5).

From the end of the 16th century to the 1870s, the use of mediated money in continental Europe was most strongly promoted and developed by old deposit banks in public own. ‘The Bank of Amsterdam represented the peak of development of public deposits banks: it was modelled after Venice’s Banco di Rialto (1587) and followed by the Hamburger Bank or Bank of Hamburg (1619). […] Most of these public banks were not doomed to fail by the superiority of note issuing banking: most were swept away only by the Napoleonic wars.’ (Vuković, 2020b, p. 4).

However, not even Napoleon managed to destroy them all. A shining example is the Bank of Hamburg. ‘In short, in the 1870s, a public bank more than 250 years old (established in 1619) operated the best developed system of transfer transactions in the new German Empire’ (ibidem). The statistics are irrefutable evidence: ‘Transactions on transfer account, which had amounted in 1875 at the Prussian Bank to only 834,000,000 marks and at the Hamburg Bank to 2,658,000,000 marks’.

London bankers were well acquainted with money transfer technology and used public deposit banks on the Continent. ‘All banks will to a certain extent economize currency, and those of Amsterdam and Hamburg have for some centuries carried on a system of transfers, the true prototype of our system’ (Jevons, p. 338). The Bankers’ Clearing House in London from 1770s was a continuation of the evolution of mediated money, with a Copernican leap – from 1841 it performed multilateral clearing, previously unknown in the world of banking.

The revolutionary nature of the London BCH is highlighted by the years of establishment of other clearing houses: New York CH established 1853, Paris CH established 1872, and Berlin CH established 1883 (Vuković, p. 10–11). At the same time, this testifies to how difficult it was for the practice of using mediated money in interbank payments, since many bankers in the 19th century failed to understand what had already been clear to Luca Pacioli and Venetian bankers in the 15th century!

Bank of England entered the London Clearing House in May of 1864, technically: ‘enter the clearing’ (Clapham, II, p. 251). Thus, the first national clearing system was created, which no longer had spatial limits in domestic payments – main clearing bankers were in-network.

International clearing was hindered by gold fetters until 1932, when the newly established BIS began ‘[…] activities on the technical cooperation between central banks (including reserve management, foreign exchange transactions, international postal payments, gold deposit and swap facilities)’. Naturally, gold deposits were not removed from central bank vault, but only a portion of them was transferred to the BIS as collateral to secure international settings. With the abolition of the gold standard in 1971, the formal link between gold and international clearing and settlement systems was severed.

At the end of this historical overview, it may be concluded that the most important places for development of mediated money were, chronologically: Babylonia (temples and palaces) – Phoenicia (trading cities and colonies) – Athenian Agora (trapezites) – Venetian Rialto (bankers) – Amsterdam (Wisselbank) – London (Bank of England and Bankers’ Clearing House) – New York (Clearing House and Federal Reserve Bank) – Basel (Bank for International Settlement).

Mediated money remains to this day the greatest monetary enigma, even though its evolution has lasted more than four millennia. The development of mediated money presented above suggests that the essential reason for its enigmatic nature lies in transactions evidence in-book, as well as in the diversity of its payment instruments and devices (from clay tablets to smartphones). The failure to distinguish between issuers and intermediaries of book-money further contributes to its monetary deceiving. Still, the biggest illusionists in this ancient story about mediated money are clearing and convertibility. That is why further research should begin with them.

Clearing is the basic method of settling mediated money, i.e. their transactions from account-to-account in the books of payment intermediaries, which represented the internal network. The payer sent money and the payee received money via the intermediary. The transaction was recorded in intermediator’s book, creating the illusion that it was executed without money, although it is clear that without payer’s money there would be no payment. This illusion is so strong that the Governor of BoE in 1992 emphasized explicit transactions in gross payment systems opposite net payment systems (Leigh-Pemberton, p. 454). Are there explicit payments vs implicit payments or explicit money vs implicit money?

The original internal clearing is probably the best example of mediated money circulation, already described through the relationship between economizing individuals and Roman argentarii (p. 11). The brilliantly simple Babylonian formula – not shift money from hand-to-hand, but from account-to-account – functioned effectively with licensed, credible, and liquid intermediaries in a stable environment of trusted money issuers. The disruption of these conditions reduces or eliminates any form of clearing (internal, bilateral, multilateral, etc.).

The logic of clearing is based on the symmetricity of payments, expressed by the equation sent payments = received payments in network. This symmetricity of payments characterizes immediate money as well, but without intermediators. Consideting the stock of mediated money is always in the intermediator possession, he can freely use deposited money. Freely using of money in-network allows incomparably expansive payments than cash, but also causes illiquidity and failure of intermediaries. Until the founding of Federal Deposit Insurance Corporation (FDIC) 1934, bankruptcy of deposit institution meant the loss of money of depositors.

Obligation of convertibility explains the rest of the enigma. Accepting mediated money has always been related to the convertibility of money-on-account into money of more stable value. Until the 20th century, metal money, weighed and coined, was considered more stable money. Non-metal money, including banknotes, had not been treated as proper-money. 

The gold standard was formally inaugurated in 1821 precisely due to the convertibility of banknotes into gold, as a way to regulate the freely print of numerous issuers. The problem of numerous issuers was brought under control only in 1844, when the Bank Charter Act was enacted. Central banks, as the sole issuers of banknotes since the 1920s, strengthened public trust in these bits of paper and made them proper-money. Consequently, the convertibility of mediated money became tied to central bank notes.

During the Great Depression of 1929–1933, this paper money became the most sought-after store of value. It was the time of the infamous run on banks, which collapsed more than ten thousand banks, mostly in the United States (over 9,000). The impracticality and legal limitations of using banknotes marginalized them as a medium of convertibility after the Global Financial Crisis, in the late 2010s. Since then, the term convertibility has referred only to the possibility that ‘one currency can be freely exchanged for another’.

It was shown that the acceptability of mediated or book money, as well as its very existence, depended on the deposit convertibility into immediate metal money up until the end of the 19th century, and during the 20th century into central bank notes. This gave deposit holders a sense of security about their transaction money in the bank. “People are so used to having payment services provided against checkable fixed nominal value liabilities, with 100% convertibility of demand deposits, […]” (Goodhart, 1988, p. 89).

Today, the convertibility of mediated money rests solely on the possibility of transferring deposits to another bank. Simply put, you can no longer get out from the network of commercial banks! Because of minimizing convertibility of demand deposits into immediate money, the only mode of exit from the network has been abolished, in case trust is lost in the chosen intermediary or the entire banking sector. ATMs remain the last resort of convertibility of our bank deposits.

The basic issuer of modern mediated money is a central bank, and it circulates between commercial banks. That central bank money is issued and circulates via commercial banks’ accounts with the central bank, the so-called reserves. Direct access to this non-cash money is available only to a smaller portion of payment intermediaries – authorized commercial banks and selected financial institutions. Other intermediaries access it through correspondent banks. All of them, along with the central bank, are interconnected through retail and wholesale payment systems in a comprehensive national clearing network.

The only money that is created and circulates in national network, as well as through all other levels of clearing (internal, bilateral, multilateral, etc), is the central bank money, denominated in national currency. “Commercial bank money” is an illusion created by the very clearing of transactions in books of banks. The bank money that Adam Smith emphasized as early as 1776 was the money of public deposit banks; during the 19th century, it evolved into central bank money. Today, this is the only shape of mediated money or bank money.

Mediated money enables the multiplication of payments within the network, provided that banks can freely use deposited money their clients. The expansion of payments is effectuated through overdraft and other loans, which are promise of money, i.e., promise of payment in the form of bank liabilities – deposits. It can be noticed that this is a process of an apparent deposit and credit multiplication, which actually represents a reflection of clearing. Without clearing and monopoly of payments, the credit potential of any bank would be equal to that of any non-deposit financial institution with the same liabilities and equity. The proposal on balance sheet reconstruction of central bank money in commercial banks (Section 9) explains this monetary deceiving.

With the suppression of cash in retail transactions, a closed payments network is being created, eliminating the limits of mediated money creation. ‘Today’s central banks have the capability of creating or destroying unlimited supplies of money and credit’ (Alan Greenspan). The terrifying monetary expansion during the 2010s and early 2020s unequivocally confirms this.

The national clearing system is a closed payments network. This comprehensive network is managed by the central bank, which is the sole issuer of modern mediated money and the lender of last resort. Mediated money circulates exclusively account-to-account in books of payment intermediaries – banks and other PSPs. The central bank is a supermediator – comprehensive and universal. The unlimited clearing and settlement power of central bank is the nuclear reactor of mediated money emission. Only thanks to such power of the central bank were wholesale depositors who were able to withdraw $140 billion uninsured deposits from the illiquid Silicon Valley Bank in just two days in March 2023.

5 Concealed differences    

The first and most important concealed difference is the very existence of two primary forms of money: immediate and mediated money. Such concealment is a consequence of neglecting the relation to their fundamental mode of circulation – off-line (hand-to-hand) versus in-network (account-to-account). That is why the belief that mediated money in circulation is a value-equivalent of hard money persists to this day as a form of metallist blindness. Current data (Table 1) indicate that all global monetary gold reserves would not be sufficient for wholesale transactions in a single day!

Immediate and mediated money have the same issuer – the central bank, but differ in their access to it – direct access (cash) and indirect access (bank accounts). Access to money is determined by their key characteristics – safety for immediate money and efficiency for mediated money.

Monetary statistics of most developed economies show that the household sector holds more than half the value of all transaction accounts (so-called demand deposits) at banks. The rest belongs to the corporate sector and financial sector. This is not surprising, as households are traditionally net creditors. Of course, for a smaller portion of retail transaction, people still use cash. With this dominant value of transaction accounts, households execute less than 0.5% of the value of total transactions! Two hundred times less than corporations and financial institutions!

Table 1 – Proportions of retail and wholesale transactions

        Source: BIS Annual Economic Report 2020, Basel, June 2020, p. 73.

