Sometime in the 1640s, a London goldsmith noticed something.
He had taken in deposits of gold from merchants who no longer trusted the Royal Mint — King Charles I had recently helped himself to £130,000 worth of merchants' gold stored at the Tower of London and called it a "loan," which was the kind of loan that is really just theft with a receipt. The merchants needed somewhere to put their gold, goldsmiths had thick vault doors, and so an informal deposit-taking business had emerged.
The goldsmith issued paper receipts for the gold. The merchants could present the receipts and collect their gold at any time. This part was straightforward.
What the goldsmith noticed was this: they rarely did. Gold came in. Gold went out. But the vault was never empty, and the full quantity of deposited gold was never claimed simultaneously. There was always a residual — gold sitting in the vault, backing receipts that were circulating in the market as though they were money. Because they essentially were. A receipt from a reputable goldsmith was trusted. People were trading them without going through the step of actually retrieving the gold.
The goldsmith's thought, at this point, was either a stroke of genius or the opening move in a four-century-long confidence game, depending on your perspective: What if I issued more receipts than I have gold?
As long as not everyone came at once, no one would ever know. And the extra receipts — claims on gold that did not exist — could be lent out at interest.
This is fractional reserve banking. It is the mechanism by which nearly all money in the modern economy is created. And it was, in its English incarnation, almost certainly an accident.
But First: Babylon
The goldsmith story is the clearest version of the tale, but it is not the beginning. The beginning is approximately four thousand years earlier, in the river valleys of Mesopotamia.
The great temples of ancient Babylon — dedicated to Marduk, Ishtar, Shamash — were not only religious institutions. They were the warehouses, administrators, and financiers of their civilizations. A farmer who produced more grain than he could immediately use brought it to the temple, where it was stored in massive granaries. The temple issued a clay tablet recording the deposit: so many kur of barley, deposited by this farmer, retrievable on presentation of this tablet.
These clay tablets were the world's first deposit receipts. And like the goldsmith receipts four millennia later, they circulated. A farmer who owed a debt to a craftsman could transfer his tablet rather than carry sacks of grain. The craftsman could redeem it, or transfer it onward, or use it to pay temple fees. The clay tablet had become, functionally, money — a bearer claim on a stored commodity.
The temples then did what institutions with large stores of wealth tend to do: they lent it out. A farmer who needed grain for planting could borrow from the temple, to be repaid after harvest with interest. The Code of Hammurabi, inscribed on a black diorite stele around 1754 BCE and now in the Louvre, dedicates extensive provisions to these transactions. Interest on grain loans: thirty-three and a third percent annually. On silver: twenty percent. Maximum rates, not minimums — the state was already trying to protect borrowers from exploitation.
This is proto-banking in almost every functional sense: deposits taken, receipts issued, loans made, interest charged, regulations applied. What the Babylonian temples almost certainly did not do — the evidence is fragmentary — was issue more receipts than they had grain. Their lending was from genuine surplus, not from fabricated claims. The crucial final step had not been taken.
China Invents Paper Money, For the Wrong Reason
The connection between paper receipts and paper money is not just logical — it happened, in the most direct possible way, in Song Dynasty China around 960 CE. And it happened partly because China had a practical problem that would strike modern readers as almost comically inconvenient.
China had long used bronze coins as currency, but in Sichuan province, bronze was scarce, and the government had authorized the use of iron coins instead. Iron coins were worth far less per unit of weight than bronze. To carry enough iron coins for significant transactions, a merchant needed not a purse but a wagon. Historians have calculated that a modest commercial payment might require several hundred pounds of iron coin.
In response, private money houses in Sichuan began offering deposit services. A merchant deposited his iron coins, received a paper certificate (jiaozi, 交子) indicating the amount, and could use the certificate in trade or redeem it for coins on demand. The certificates were lighter than the coins by approximately the entire weight of the coins, which was the point.
By the early eleventh century, the government recognized what was happening, took over issuance, standardized the certificates, and created the world's first official paper money. By 1023, a government bureau was printing jiaozi with controlled circulation. By the time the Mongol Yuan Dynasty consolidated power in the thirteenth century, paper money had become the primary currency across China, and the Mongols issued their own — the chao (鈔) — which Marco Polo encountered on his travels and described with astonishment in his journals.