The drivers of the eruption of payment transactions and mediated money are clearly wholesale (commercial and financial) transactions, which have completely marginalized retail payments (Table 1). As already emphasized above, freely using of mediated money in-network enables incomparably more expansive payments than cash, but also causes illiquidity and failure of intermediaries. From time to time, the entire system faces collapse, which is why systemic liquidity risk is one of the more frequent topics in monetary literature.

At first glance, it is evident that there is no problem of household sector liquidity, but only wholesale sectors liquidity – commercial and especially financial. There is no longer a danger of withdrawing household deposits from banking network. Thanks to the crowding-out of cash, the depositor can no longer exit the network, but can only switch banks. Thus, mediated money is locked in bank system.

The rapid growth of mediated money, followed by the eruption of transactions in-bank, encouraged theoretical equalization of issuers (central banks) and intermediaries (commercial banks). The main product of this confusion is the promotion of commercial bank money as equal or even superior to central bank money. In public attitudes about money, commercial bank money does not exist. ‘Most of the public intuitively knows that a money issuer cannot go bankrupt. But commercial banks go bankrupt. Despite the access to central bank’s liquidity, prudential regulation, supervision and effective resolution regimes (Financial Stability Board), there will still be bank failures’ (Vuković, 2023, p. 9). The same applies to other fictions from mediated money – there is no such thing as clearinghouse money or netting money, not even as ‘a very close substitute’ (BIS, 1989).

Lastly, concealed differences between money issuers and money intermediaries are revealed by the interbank settlement process itself. This process has always relied on an acceptable settlement asset; without such an asset, there could be no settlements, and therefore no mediated money. Commercial banks have never accepted claims from other banks for settlement. The only settlement asset that all banks have accepted was and remains the central bank money – immediate (until the mid-19th century) and mediated. That is why central banks are the ultimate providers of interbank settlement assets.

The comprehensive dynamic approach explains all the diversity and peculiarity of mediated money. Dynamic analysis focuses on the stock-flow relationship. The stock of money is the result of flow; money flow contains a series of payment transactions; transactions occur in network between payer and payee; transaction outflow of payer = transactions inflow of payee (symmetric); each transaction changes the stock of money for both the payer and payee; the transaction does not change the stock of money in the overall system; transactions are successive and interdependent – previous transactions determine the next ones; full information on money stock and flow in the network is held only by the system mediator; stock of money enables the start of flow; the difference between Stock(t) and Stock(t-1) does not show the total value of flow.

Why is the flow of money important? Most monetary analyses are static. The first sentence of one of the most well-known monetary studies is enough: ‘This book is about the stock of money in the United States’ (Friedman and Schwartz, p. 3). However, money stock is not the only measure of money supply. A more complete explanation requires a systemic approach, i.e. money flow analysis. Cash-flow analysis is partial because it focuses on individual units within the system.

The fundamental stock-flow relationship equation is:

Stock(t) = Stock(t-1) + ∑Inflow – ∑Outflow

The most important flow variable is Net flow:

Net flow = ∑Inflow – ∑Outflow

Consequently, the key stock variable follows:

Stock(t) = Stock(t-1) ± Net flow (Net position)

The above equations and variables are relevant for every intermediary and every payment system.

In a closed interbank payment network:

∑Inflow = ∑Outflow => Stock(t) = Stock(t-1)

Two main conclusions arise: first, central bank money determines and constitutes the stock of money (base money); second, transaction accounts with commercial banks (so-called demand deposits) are promise money (most of M1). The relationship between these two monetary aggregates is usually shown as the money multiplier:

M1 = m·BM

However, a more accurate term would be promises multiplier.

Mediated money, by itself, is not the cause of the expansion of promise money. The cause is its free use by intermediaries under conditions of insufficiently clear legislation. The advantage of this freedom of use was the efficiency in payments and power of netting.

Understandably, the multiplication of payments with immediate money is impossible due to its fundamental characteristics – circulation hand-to-hand and because the cash is always in the holder’s possession. Therefore, it is unfounded to equate the safety of cash and mediated money. Even when its ‘immediate’ settlement is emphasized, mediated money remains promise money. Immediate money is cash in hand, whose transactions are instantaneous, final, irrevocable, and unconditional.

6 Clearing and settlement power

The power of mediated money that circulates in books of intermediaries has always been understandable and appealing to bankers, just as much as it has been misunderstood and off-putting to the general public. The true extent of that power—without banking bias and public aversion—can be determined through its quantification.

Keynes was explicit and precise in quantification. ‘[I] propose to employ two terms, namely Velocity (V) and Efficiency (E); of which the latter represents the ratio of the Bank Clearings to the Total Deposits. This leaves us free to use the expression ‘velocity of circulation’ to denote unambiguously the velocity or rate of turnover of what is truly serving the purposes of cash, namely the Cash Deposits. It follows that E = Vw, where w is the proportion of the Cash-deposits to the total deposits’ (Keynes, 1930b, p. 22). By emphasizing Bank Clearing, Keynes stepped out from a static standpoint into a dynamic analysis. A purified and modernized version of his ratio would be: Clearings Efficiency = Bank Clearing to Demand Deposits.

The official formulation of this ratio, published in December 2023, is nearly identical, although it carries a distinctly pejorative name. ‘Liquidity recycling is the ratio between value of payments sent and liquidity usage’ (BoE & PRA, p. 15). For illustration, if a gross outflow is ten times bigger than a liquidity usage, then there is a liquidity recycling factor of 10. In this way, the greatest economic power has been reduced to ‘the level of liquidity recycling’ (ibidem).

The efficiency ratio at the system level is essentially the same as the previous two: ‘[…] intraday liquidity efficiency Qs to be the ratio of aggregate payment values over aggregate intraday liquidity used at the system level. Qs ≡ Ps / Ls. This ratio captures the value of payments that are made for each unit of intraday liquidity used. If system participants can meet their daily payment obligations with minimal liquidity usage, the ratio takes on higher values and the system is more liquidity efficient’ (Kabadjova et al, p. 11). Here, the greatest economic power is subordinated to the ‘minimization of liquidity usage’.

The results of empirical research illustrate the level of efficiency of mediated money. In the UK’s CHAPS, the efficiency ratio before the GFC 2007–9 averaged 15 times, and after the crisis fell to 11 times (Benos et al, p. 166). Static (stock) indicators of in-network efficiency, such as the ratio of reserves to banks’ assets, show a greater range of oscillations. In Fedwire, from 8% (2010 and 2019) to 19% (2014 and 2021), which is primarily the result of oscillations in the expansiveness of monetary policy (Afonso et al, 2024, p. 13). These results show that the power of payments in network has not diminished, despite conventional wisdom.

The method for economizing of liquidity is the most common synonym for the power of netting transactions in bank networks (for example, see Johnson et al). This economizing of funds creates the illusion of transactions without money. Interestingly, the flow of money remains obscured in the clearing-netting process (“implicit” transactions) and becomes visible to observers only in the settlement phase (“explicit” transactions). Despite this, it should be understandable, at least intuitively, that in every clearing transaction our money, circulating via intermediaries, not imaginary money, but central bank money.

Clearing and settlement power in banking practice is often perceived as the magic of money transfers in banks’ books. And magic lending of deposit money. ‘But the English money is “borrowable” money. Our people are bolder in dealing with their money than any continental nation, […] their money is deposited in a bank makes it far it more obtainable’ (Bagehot, p. 5).

In the economic literature, clearing and settlement power is defined as multiplication of deposits and loans. Monetary collapse during the Great Depression exposed and discredited this power as an uncontrolled multiplication of deposit money. Irving Fisher, a pioneer of modern monetary analysis, considered that the instability of demand deposits is the primary cause of booms and depressions, which is why he advocated a full cash reserves or 100% money (Fisher, p. 119–120). ‘The destruction of check-book money was not something natural and inevitable; it was due to a faulty system’ (ibid, p. 7). Similarly, Hayek saw clearing and settlement power as the ‘perverse elasticity of bank deposits’ and ‘the seat of the trouble’.

Obviously, it is widely known that bank deposit business led to cyclical recessions and depressions, but it is also indisputable that it was the driving force behind the leading economies—from Venice and Amsterdam to London and New York.

The clearing and settlement power is vividly illustrated by the data from the world’s largest central bank—Federal Reserve Bank of New York, about the activities in Fedwire of the top 15 banks during the first 100 business days in 2020. These 15 banks were sending 76% of the dollar value of all payments sent by the top 100 entities. The same 15 banks held approximately 45% of the reserves held from the same 100 entities during the same period (Afonso et al, 2022, p. 7).
By comparing the share of sent payments (76%) and reserves (45%), we get a coefficient of 1.69, meaning that the top 15 banks used reserves 69% more efficiently than the top 100 banks! Or conversely, to send the same value of payments, they required 40% less reserves compared to the top 100 banks. Clearly, a higher clearing level and scope induces even greater power of payments. A possible gradation of clearing, based on the criterion of its level and scope, is given below.

CLEARING LEVEL AND SCOPE

INTERNAL clearing

BILATERAL clearing

MULTILATERAL clearing

SYSTEM clearing

NATIONAL clearing

INTERNATIONAL clearing

It is clear that the distinction of clearing cannot be strict due to overlaps – for example, multilateral with system or system with national. The only pure clearing is internal, at the level of an individual intermediary, i.e., a single bank. This basic clearing is characterized by the absence of other intermediaries and independence from the liquidity level of the intermediary itself, described by their stock-flow identities: ∑Inflow = ∑Outflow => Stock(t) = Stock(t-1). These two characteristics eliminate settlement risk and make internal set-off the safest clearing after national clearing under the control of the central bank. As previously explained, the greatest clearing and settlement power is held by the national clearing system – a comprehensive network with the central bank as the supermediator without technical limits. The inflation, i.e., purchasing power of money, is the economic limit.

Consequently, the most comprehensive and important indicator of clearing and settlement power is the ratio of daily payment value to annual GDP. ‘In 2020, the Fedwire Funds Service handled a daily payment value of over $3.3 trillion dollars, meaning that a sum equivalent to the (annual) GDP of the United States was turned over every 7 days or so’ (Afonso et al, 2022, p. 1). Although already hypertrophied relative to the needs of the real sector, this monetary power continues to grow unstoppably.

The evidence of the absence of technical limits in the expansion of clearing transactions using mediated money is provided below. For this purpose, a bilateral clearing model was developed, based on the BIS – BCBS practical example.

Evidence based on BIS – BCBS practical example

The Basel Committee on Banking Supervision (BCBS) at the Bank for International Settlements (BIS), ‘has developed a set of quantitative tools to enable banking supervisors to monitor banks’ intraday liquidity risk’ (BIS, 2013b, p. 1). To illustrate the use of these tools, the Basel Committee also provided Practical example, which served as the basis for the construction of our interbank payments model – the main proof of clearing and settlement power.

‘Practical example of the monitoring tools would operate for a bank on a particular business day’ BIS-BCBS was designed for two types of banks: 1. direct participant, and 2. bank that uses a correspondent bank. Substantially and numerically, both examples are identical (ibid, p. 13–14). For the description of clearing and settlement effects, it is also irrelevant to distinguish between payments for time-specific obligations and regular payments (ibid, p. 13). All details from the BIS – BCBS practical example, which were also used in our analysis of payments in the interbanking network, are listed below.

‘Details of the bank’s payment profile are as following: The bank has 300 units of central bank reserves and 500 units of eligible collateral. A(i) Daily maximum liquidity usage: largest negative net cumulative positions: 550 units; largest positive net cumulative positions: 200 units. A(ii) Available intraday liquidity at the start of the business day: 300 units of central bank reserves + 500 units of eligible collateral (routinely transferred to the central bank) = 800 units. A(iii) Total payments: Gross payments sent: 450+100+200+300+250+100 = 1400 units. Gross payments received: 200+400+300+350+150 = 1400 units’ (ibid). 

This BIS – BCBS practical example is schematic focused on an individual bank (Bank A), thus neglecting the nature of network payments. Therefore, our analysis introduces an additional bank (Bank B), taking into account the fundamental equation: sent payment A = received payment B, and vice versa. The result of such an extension of this simple model is presented below (Table 2).

       Table 2 – Bank A (initial reserves 300 units) vs Bank B (initial reserves 300 units)

      Sources: BIS (2013b), p. 13-14, and author’s calculation.

Both banks have the same payment profile: identical initial reserves (300 units), identical available intraday liquidity (800 units), and identical total payments (1,400 units). Only the payment dynamics throughout the business day differ, hence the usage of central bank reserves and intraday credit differs. Their net cumulative positions, negative and positive, are also different. More precisely, their signs differ: minus or plus (Table 3). ‘For the calculation of the net cumulative position, ‘payments received’ do not include funds obtained through central bank intraday liquidity facilities’ (ibid, p. 5).

The BIS – BCBS practical example is extremely simplified, and this derived model of bilateral intraday payments shares that trait. For instance, the assumed equality of sent payments and received payments within a single business day is an extremely rare occurrence, while typically there is some difference between them. The good side of this simple model of interbank payments is that it highlights the interdependence between the payer bank and payee bank and the fundamental principles of payments in network.

The initial distribution of available intraday liquidity is ideal: fifty-fifty. For certain reasons, Bank A sends more than half of the total intraday payments already by 10:00 (the first and largest payment of 450 units is not explained), while its received payments are almost four times smaller. For such early payments, it had to use intraday credit from the central bank. On the other hand, Bank B nearly tripled its initial reserves, achieving a noticeable concentration of liquidity. By the end of the business day, total payments were equalized, so reserves returned to the initial level. The beauty of this example is that it clearly shows how, in an RTGS system, early payment is penalized, while late payment is rewarded: Bank A had to use intraday credit, Bank B did not!

The most important proof is that all payments remained within the network, i.e., in bank books. Simply put, received payments are used by the bank for sent payments (so-called ‘liquidity recycling’), the missing amount is covered by intraday credit, while excess reserves are offered to other banks through the money market. All payments were executed with short-term credit support from the central bank, and both banks retained the same reserves as at the beginning of the day. This is not the alchemy of payments, but the clearing and settlement power held by all payment intermediaries. The power of mediated money.

            Table 3 – Net cumulative positions – negative and positive (Bank A vs Bank B)

            Sources: BIS (2013b), p. 13-14, and author’s calculation.   

Net cumulative position (received payments minus sent payments, transaction-by-transaction), negative and positive, directs the use of available intraday liquidity. ‘The largest net negative position during the business day on the account(s), (i.e. the largest net cumulative balance between payments made and received), will determine a bank’s maximum daily intraday liquidity usage.’ Maximal net negative position Bank A (-550 units) and Bank B (-200 units) are within their available intraday liquidity, which means there is no liquidity risk. Conversely, if the largest net negative position were to exceed available intraday liquidity, there would be a danger of gridlock in payments. In such a case, it would be necessary to seek additional intraday liquidity sources. That is why banking supervisors continuously monitor any intraday liquidity shortfall.

The model sketched in Table 3 also clearly demonstrates the interdependence of sent and received payments in a bilateral interbank network and two fundamental rules:

Gross payments sent A = Gross payments received B, and vice versa and

Net negative cumulative position A = Net positive cumulativ position B and vice versa.

The previously emphasized the largest net negative cumulative position (LNNCP) of Bank A necessarily creates the largest net positive cumulative position (LNPCP) of Bank B (abbreviations according to ECB, 2024). More precisely, LNNCP Bank A = LNPCP Bank B, and vice versa. The conclusion is obvious: liquidity has not disappeared, but its distribution has significantly changed, with reserves concentrated in Bank B’s account. Therefore, the BIS-BCBS view of deposit outflows and similar phenomena is too narrow. The real question is where the deposits flowed—specifically, which portion was transferred to other banks, and which portion left the banking sector (e.g., by purchasing government bonds and central bank securities in primary markets, or through tax payments). Most often, the majority of deposits remain within the banking sector; only the deposit-holding bank changes. This phenomenon results from the systematic suppression of cash from payment flows, which escalated dramatically in the 2010s. Consequently, an analysis of payments and settlement processes in the banking sector becomes not only as important as the investigation of the intraday liquidity of individual banks, but even more important.

The banking sector (B) is a set of banks (b). According to set theory, it follows that b ∈ B, and ∈ is referred to as the membership relation. Mathematically, ‘all relevant facts about sets can be expressed in terms of the membership relation.’ The BIS–BCBS practical example is presented without the membership relation—i.e., without relevant facts about the banking sector! By focusing on the intraday liquidity of an individual bank, the practical example neglects the essential membership relation within the banking sector.

Introducing a larger number of banks into the model does not alter the aforementioned principles of bilateral payments. Assuming there are no inflows or outflows of reserves into or out of the banking sector (BS), and that intraday credit from the central bank is repaid by the end of the business day, the rules remain the same:

Gross payments sent BS = Gross payments received BS,

            Net negative cumulative position BS = Net positive cumulative position BS.

For the sake of analytical precision, the net negative cumulative position of the banking sector represents the aggregate of such positions across all individual banks, whereas the net positive cumulative position reflects the sum of all positive positions held by banks at a given point in time. While the absolute magnitudes of these positions fluctuate continuously throughout the business day, they remain perfectly offsetting in value, differing solely by sign—negative or positive. The same rule applies to the largest net negative cumulative position of the banking sector: LNNCP BS = LNPCP BS. By definition, their resultant is zero, providing further empirical validation of the inherent network effects of payment systems. This also reinforces the argument regarding the strong dependence between interbank payments.

In the presented model of interbank payments (Table 2), the banking sector is composed of Bank A and Bank B. The sector’s transactional relationship with the central bank during the payment process is primarily represented through two key variables: reserves and intraday credit for liquidity. Within this framework, the principal indicator of sectoral payment transfers is the volume of sent payments, which—by definition in this model—corresponds precisely to the volume of received payments. Table 4 below outlines the projected changes in these variables under three scenarios: initial sector reserves of 600 units (as in Table 2), then reduced initial reserves of 300 units (Bank A = 150 units + Bank B = 150 units), and finally, a scenario with no initial reserves – 0 units. All projections are derived from data presented in the Basel Committee Practical example.

Table 4 – Banking sector: total initial reserves 600 units vs 300 units vs 0 units

              Sources: BIS (2013b), p. 13-14, and author’s calculation.  

A comparison of reserves and credit reveals that the sum of intraday credit determines the level of aggregate reserves, thereby confirming that these credits represent the most important source of external liquidity for the banking sector. This is most clearly demonstrated in the third scenario with zero initial reserves, where the relationship aggregate reserves = total credit holds true. This final example illustrates the extent of clearing and settlement power even in the absence of initial reserves.

The ultimate evidence of the superiority of payments in network lies in the fact that banks can sustain operations for several days, or even weeks, without borrowing from the central bank or accessing the money market, without liquidating assets, and with minimal reserves. However, without incoming payments, they would not survive even half a working day. Put simply, incoming payments cover approximately 85–95% of a bank’s outgoing payments.

7 RTGS: the way backward

The net settlement model dominated operations in clearing houses for over two centuries, up until the early 1990s. At that period, central bankers began favoring explicit transactions with explicit money at the account, i.e., gross settlement (Leigh-Pemberton, p. 454). ‘I am quite clear, therefore, that real-time gross settlement is the way forward, […]’ (ibid, p. 452). Since then, the stigma of implicitness has followed the net settlement model, although it has managed to survive in the shadow of RTGS as the most efficient method to reduce settlement risk.

Among central bankers, the belief took hold that they had discovered a quicker and more efficient system. ‘In this case, large-value payment systems would be able to run more quickly and efficiently on the basis of sound banking principles: funds must be in an account at a central bank before payment transfers will be honored’ (Angell, p. 7). This became the core idea behind the construction of RTGS systems. Therefore, in the RTGS model, ‘[…] offsetting payment obligations cannot be netted and must instead be pre-funded. […] Banks do not have to fund all of their payments directly, however. Liquidity is recycled throughout the day as banks use incoming payments to fund outgoing ones’ (Benos et al, p. 144–145). The cited sound banking principles represent a rejection of the clearing house concept and its practice. Of course, netting could not disappear entirely—it was merely suppressed (via incoming payments) and made more difficult (by requiring funds at account).