"All his Majesty's armies are paid with this currency, and all those who have goods or who sell articles of food or clothing get paid in it." — Marco Polo, Il Milione, ca. 1300
What Polo witnessed was a state-operated paper currency that worked because the government backed it — and that catastrophically inflated when the government issued more than trade required without adequate reserves. The Yuan Dynasty's monetary problems were a preview of every sovereign currency crisis that followed: paper money is only as good as the restraint of whoever prints it.
The Chinese experience with jiaozi was parallel to but independent of the goldsmith story. Both followed the same logic — paper claim on something real circulates as money — and both discovered the same temptation: issue more paper than you have real thing.
The Italians Build a System Without the Physical Stuff
Between the Babylonian temples and the London goldsmiths, the most sophisticated financial innovators were the Italian banking houses of the thirteenth through fifteenth centuries — the Bardi, the Peruzzi, and above all the Medici.
The Italians did not invent fractional reserve banking, but they invented something arguably more important: credit instruments that allowed money to exist as a relationship rather than as a physical thing at all.
The lettera di cambio — the bill of exchange — allowed a merchant in Florence to deposit funds with a Medici agent, receive a written instrument, travel to London, and present that instrument to a Medici agent there in exchange for English sterling. No coins crossed the Alps. No gold was physically moved. A piece of paper, backed by the reputation and network of the Medici house, substituted for the physical transfer of wealth across a thousand miles of dangerous road.
This is credit in its purest form: value existing as a promise, guaranteed by an institution rather than by physical reality. The Medici bank was the most sophisticated financial institution in the world in the fifteenth century, with branches across Europe and relationships with popes and kings. It also failed catastrophically in 1494, partly due to bad loans to Edward IV of England that were never repaid — a reminder that the difference between a sophisticated credit instrument and an unsecured loss depends heavily on whether the borrower pays.
The Italian banking houses also developed the deposit account in something closer to its modern form — money held at a bank, earning a return, accessible by written instruction. They did not call the return "interest," because usury was still formally prohibited. They called it discrezione — discretionary compensation — which is one of the more creative theological work-arounds in financial history.
Sweden Has a Problem With Copper
The direct precursor to the English goldsmith story, and the site of the first European paper currency and the first documented European bank run, is an unlikely location: Sweden.
Sweden in the mid-seventeenth century had vast copper deposits and a government that had decided copper would be the basis of its currency. This was fine in principle and catastrophic in practice. Copper is worth far less per weight than silver, so a Swedish copper daler had to contain enough copper to equal the purchasing power of a silver coin. A ten-daler copper plate — and these were literal plates, not coins — weighed approximately forty-three pounds.
Carrying enough money to buy a horse required, itself, something like a horse to carry it.
Johan Palmstruch founded Stockholms Banco in 1657 and in 1661 introduced the Kreditivsedlar — credit notes — as a solution. Deposit copper plates, receive paper. The paper was lighter. Sweden's copper problem was manageable.
Palmstruch then made the leap: he issued more notes than he had copper. The loans he made with phantom copper notes generated interest income. For a few years, this worked. Then redemption demands exceeded reserves, the bank could not pay, and in 1664 Stockholms Banco collapsed. Palmstruch was convicted of fraud and sentenced to death. (The sentence was commuted; he died in prison in 1671.)
The Swedish parliament, rather than concluding that paper money backed by fractional reserves was inherently dangerous, concluded that Stockholms Banco had been managed by the wrong person. In 1668 they founded the Riksbank — the Bank of Sweden — which is the oldest continuously operating central bank in the world. It still exists. The Riksbank's founding charter specifically prohibited it from issuing more notes than it held in reserves.
It did not hold to that prohibition for very long.
London, the 1640s, and the Goldsmith's Realization
The English goldsmith story happens in a specific political context that matters.
In 1640, King Charles I was at war with Scotland and desperate for money. Gold and silver belonging to London merchants — approximately £130,000 — was stored in the Royal Mint at the Tower of London for safekeeping. Charles seized it. He called it a forced loan and promised repayment. The merchants called it theft and never fully forgot it.