As a result, all participants were effectively pushed into a waiting room. A simple example: two banks hold $10 million in mutual payments in the morning, both on their own behalf and on behalf of their clients. However, the smallest individual payment is $800,000, and neither bank has that amount in its account at the central bank at the beginning of the day. The banks are formally illiquid, despite holding sufficient settlement assets to meet their total obligations, and they face congestion. Each bank is required to send the full amount of money to the other bank at account. From this metallist’s view, does it follow that a clearing instruction is not a payment order, but rather a worthless message? In order to initiate the chain of payments, banks must wait for incoming payments from other banks, or borrow from the central bank or on the money market.

This simple example highlights a fundamental design flaw of RTGS payment systems. ‘The fundamental problem of a payment system is whether promised payments will actually occur’ (Kahn et al, p. 3). Hence, the gridlock of payment flows is the chronic illness of RTGS systems. This illness was not chronic in the clearing houses of previous centuries: ‘[a] key advantage of net settlement over RTGS, which is that net settlement can eliminate gridlock equilibria, […]. Net settlement is effective because it de facto gives absolute priority to offsetting claims, since such claims are automatically discharged under net settlement rules’ (ibid, p. 17).

Notions such as gridlock, congestion, delays, cascades, shortfalls, deadlocks, queuing, hoarding, and many others are frequently found in titles of research papers on RTGS systems, describing various functional problems. The primary cause lies in the metallist concept of explicit transactions, or so-called gross settlement, which operates on a transaction-by-transaction basis. ‘In a gross settlement system, on the other hand, the settlement of funds occurs on a transaction-by-transaction basis, that is, without netting debits against credits’ (BIS, 1997, p. 5). In short, it functions without clearing, on a purse-to-purse model.

Despite the dominance of RTGS systems in wholesale payments, the practice of gross settlement could not eliminate the interdependence between incoming and outgoing transactions, nor the restrained use of reserves. ‘[W]e find evidence that banks in the U.S. still economize in the use of intraday liquidity and rely on incoming payments to make outgoing payments, generating a high degree of strategic complementarity in payment decisions’ (Afonso et al., 2022, p. 15). The persistence of clearing or netting technology, as already emphasized, is rooted in a fundamental identity: sent payments = received payments in network. From this identity, one can derive a function that describes the position of each individual payment intermediary within the network of intermediaries: total payments = f (cumulative inflows). At the same time, this provides further evidence that in any settlement system, whether gross or net, payment decisions are governed by absolute complementarity—the liquidity of one participant depends on the payments of all others and on the liquidity of the entire system.

Net settlement has been accompanied by two pejorative attributes – implicit transactions and deferred payments. The first has already been clarified as a product of monetary blindness, which fails to recognize clearing as a legitimate method of payment. The second is tied to the early days of modern clearing houses, in the late 18th and early 19th centuries, when bankers would meet at the end of the business day to perform clearing of mutual claims and, finally, settlement. This practice persisted for the next two centuries, and such systems became known as Deferred Net Settlement (DNS). Even today, there are similar private payment systems operating as hybrids between the RTGS and DNS models. By far the most well-known and largest such system, both in the United States and globally, is the CHIPS network. However, deferred settlement does not mean delays of payment—it simply refers to the system rules. Technically, net settlement can be performed every minute throughout the business day. Therefore, RTGS systems have since been upgraded by introducing netting schemes to bridge illiquidity.

All available evidence points to the conclusion that RTGS systems are not superior, but rather inherently inferior to net systems. Nevertheless, gross settlement systems have become dominant; their expansion during the 1990s was driven by the need for safer foreign exchange transactions and securities trading. The imposition of RTGS as the standard model for large-value payment systems had two key motivations. First, to minimize the risk of default in the largest financial market—the global foreign exchange market—through the use of the Payment versus Payment (PvP) mechanism. Second, to mitigate problems of securities delivery and funds transfer through the Delivery versus Payment (DvP) mechanism. Technically, both PvP and DvP would not be able to function without RTGS. Yet, this trinity of gross settlement mechanisms neither prevented nor mitigated the GFC 2007–9.

Thus, these anachronistic gross settlement models (RTGS with PvP and DvP) have proven to be the way backward. A potential solution could lie in Real-Time Net Settlement (RTNS), based on accelerated sequential netting. A proposal for an RTNS payment system is outlined in the next section.

8 RTNS solution: accelerated sequential netting

The basic prudence post-crisis measures focus on short-term resilience of a bank’s liquidity and longer-term sustainability of its solvency. These measures involve two minimum standards: Liquidity Coverage Ratio (LCR) and Net Stable Funding Ratio (NSFR), which form the backbone of the Basel III framework (BIS, 2010). For the analysis of the RTNS solution (Real-Time Net Settlement), the LCR standard is primarily relevant.

LCR = Stock of high-quality liquid assets / Total net cash outflows over the next 30 calendar days ≥ 100%

The numerator of this ratio is commonly referred to by the acronym HQLA (High-Quality Liquid Assets). ‘At a minimum, the stock of liquid assets should enable the bank to survive until Day 30 of the stress scenario’ (ibid, p. 3). The denominator of this ratio = outflows – Min {inflows; 75% of outflows}. Cash outflows include retail deposit run-off, outflows of other deposits, as well as unsecured wholesale funding run-off (ibid, p. 12–16).

The definition of the LCR standard reveals the theoretical basis of a partial approach, previously analyzed through the BIS–BCBS practical example. Deposits, by definition, can only flow into other banks, not outside the banking sector, just like most wholesale funding. An exception is the portion that is redirected toward the purchase of government bonds and central bank securities.

However, even under a more comprehensive approach that evaluates the level of systemic liquidity, a high level of reserves in the banking sector does not imply timely payments. On the contrary, in a stress scenario, bankers instinctively engage in hoarding of liquidity, driven by precautionary motives and speculative purposes (Acharya and Merrouche, p. 152). This leads to an uneven distribution of liquidity within the banking sector and a slowed movement of reserves, nearly to the point of freezing. Nevertheless, the most important generator of crisis is fear, although it is largely overlooked in analyses of systemic illiquidity. Therefore, it is specifically emphasized here in the development of a liquidity crisis.

The pattern of liquidity crisis:
Distress ⇒ Uncertainty ⇒ Untrust ⇒ FEAR ⇒ Hoarding ⇒ Panic ⇒ Sale ⇒ Freezing

With the onset of hoarding of reserves, the links between incoming and outgoing payments—which are the core of mediated money, or payments in a network—rapidly break down. In the panic phase, comprehensive de-networking threatens the liquidity of all banks, even though they attempt to isolate themselves. However, the illiquidity of banks is likely the only ‘contagion’ which is actually worsened by the isolation.

The description of the banking panic of 1933 best illustrates this process of breakdown:
‘The System was demoralized. Each Bank was operating on its own. All participated in the general atmosphere of panic that was spreading in the financial community and the community at large’ (Friedman–Schwartz, p. 391). It is evident that in interbank payment systems, all banks are mutual counterparties, regardless of whether they have direct payment relationships or not.

Therefore, the spirit of cooperation and mutual trust among participants is necessarily stronger in the CHIPS network than in Fedwire, or in CHAPS than in the BoE’s RTGS system. After all, the first clearing house was created precisely due to the spirit of cooperation among participants and their mutual trust (Thornton, p. 101). Respect for clearing house rules has always been unquestioned, even when those rules limited excessive balances by defining a “maximum available balance” (PRC).

A more detailed comparison of clearing house rules and the LCR minimum standard shows that internal rules have been more important than the official standards in achieving a more balanced distribution of systemic liquidity and in preventing precautionary hoarding of reserves. Nevertheless, neither rules nor standards would have had significant impact on the prevention of systemic liquidity risk without the traditional measure – expansionary monetary policy, but on an unprecedented scale. ‘The massive expansion of reserve balances since fall 2008 and the payment of interest on reserve balances have altered the intraday liquidity management practices of financial institutions’ (Bech et al, p. 2). This was a true Big Bang of central bank mediated money:
‘Consequently, the level of reserve balances ballooned from $10 billion on average during August 2008 to $850 billion by year-end’ (ibid, p. 12).

Despite such excessive reserves in the payment network, which remained largely abundant during the 2010s and early 2020s, there are no guarantees that systemic illiquidity contagion has been preempted. This was clearly demonstrated by the eruption of the international liquidity crisis in March 2023. The explanation lies in the pattern of liquidity crisis, which reflects the behavior of payment intermediaries, as well as game theory, which assists in analyzing their strategies in cooperative and noncooperative games characterized by strong interdependence.

Up to the phase of untrust in the crisis, banks behave relatively cooperatively, and then conflict actions prevail along with the domination of own interest. Understandably, fear triggers the survival instinct, which ignores a common interest and leads to deeper isolation from other participants. A collapse ensues if the central bank and government do not intervene quickly and decisively.

Hence, the eternal search continues for the ‘alchemical liquidity stone’: ‘[A]n adequate regulatory mechanism should be in place to prevent banks from being “too interconnected to fail” and thus reduce the likelihood of a system-wide liquidity strain occurring’ (Silva, p. 2). Unfortunately, such a mechanism cannot exist due to the total interdependence in the payment system: all actions of all payment agents are interdependent. The gross settlement system of transaction-by-transaction is precisely a product of such an alchemical attempt, which survives only on the support of expansionary monetary policy.

On the other hand, the gross settlement system of transaction-by-transaction has reinforced a competitive spirit and blurred the conditionality of own interest by common interest. Since the early 1990s, the payment game has become increasingly noncooperative.

The insoluble problem is the conflict-dominated own interest in a system of strongly interdependent participants. At first glance, the dominating strategy of each participant in the payment system is to release outgoing payments at the end of the working day. In essence, this is a self-destructive strategy. In the payment system, a common interest maximizes the own. The most important components of common interest are system liquidity and payment efficiency. These goals are inevitably threatened by competitors’ confrontation in a noncooperative game. Even in the case of an oligopoly game, the oligopolist succumbs to the pressures of uncertainty and untrust. In the end, a cooperative game can to some extent be externally imposed on participants—through law, central bank regulation, and payment system rules.

The gross-settlement system of transaction-by-transaction causes a form of dictatorship of incoming payments, i.e., de-networking of interbank payments. As a result, only reserves at the bank account are perceived as one’s own liquidity, while incoming payments are viewed as free funds from other participants. In doing so, they separate their outgoing payment from our incoming payment, even though it is in fact a single unified transaction between the payer bank and the payee bank. This form of de-networking in gross settlement can be described in a simple schematic way.

The elucidation of gross settlement system:

Account A => outgoing payment / incoming  payment => Account B => outgoing payment / incoming  payment => Account A.

Opposite to him is the elucidation of net settlement system:

Account A => outgoing payment   => SETTLEMENT  <= outgoing payment  <= Account B.

The proposition of the RTNS (Real-Time Net Settlement) payment model is based on the presented net settlement system. Briefly – banks send their own outgoing payments to the system operator by payment messages for sequential multilateral netting and settlement at the end of predefined cycles – minimum every two minutes, maximum fifteen minutes. The two-minute sequence corresponds to the maximum cognitive capabilities of liquidity managers for monitoring a payment algorithm and dynamic decision-making, although the technological capacity for processing payments is nearly unlimited in the AI era. It is evident that such accelerated sequential netting is incomparably less exposed to systemic liquidity risk than deferred net settlement. For example, the liquidity risk of a two-minute cycle is approximately 300 times lower than end-of-day settlement for ten working hours. Likewise, credit risk is proportionally reduced, as well as overall settlement risk.

The fundamental rule is that each payment message = payment order. At the end of the cycle, the system operator calculates interbank obligations from payment messages through multilateral netting. Offset payments are immediately entered into system records and bank books as final, irrevocable, and unconditional, without any deferment. Remaining non-settled payment messages are placed in queuing according to the system’s rules and instructions from the payer bank.

The main motive of banks for a deferred cycle is to collect as many payment messages as possible for facilitated netting and settlement. However, for high-value payment systems, the problem is not the accumulation of payments, but rather their restriction. For instance, the value of a single payment in Fedwire is limited to one cent less than $10 billion. The bigger problem is the insufficiency of small transactions to complete the multilateral mosaic of clearing. Prefunded balances also can help to complete clearing, as the maximum possible net debit position. These balances in net settlement systems are smaller than in RTGS, by a definition (Dent and Dison, p. 236). In addition, intraday credits of the central bank can be used.

In the proposed RTNS payment system, participants will be strongly motivated to send as many payment messages earlier, instead of transferring scarce and costly reserves from their transaction account. Friedman’s monetary deceiving – the interpretation of netted payments as implicit transactions – would have a strong effect here. Of course, the truth is the opposite: netted payments are explicit transactions, just like all other easily visible moneyness transactions. From this arises the general rule of payment: payment flows = money flows.

For a comparison of the RTNS model’s performance with prevailing RTGS payment systems, the BIS – BCBS practical example, already analyzed in Section 6 (pp. 20–24), can again be used. In the presented projection of interbank payments (Table 2), the banking sector, i.e., the payment network, consists of Bank A and Bank B. The entire analysis was conducted according to RTGS rules. In the upcoming Table 5, alongside RTGS payments, we will simulate net transactions in the payment network under RTNS assumptions – columns Clearing and Queuing. It should be noted that the RTGS simulation used bank reserves and intraday credits, while the RTNS projection excluded these, relying in this case on pure clearing and a queuing mechanism.

             Table 5 – RTGS vs RTNS  

            Sources: BIS (2013b), p. 13-14, and author’s calculation.    

The simulation shows that RTGS payments were processed faster, which is expected given the use of bank reserves and intraday credits. An additional influence on this result was the unusual distribution of payments in the BIS – BCBS practical example – the accumulation of payments (50%) in the first quarter of the working day. Even under such adverse conditions, the RTNS model proves superior, as payments are executed in short cycles, which in this case were projected at five minutes.

The following Table 5a vividly illustrates the dynamics of RTGS and RTNS payments per each five-minute settlement cycle and in total. All payments in the cycles represent the sum of payments by Bank A and Bank B (1,400 + 1,400 = 2,800). In the RTNS model, all netted payments are summed within the predefined cycle according to the rule:

Account A => outgoing payment => SETTLEMENT <= outgoing payment <= Account B.

Table 5a – Payment dynamics: RTGS (with reserves and  intraday credits)                   vs RTNS (without  reserves and intraday credits)

                  Sources: BIS (2013b), p. 13-14, and author’s calculation.     

The comparison of payment effectiveness between the RTGS system, with reserves and intraday credits, and the RTNS model, without reserves and intraday credits, testifies to the superiority of the proposed model. It is necessary to emphasize that the author’s motive for proposing the RTNS model is not ‘saving’ liquidity or faster liquidity ‘recycling’, but the creation of more reliable, liquid, and efficient wholesale and retail payment systems.

The superiority of net settlement systems is proven in practice by the record-setting efficiency of the largest system of its kind in the world – the CHIPS network. ‘The liquidity efficiency of CHIPS averaged 26:1 in 2023, meaning that $1 contributed to the network in funding supported $26 in settled value, with most payments settling in mere seconds. This contrasts with the average of 6.6:1 liquidity efficiency in other major large-value payment systems around the world’ (CHIPS News, April 02, 2024). The cited CHIPS network payment efficiency can be explained by its fundamental characteristics. ‘CHIPS is a real-time final settlement system that continuously matches, nets and settles payment messages. […] On a daily basis, the system provides real-time finality for all payment orders released from the CHIPS queue’ (BIS-CPSS). Additional support for the functioning of this system is provided by the CHIPS prefunded balance account (ibidem).  

The proposed RTNS model differs from the CHIPS network by its own accelerated sequentiality and finality at the end of a predefined cycle. However, the most important thing is that funds on account and funds in a payment order are treated equally in settlement processes. Thus, money flow is not hidden behind money stock.

According to an official source from the Federal Reserve, RTNS systems already exist.
‘Some countries use real-time net settlement, where transactions net every few seconds, essentially in real time’ (FRB services, p. 5). However, a detailed internet search reveals no such payment system.

Finally, the proposed RTNS model can be succinctly defined as a real-time net settlement payment system with accelerated sequential multilateral netting/clearing in predefined cycles (from minimum two minutes to maximum fifteen minutes) on a final offsetting basis, along with the use of prefunded balances and intraday credits.

Immediate settlement certainly does not convert mediated money into immediate money. It still remains in the intermediaries’ possession. Solving this structural problem of mediated money could be helped by the balance sheet reconstruction, presented in the following section.

9 Balance sheet reconstruction

Let us revisit the LCR minimum standard: LCR = Stock of HQLA to Total net cash outflows over the next 30 calendar days ≥ 100%. This total net cash outflows = outflows – Min {inflows; 75% of outflows}. Cash outflows include the retail deposit run-off and other deposit outflows, as well as unsecured wholesale funding run-off (BIS, 2013a). Deposits cannot flow out of the banking sector, so systemic liquidity can only be threatened by wholesale funding run-off, i.e., the freezing of the money market. From experience, we know that in such a scenario, the official haircut: 75% of outflows can quickly drop to 50%, then to 25%, and finally to 0% of inflows. Under these conditions, even LCR ≥ 200% or 300% is insufficient, as shown by the international banking crisis in March 2023. Liquidity in freeze, like any fluid, dramatically slows its movement — that is, its circulation.

The above-mentioned problems with external wholesale funding are most often referred to as dysfunctionality of the financial market. They are less frequently described as liquidity evaporation, which is incorrect — liquidity cannot evaporate; that illusion results from its freezing. Regardless of the different interpretations, the systemic liquidity problem requires critical support from the central bank to prevent the collapse of the financial system, as in September 2007. For that purpose, a large-scale central bank balance sheet expansion is used to quickly inject additional liquidity into the financial system and the economy (Bernanke, p. 9).

The central bank balance sheet before September 2007 was of interest only to its accountants. After that, it moved to the epicenter of policymakers’ attention, although even earlier, all monetary policy decisions directly caused changes in the size and composition of the central bank’s balance sheet. Nevertheless, it is more appropriate to emphasize the level of the central bank balance sheet than the notion of ‘printing’ central bank money or unprecedented monetary expansion. In this way, the central bank balance sheet rose to the highest pedestal of monetary policy.

‘The central bank’s balance sheet plays a critical role in the functioning of the economy. The main liabilities of the central bank (banknotes and commercial bank reserves) form the ultimate means of settlement for all transactions in the economy. Despite this critical role the central bank’s balance sheet remains an arcane concept to many observers’ (Garreth, p. 1). The attributes arcane or mysterious form a veil that conceals the true nature of mediated money in this balance sheet.

Thus, it became possible to rename the power of mediated money into ‘the power of the central bank balance sheet’ (Orphanides, p. 2), which is, after all, just a sheet. And that power is indeed nuclear. ‘Between March and June 2020, the Fed expanded its balance sheet by $3 trillion. In three months, the Fed ‘printed’ as much high-powered money as it did over the first 100 years of its history, from 1913 to 2013. The Fed was not alone; the BOJ and the ECB engineered similarly massive balance sheet expansions’ (ibidem).

During the GFC, an exit strategy was devised.‘When the economic outlook has improved sufficiently, we will be prepared to tighten the stance of monetary policy and eventually return our balance sheet to a more normal configuration’ (Bernanke, p. 9). Despite all monetary policy measures and policymakers’ expectations, the configuration of the balance sheet became increasingly abnormal. In the end, it turned out that the ‘balance sheet would ultimately have to be larger than before, to take account of long-term changes in the demand for liquidity’ (Hauser, 2021), although the effort to reduce the oversized balance sheet continues (Fed, 2024).

Consequently, the central bank balance sheet has become a countercyclical policy tool.
‘Policy actions taken for both monetary policy and financial stability purposes will typically imply that balance sheet adjustments are countercyclical. […] However, the case that monetary policy operations should give rise to a counter-cyclical balance sheet is less well explored. In part that is because policy actions that give rise to large changes in the balance sheet have become commonplace only relatively recently’ (Bailey et al, pp. 25–26).

The proposed balance sheet reconstruction is intended precisely to mitigate too big changes in the balance sheet and frequent cyclical oscillations, despite the fact that such approaches are still less well explored and relatively recently used. The very term balance sheet reconstruction implies changes in the composition of the central bank balance sheet, as issuer and supermediator of central bank money, on one hand, and commercial banks, as the most important payment service providers, on the other.

For this purpose, stylized balance sheets were used, containing only the relevant items to enhance understanding of the structural changes and to enable a clear comparative analysis. For the same reason, sectors (Households, Firms, Financial Sector, Government, Foreign) are not separately listed, but are implicitly included in the balance sheet projections. In other words, it was necessary to project average proportions that approximately correspond to the actual structure of representative central bank and commercial bank balance sheets. The commercial banks’ balance sheet is consolidated (interbank liabilities are excluded).

The stylized balance sheets are presented in parallel, so that the key relationships (identities) are easily visible—above all, the reserves on the liabilities side of the central bank and those same reserves on the assets side of commercial banks. ‘The reserve balances held by commercial banks are the highest quality, most liquid asset they could hold. Pre-crisis, the main use of reserves was to meet interbank payments flows, cleared and then settled across reserve accounts’ (Paul Fisher, p. 2).

However, what is most important for this analysis is that the presented stylized balance sheets clearly express the two primary forms of money: immediate money (banknotes) and mediated money (reserves), as well as their relation to commercial bank promise of money, i.e., fractional money (current accounts).

Stylised balance sheets – existing shapes

Stylised central bank balance sheet                        Stylised commercial banks balance sheet

Assets                      Liabilities                                Assets                           Liabilities

Loans            300     Banknotes     100                     Reserves        400          Current acc.       800

Securities      500     Reserves        400                     Loans          1,800          Deposits          1,500

Other assets   200     Other liabil.   495                    Securities     1,200          Other liabil.    1,500

                                  Capital              5                     Other assets    600          Capital              200

Total:          1,000     Total:         1,000                     Total:          4,000          Total:             4,000

It is necessary to emphasize clearly that current accounts correspond to “reservable transaction accounts”, while deposits represent “non-reservable saving deposits” (according to Fed’s Regulation D), as well as other deposits (time deposits, wholesale deposits). This terminological distinction between accounts and deposits is not merely formal, but essential for understanding the difference between money and financial investments, regardless of how much they are “near money”.

The mutual relationships between banknotes, reserves, and current accounts (i.e., demand deposits) form the hub of most monetary research and analysis, cause numerous controversies and doubts about the nature of money, and trigger all panics and bank runs. They are also the origin of various proposals and solutions, such as the gold standard, 100 percent reserves, requirement reserves, narrow banking, reserve-to-deposit ratio, deposit insurance, and the latest — the liquidity coverage ratio and net stable funding ratio.. That is why these relationships remain an inspiring basis for research into the ‘mystery and paradox’ of money. 

The first and most important insight provided by the existing structure of commercial banks’ balance sheets is that banks fully possess the mediated money of their clients (current accounts on the liabilities side). The second key insight is that commercial bank reserves are significantly smaller than current accounts (fractional reserves).

All previously mentioned solutions have aimed at mitigating the problem of fractional reserves, but none have been sufficiently effective, because they did not remove the fundamental cause – the relative free bank’s disposal of balances on current accounts their clients. Against these partial solutions, the proposals based on full reserves would lead to disintermediation and endangerment of the bank lending mechanism.

In fact, what is needed is to straighten bank balance sheet’s tangles that date back to 1844. ’The different treatment of cash and non-cash money, or banknotes as a regulated business and bank deposits as a free market activity, was first legislated in the UK’s Bank Charter Act 1844. The intention was to split the Bank of England’s note issuing operations from its general banking business by creating distinct Issue and Banking Departments. The Act therefore separated central bank money for the public (Issue Department) from central bank deposit money, available only to banks (Banking Department). So, the general public, denied access to risk-free central bank deposits for transactions, was forced to use fractional bank deposits for non-cash payments, although before the Act was passed the Bank used to accept individual deposits’ (Vuković, 2020b, p. 3-4). 

Resolving of this 1844 tangle clearly lies in restoring public access to risk-free central bank money for public. This does not mean that commercial banks would lose their status as PSPs. On the contrary, their role in payment systems would be strengthened. ‘Banks are an inevitable payment intermediary, and so safe transaction accounts would have a positive effect on the stability of their business and improve their less than shiny perception in public’ (Vuković, 2020b, p. 6). The only difference is that banks would no longer be able to directly use the money from their clients’ current accounts, but only indirectly – through the central bank loans.

The potential solution is relatively simple. Current accounts, i.e., demand deposits, would be transmitted from the commercial banks liabilities to the central bank liabilities. In return,
banks would receive compensatory loans, proportionate to transaction accounts balances their clients. These loans in the assets of the central bank would be permanent and could be calculated based on account balances at the end of the business day or based on the daily average at the end of the business week. All other balance sheet items would remain unchanged after this reconstruction, as illustrated in the projected balance sheets below.

Stylised balance sheets after reconstructions

Stylised central bank balance sheet                   Stylised commercial banks balance sheet

Assets                          Liabilities                       Assets                      Liabilities

Loans             300        Banknotes       100           Reserves       400      Compensatory loans     800

Securities        500         Reserves          400            Loans          1,800      Deposits                       1,500

Comp. loans   800        Current acc.     800           Securities    1,200     Other liabilities          1,500

Other assets    200       Other L + cap.  500          Other assets    600     Capital                          200

Total:          1,800        Total:             1,800           Total:          4,000      Total:                        4,000

Consequently, the size of the central bank balance sheet would be increased by the amount of the transferred current accounts in liabilities and compensatory loans in assets. The projected increase of 80% (which in practice would certainly be twice as small) may seem too big, but it is significantly smaller than the extreme oscillations of the central bank balance sheet since the GFC. However, this would represent zero monetary expansion, because the compensatory loans would be entirely financed from current accounts funds.

Banks would still record and manage their clients’ current accounts, but exclusively as PSPs, charging for their services as usual. This means that banks would not lose revenue from payment services, nor would there be even the slightest disintermediation. On the contrary, banks would retain a surplus reserves originally intended for current accounts transactions. It is clear that bank payment services would be performed off-balance sheet. Banks would retain access to the same amount of reserves for their transactions with savings, time, and wholesale deposits, as well as for all other payments. However, they would no longer be able to directly use the current accounts of their clients. It is understood that current accounts, as a liability of the central bank, would be interest-free.

The transactions involving current accounts would be executed within a pure clearing payment system, in which clearing = settlement. This payment system would be managed by the central bank, acting as a supermediator, without any need for additional liquidity or loan interventions.
It would be a kind of internal clearing within the central bank payment network, without frictions. Payments would depend solely on the liquidity of the payer, who would not be able to issue a payment order without full coverage on their current account. Of course, from the same current account, the payer could transfer funds to a bank as savings or time deposits, and could also receive loans from the bank into the same account. The pure clearing payment system would be connected with all other domestic payment systems, in the same way that, for example, the Fedwire and CHIPS networks are interconnected.

It is evident that in such a pure clearing payment system there would be no need to introduce a special central bank digital currency (CBDC), as it would already be created through the proposed reconstruction. The only digital currency still lacking would be a retail CBDC for offline transactions, primarily intended for unbanked people (Vuković, 2021).

Additional benefits of the proposed reconstruction are numerous: containment of digital bank runs in the future, significantly lower pressure to shift deposits between banks, proportionally reduced need for deposit insurance, transfer of credit risk from account holders to the central bank, reduction of settlement risk in existing payment systems, stronger interbank competition in the payment services market, improved supervision of banks and increased perception of financial stability, and more.

For the general public, these changes would be almost imperceptible. ‘Unsurprisingly, most did not see the difference between central bank money and commercial bank money’ (Kantar Public, 2022, p. 7). Nevertheless, the public would need to be clearly and simply informed about the safety of their money under the direct umbrella of the central bank, as well as about the efficiency of payment transactions via their current accounts. The only thing that could threaten such ultimate security and efficiency is the central bank itself, if would by extremely expansionary monetary policy to trigger undesired inflation and thus eroding the value of money.

By granting the public direct access to non-cash central bank money, the sinister threats of a cashless society would be averted. ‘The Swedish central bank has an answer as to what that society will look like: ‘In a cashless society, a restricted group of financial corporations has access to risk-free central bank money. The general public, in contrast, does not’’ (Vuković, 2020a).

The changes in the composition of central bank and commercial bank balance sheets would represent a kind of defractionalization of the banking system. Such a reconstruction would require clear political support, in favor of both the public and the banks. After the expert and political debate, the proposed changes would need to be incorporated into relevant laws and regulations, and subsequently into the central bank’s supervision rules and banking regulations.

Finally, in summary: the stock of mediated money would no longer be in the possession of intermediaries. The intermediary would be a custodian, disable to freely use money on current accounts. Certainty and finality of payments would not be depend from liquidity of intermediary, but only from payer. And most importantly, the public would gain direct access to mediated central bank money.

10 Conclusion 

The main finding of this research is that money has two primary forms – immediate money and mediated money. The former circulates off-line, and the latter in-network, via intermediaries. Immediate money performs its functions without intermediaries, circulating hand-to-hand. Mediated money circulates exclusively account-to-account in the books of payment institutions (in-network). Payments and other functions of mediated money are realized only in-network.

The payment function of immediate money is performed instantaneously, finally, irrevocably, and unconditionally. The stock of immediate money is always in the possession of the money-holder. If the holder entrusts their stock of coins and/or banknotes to a deposit institution, it transforms into mediated deposit money.

Mediated money processes payments internally (in-bank) and via clearing houses and settlement systems. Certainty and finality of payments depend on the liquidity of the payer and its intermediary. The stock of mediated money is always in the possession of intermediaries. Intermediaries are not custodians, and can freely use the deposited money.

These two forms of money have coexisted, intertwined, and conditioned one another throughout their multi millennia-long evolution. Both primary forms embody the same functions of money, and illuminate their different creation and circulation, without being obscured by material forms. Their shared evolution has often hidden their fundamental differences caused by their mode of circulation – off-line versus in-network. The decisive difference lies in the power of clearing and settlement in payments, although many well-known textbooks about monetary economics and banking do not even contain the terms clearing and settlement.

*****

The oldest archaeological evidence comes from the temples in Mesopotamia, which first developed immediate metal money as a standard of weight and then mediated (accounting) money in the third millennium BC. Weighed metal is by all accounts the embryo of money – known units of weight (such as the shekel, the talent) served as monetary units for at least two thousand years before the first coins. The weighed metal money, which regularly stamped, enabled the development of wholesale barter, but above all as a means of settlement.

Coined metal money never had such importance for the wholesale trade. Nevertheless, coins played an epochal role by introducing people across the known Antique world to coined money and its military, commercial and sacral powers. Coins remained on the pedestal of true or proper money for over two and a half millennia – they were the benchmark for representative money until the 1930s. Banknotes are the youngest and most advanced shape of non-mediated money, but have never been the dominant means of payment.

According to their issuers, coins and banknotes are classified as central bank money. Consequently, with their disappearance, immediate central bank money will also vanish, and retail payments will be conducted exclusively in-network.

*****

Mediated money remains to this day the greatest monetary enigma, even though its evolution spans more than four millennia. The previously presented development of mediated money suggests that the essential reason for its enigmatic nature lies in the in-book evidence of transactions, as well as the diversity of its payment instruments and devices (from clay tablets to smartphones).

The brilliantly simple Babylonian formula – not shift money from purse-to-purse, but from account-to-account – functions effectively with licensed, credible and liquid intermediaries in a stable environment of trusted money issuers. When those conditions are disrupted, all clearing (internal, bilateral, multilateral, etc.) is reduced or eliminated. The logic of clearing is based on the symmetry of payments, expressed as the equation: sent payments = received payments in network. This symmetry of payments also characterizes immediate money, but without intermediaries.

The only money that is created and circulated within a national network, as well as on all other levels of clearing (internal, bilateral, multilateral, etc.), is central bank money, denominated in national currency. Mediated money enables the multiplication of payments in the network through overdrafts and other loans, which is a promise of money, i.e., a promise of payment in the form of bank liabilities – deposits. Seemingly, this is a process of apparent deposit and credit multiplication, which in reality represents a reflection of clearing.

Commercial banks have never accepted claims from other banks for settlement. The only settlement asset accepted by all banks has always been central bank money – immediate (until the mid-19th century) and mediated. That is why central banks are the ultimate providers of interbank settlement assets.

The central bank is the supermediator – comprehensive and universal. The unlimited clearing and settlement power of the central bank is the nuclear reactor of mediated money emission. Only thanks to such power was it possible for wholesale depositors to withdraw $140 billion in uninsured deposits from the illiquidity-stricken Silicon Valley Bank in just two days in March 2023.

Based on available historical evidence, the most important places for the development of mediated money, chronologically, are: Babylonia – Phoenicia – Athens – Venice – Amsterdam – London – New York – Basel.

Immediate and mediated money have the same issuer – the central bank, but they differ in access to it – direct access (cash) and indirect access (bank accounts). Access to money determines their key attributes – safety for immediate money and efficiency for mediated money.

The drivers of the eruption in payment transactions and mediated money are clearly wholesale (commercial and financial) transactions, which have completely marginalized retail payments (see Table 1). With the dominant value of transaction accounts, household payments represent  for less than 0.5% of the total transaction value – two hundred times less than corporations and financial institutions! It is evident that there is no problem of liquidity in the household sector, but only in the wholesale sectors – especially financial.

*****

This dynamic analysis focuses on the stock-flow relationship, although most monetary analyses are static. However, the money stock is not the only measure of money supply. The flow of money remains obscured in the clearing-netting process (‘implicit’ transactions) and becomes visible to observers only in the settlement phase (‘explicit’ transactions).

The net settlement model dominated in clearing houses for more than two centuries, until the early 1990s. At that time, central bankers concluded that explicit transactions with explicit money at the account, i.e., gross settlement, were more desirable. Since then, the stigma of implicitness has followed the net settlement model, although it has continued to survive—even in the shadow of RTGS—as the most efficient way to reduce settlement risk.

Despite the dominance of RTGS systems in wholesale payments, the practice of gross settlement has not been able to eliminate the linking of incoming and outgoing transactions, nor the cautious use of reserves. All available evidence suggests that RTGS systems are not superior, but rather inherently inferior to net systems.

On the other hand, gross settlement, in a transaction-by-transaction model, has intensified competitive behavior and blurred the conditionality of own interest by common interest. Since the early 1990s, the payment game has become increasingly noncooperative. RTGS causes a kind of dictatorship of incoming payments, i.e., de-networking of interbank payments. Thus, only reserves at the bank account are seen as own liquidity, while incoming payments are treated as free funds provided by other participants. At the same time, there is a separation between our incoming payment and their outgoing payment, even though this is a single transaction between the payer bank and payee bank.

Elucidation of the gross settlement system: Account A => outgoing payment / incoming payment => Account B => outgoing payment / incoming payment => Account A. Opposite to this is the net settlement system: Account A => outgoing payment => SETTLEMENT <= outgoing payment <= Account B.   

A high level of reserves in the banking sector does not imply timely payments. On the contrary, in stress scenarios, bankers instinctively engage in hoarding of liquidity, driven by precautionary and speculative motives. Yet, the most important generator of crisis is fear.
The pattern of a liquidity crisis: Distress => Uncertainty => Untrust => FEAR => Hoarding => Panic => Sale => Freezing. Once hoarding of reserves begins, the connections between incoming and outgoing payments, which constitute the essence of mediated money, i.e., payments in network—rapidly break down. In the panic phase, comprehensive de-networking threatens the liquidity of all banks, even though they attempt to isolate themselves. However, bank illiquidity is probably the only contagion that is worsened by isolation.

A potential solution could be Real-Time Net Settlement (RTNS). The proposed RTNS model can be succinctly defined as a real-time net settlement payment system with accelerated sequential multilateral netting/clearing in predefined cycles (from a minimum of two minutes to a maximum of fifteen minutes) on a final offsetting basis, along with the use of prefunded balances and intraday credits.

*****

Of course, immediate settlement does not transform mediated money into immediate money. It still remains in the possession of intermediaries. A solution to this structural problem of mediated money could be found in balance sheet reconstruction.

The central bank balance sheet before September 2007 was of interest only to its accountants. After that, it became the focal point for policymakers. As Garreth observed: ‘The central bank’s balance sheet plays a critical role in the functioning of the economy.’ This made it possible to rename the power of mediated money as the ‘power of the central bank balance sheet’ (Orphanides),
– even though it is just a sheet. Consequently, the central bank balance sheet has become a countercyclical policy tool.

Balance sheet reconstruction was proposed precisely to mitigate too big changes in bank balance sheets and frequent cyclical oscillations. The very name indicates changes in the composition of the balance sheets of both the central bank (as issuer and supermediator of central bank money) and commercial banks (as the most important payment service providers).

A possible solution is relatively simple. Current accounts, i.e., demand deposits, would be transferred from the commercial bank liabilities to central bank liabilities. In return, banks would receive compensatory loans, proportional to the transaction balances of their clients. These loans would be permanent and could be calculated based on the account balances at the end of the business day or week. All other balance sheet items would remain unchanged after this reconstruction.

Consequently, the size of the central bank balance sheet would increase by the amount of assumed current accounts (on the liabilities side) and the corresponding compensatory loans (on the asset side). However, this would represent zero monetary expansion, since the loans would be entirely financed from current account funds.

Banks would still maintain current accounts for their clients, but solely as PSPs—without losing revenue from payment services and without any disintermediation. They would retain the same amount of reserves for transactions involving savings, time, and wholesale deposits, as well as all other payments. However, they would no longer be able to directly use the current accounts of their clients. It is understood that current accounts, as a central bank liability, would be interest-free.

Transactions using current accounts would be executed in a pure clearing payment system, where clearing = settlement. This payment system would be managed by the central bank as a supermediator, without any additional liquidity or loan interventions. Payments would depend solely on the liquidity of the payer, who could not issue a payment order without full coverage in their current account. Of course, from the same account, the payer could transfer their savings and time deposits or receive loans from the bank. This pure clearing payment system would be connected with all other domestic payment systems.

It is clear that within such a pure clearing payment system, there is no need for the introduction of a separate central bank digital currency (CBDC), because it would be implicitly created by the proposed reconstruction. The only missing digital currency would be a retail CBDC for offline transactions, primarily for the unbanked population.

Additional benefits of the proposed reconstruction are numerous: containment of bank digital run in the future, significantly lower pressure to reallocate deposits from one bank to another, proportionally reduced need for deposit insurance, transfer of credit risk from account holders to the central bank, reducing settlement risk in existing payment systems, stronger interbank competition on the payment services market, improved supervision of banks, increased perception of financial stability and similar.

Changes in the composition of balance sheets – central bank and commercial banks – would represent a kind of defractionalization of the banking system. Such a reconstruction would require clear political support, in favor of the public, but also of the banks.

In summary: stock of mediated money would not be in an intermediary’s possession. The intermediary would be a custodian, who cannot freely use money on current accounts. Certainty and finality of payments would not depend on the liquidity of the intermediary, but only on the payer. And most importantly, the public would gain direct access to mediated central bank money.

Naturally, direct access to mediated central bank money must not be a pronouncement for the abolition of cash, i.e., immediate money. Cash is the anchor which guarantees the stable nominal value of money. Without cash, £10 will not always be £10, €10 will not always stay €10, nor will $10 always remain $10. This is why the severing of the multi-millennium connection between immediate money and mediated money is economically and socially unacceptable.

First published on https://centralbankmoneyresearch.com/

References

Acharya, Viral and Quarda Merrouche (2012), Precautionary hoarding of liquidity and interbank markets: Evidence from the Subprime crisis, Review of Finance, No 17, pp. 107-160.  

Afonso, Gara, Darrell Duffie, Lorenzo Rigon, and Hyun Song Shin (2022), How Abundant Are Reserves? Evidence from the Wholesale Payment System, Staff Reports no. 1040, FRB of New York, November 2022.       

Afonso, Gara, Domenico Giannone, Gabriele La Spada, and John Williams (2024), Scarce, Abundant, or Ample? A Tyme-Varying Model of the Reserve Demand Curve, Staff Reports no. 1019, FRB of New York, Revised April 2024.     

Andreau, Jean (1999), Banking and business in the Roman world, Cambridge (MA): Cambridge University Press.  

Angell, Wayne (1992), Large-Value Payment Systems: What have we learned? Remarks, Board of Governor of the Federal Reserve System.  

Arnon, Arie (2011), Monetary Theory and Policy from Hume and Smith to Wicksell: Money, Credit, and the Economy, New York: Cambridge University Press.

Bagehot, Walter (1896), Lombard Street: A Description of the Money Market, London: Kegan, Paul, Trench & CO. Ltd. 

Bailey, Andrew, Jonathan Bridges, Richard Harrison, Josh Jones, and Aakash Mankodi (2020), The central bank balance sheet as a policy tool: past, present and future, Staff Working Paper, No 899, Bank of England, December 2020.

Bech, Morten, Antoine Martin, and James McAndrews (2012), Settlement liquidity and monetary policy: Lessons from the Financial Crisis, FRBNY Economic Policy Review, FRB of New York, March 2012, pp. 1-25.

Benos, Evangelos, Rodney Garratt, and Peter Zimmerman (2014), The role of counterparty risk in CHAPS following the collapse of Lehman Brothers, International Journal of Central Banking, Vol. 10, No. 4, December 2014.

Bernanke, Ben (2009), The Federal Reserve’s Balance Sheet – An Update, Speech, October 8, 2009, Chairman of Federal Reserve Board.

BIS (1989), Report on netting schemes, Group of experts on payment systems of the central banks of Ten countries, Bank for International Settlements, February 1989.

BIS (1997), Real-time gross settlement systems, Report, Committee on Payment and Settlement Systems, Bank for International Settlements, March 1997.

BIS (2010), Basel III: International framework for liquidity risk measurement, standards and monitoring, Basel Committee on Banking Supervision, Bank for International Settlements, December 2010. 

BIS (2013a), Basel III: The Liquidity Coverage Ration and liquidity risk monitoring tools, Basel Committee on Banking Supervision, Bank for International Settlements, January 2013.

BIS (2013b), Monitoring tools for intraday liquidity management, Basel Committee on Banking Supervision, Bank for International Settlements, April 2013.

BIS (2023), Report on the 2023 banking turmoil, Basel Committee on Banking Supervision, Bank for International Settlements, October 2023.

BoE (1969), The Bank of England note: a short history, Quarterly Bulletin, Q2 (June), Bank of England. 

BoE (2021), New forms of digital money, Discussion Paper, June 2021, Bank of England.

BoE (2024), The BoE’s approach to innovation in money and payment, Discussion Paper, Bank of England, July 2024. 

BoE & PRA (2023), Pillar 2 liquidity, Statement of Policy, Bank of England and Prudential Regulation Authority, December 2023. 

Clapham, John (1970), The Bank of England: A History, 1694-1914, Volume I – II, Cambridge University Press.

Dale, Gareth (2013), ’Marketless Trading in Hammurabi’s Time’: A Re-appraisal, Journal of the Economic and Social History of the Orient, No. 56, pp. 159-188.

Dent, Andrew and Will Dison (2012), The Bank of England’s Real-Time Gross Settlement infrastructure, Quarterly Bulletin, Volume 52, Number 3 (September), Bank of England, pp. 234-243.

Einzig, Paul (1949), Primitive Money in its ethnological, historical and economic aspects, London: Eyre & Spottiswoode.

Finley, Moses (1999), The Ancient Economy, Berkeley: University of California Press.

Fisher, Irving (1936), 100% Money, New York: Adelphi Company.

Fisher, Paul (2018), The future for central bank balance sheets and their potential use as a macroprudential tool, Policy Note, Issue No 38, SUERF, June 2018.

FRB services (2024), Unpacking clearing and settlement, The Federal Reserve – FRB services, 3/13/2024.

Friedman, Milton and Anna Schwartz (1993), A Monetary History of the United States, 1867-1960, Ninth Ed., NBER & Princeton University Press.

Garreth, Rule (2015), Understanding the central bank balance sheet, Centre for Central bank Studies, Bank of England, 2015.

Goodhart, Charles (1988), The Evolution of Central Banks, MIT Press.

Grierson, Philip (1978), The Origins of Money, Research in Economic Anthropology, Volume 1, pp. 1-35  

Harris, William (2006), The Nature of Roman Money, pp. 174-207, In: Harris, William (ed.) (2006), The Monetary Systems of the Greeks and Romans, New York: Oxford University Press. 

Hayek, Friedrich (1937), Monetary nationalism and international stability, London: Lognmans.

Henckel, Timo, Alain Ize, and Arto Kovanen (1999), Central Banking Without Central bank Money, WP/20/254, International Monetary Fund, July 1999.

Hicks, John (1967), Critical Essays in Monetary Theory, Oxford: Clarendon Press.

Jevons, Stanley (1875/1896), Money and the Mechanism of Exchange, New York: D. Appelton and Company. 

Johnson, Kurt, James McAndrews and Kimmo Soramaki (2004), Economizing on liquidity with Deferred Settlement Mechanisms, FRBNY Economic Policy Review, FRB of New York, December 2004, pp. 51-72. 

Kabadjova, Biliana, Anton Badev et al (2023), Intraday liquidity around the world, Working paper No 1089, Bank for International Settlements, April 2023. 

Kahn, Charles, James McAnrews, and William Roberds (1999), Settlement risk under Gross and Net settlement, Working Paper, 99-10a, Federal Reserve Bank of Atlanta, November 1999.

Kantar Public (2022), Study on New Digital Payment Methods, March 2022.

Keynes, John Maynard (1930a), A Treatise on Money – Volume I – The Pure Theory of Money, London: Macmillan and Co. 

Keynes, John Maynard (1930b), A Treatise on Money – Volume II – The Applied  Theory of Money, London: Macmillan and Co. 

Kindleberger, Charles (1998), Economic and financial crises and transformations in sixteenth-century Europe, Essays in international finance: no. 208, Department of Economics, Princeton University.   

Kohn, Meir (1999), Medieval and early modern coinage and its problems, Working paper 99-02, Dartmouth College.

Kroll, John (2006), The Monetary Use of Weighed Bullion in Archaic Greece, p. 12-37, In: Harris, William (ed.) (2006), The Monetary Systems of the Greeks and Romans, New York: Oxford University Press. 

Leigh-Pemberton, Robert (1992), Developments in wholesale payments systems, Quarterly Bulletin, Volume 32, Number 4 (November), Bank of England, pp. 449-457.

Menger, Karl (1976), Principles of Economics, New York University Press.

Mueller, Reinhold (1997), The Venetian Money Market: Banks, Panic and Public Debt, 1200-1500, Baltimore (MD): Johns Hopkins University Press.

Norman, Ben, Rachel Shaw and George Speight (2011), The history of interbank settlement arrangements: exploring central banks’ role in the payment system, Working Paper No. 412, Bank of England, June 2011.   

Orphanides, Athanasios (2021), The power of central bank balance sheets, Discussion Paper, No. 2021-E-10, Institute for monetary and economic studies – IMES & Bank of Japan, August 2021.

Peacock, Mark (2006), The origins of money in Ancient Greece: the political economy of coinage and exchange, Cambridge Journal of Economics, No 30.

PRC (2000), Intraday liquidity management in the evolving payment system: A study of the impact of the euro, CLS Bank, and  CHIPS finality, The Payment Risk Committee, New York, April 2000.

Roberds, William and Francois Velde (2014a), The Descent of Central Banks (1400-1815), Norges Bank 2014 conference, May 2014.

Seaford, Richard (2024), Money and the Early Greek Mind, Cambridge University Press.

Silva, Andre (2016), Strategic complementarity in banks’ funding liquidity choices and financial stability, Working Paper Series, No 19, European System Risk Board, July 2016.

Simmel, Georg (2004), The Philosophy of Money, London: Routledge.

Smith, Adam (1776/1937), An Inquiry into the Nature and Causes of the Wealth of Nations, New York: The Modern Library, 1937.

Spufford, Peter (1988), Money and its use in medieval Europe, Cambridge University Press.

Thornton, Henry (1802), An inquiry into the nature and effects of the paper credit of Great Britain, Reprint, New York, 1965.

Turocy, Theodore and Bernhard von Stengel (2001), Game Theory, CDAM Research Report, 2001-09, LSE, October 2001.

Ugolini, Stefano (2017), The evolution of central banking: theory and history, Palgrave & Macmillan. 

Vuković, Vlastimir (2020a), Abolition of Cash or Loss of Anchor, Paper 3, Central Bank Money Research, April 2020.

Vuković, Vlastimir (2020b), Fractional bank money, Paper 4, Central Bank Money Research, April 2020.

Vuković, Vlastimir (2020c), The illusion of paper money as ’the dominant means of payment’, Discussion Paper, No. 1, Central Bank Money Research, October 2020.

Vuković, Vlastimir (2021), CBDC as a Solution: for billions of unbanked people, Discussion Paper, No. 2, Central Bank Money Research, October 2021.

Vuković, Vlastimir (2022), Hammurabi vs cryptocurrencies: anonymity of payments, Discussion Paper, No. 3, Central Bank Money Research, November 2022.

Vuković, Vlastimir (2023), The perception of people about their money and the standpoint of monetary theory and policy, with case study: why digital euro, Discussion Paper, No. 4, Central Bank Money Research, Deceber 2023.