The lesson was obvious: do not store your valuables where the king can reach them. London's goldsmiths, who kept their own gold and silver stock in purpose-built vaults, began offering storage services. The merchants obliged, and by the 1640s the goldsmiths had become de facto bankers.
The receipt system worked. The receipts circulated. The crucial observation was made — by which specific goldsmith, and in which year, history does not record, because the men who made this discovery had every incentive not to document it publicly.
The observation: in a vault holding deposits from many different merchants, the daily net flow of gold in and out is small relative to the total held. Individual depositors withdraw and deposit, but the aggregate barely moves. There is always a substantial residual — "idle" gold, backed by outstanding receipts, that simply sits.
The insight: issue additional receipts against that idle gold. Lend those receipts out at interest. Keep enough gold on hand to meet expected redemption demand, plus a margin for safety. The fraction of total issued receipts held as actual gold in reserve: perhaps twenty to thirty percent. Maybe less.
This is the moment. The goldsmith has not created a commodity, performed a service, or produced anything. He has created money — claims on gold that does not exist — and lent it at interest. The modern banking system descends from this realization in essentially unbroken lineage.
The most prominent goldsmiths-turned-bankers of the period — Edward Backwell, Francis Child, Richard Hoare — became enormously wealthy. Their "running cash notes," as the receipts came to be called, were effectively the currency of London commerce. The Bank of England, founded in 1694, institutionalized this practice at national scale: a group of merchants lent the government £1.2 million and received a charter to issue banknotes of the same value — backed not entirely by gold but by the government's promise to repay.
The nation-state had adopted the goldsmith's trick.
The Bank Run: What Happens When Everyone Comes at Once
The existential risk of fractional reserve banking is not obscure. It was apparent from the moment the first goldsmith issued his first excess receipt.
If every depositor presents their receipt for gold simultaneously, the goldsmith cannot pay. He has gold for twenty cents of every dollar his receipts represent. He defaults. This is a bank run, and it is the recurring catastrophe of financial history — from the collapse of Stockholms Banco in 1664 to the failure of Northern Rock in 2007, the first British bank run in 140 years, in which depositors queued around the block to withdraw savings from a bank that had run out of money.
The stability of fractional reserve banking depends entirely on confidence: the reasonable expectation, held by all depositors simultaneously, that they could withdraw but that most of them won't. When that expectation breaks — when enough depositors decide to be first in line rather than trust the system — the system collapses. The gold in the vault was never there. The money was a collective agreement to act as though it was.
Central banking, in its modern form, exists largely to manage this vulnerability. The "lender of last resort" doctrine — developed by Walter Bagehot in Lombard Street (1873) — holds that a central bank should lend freely to solvent banks facing a liquidity crisis, at a penalty rate, against good collateral. The theory is that bank runs are self-fulfilling panics that can be arrested by the credible promise of unlimited central bank support. The practice, as 2008 demonstrated, involves decisions about which institutions are "solvent" that are as much political as financial.
How Much Money Is Real?
The modern answer to this question is stranger than most people realize.
In the United Kingdom, approximately 97% of the money in circulation exists not as printed notes or minted coins but as numbers in bank accounts — digital entries created when banks make loans. In the United States, the figure is similar. When a bank makes a mortgage loan, it does not transfer gold from a vault. It creates a deposit account in the borrower's name and credits it with the loan amount. The money did not exist before the loan. It was created by the act of lending.
The Bank of England confirmed this in a 2014 bulletin with unusual directness: "Rather than banks receiving deposits when households save and then lending them out, bank lending creates deposits... Whenever a bank makes a loan, it simultaneously creates a matching deposit in the borrower's bank account, thereby creating new money."
The Babylonian temple issued clay tablets for grain that existed. The London goldsmith issued paper receipts for gold that mostly existed. The modern commercial bank issues loans backed by reserves that amount to a small fraction of total deposits — and the reserves themselves are, in most modern systems, entries in a central bank's ledger rather than physical anything.
Each generation of this story removed one more link between the claim and the underlying reality. The clay tablet represented actual grain. The goldsmith receipt represented mostly actual gold. The modern bank deposit represents a promise backed by a fraction of a fraction, guaranteed ultimately by the state's willingness to stand behind it.
The Accident and What It Built
Whether you consider fractional reserve banking an ingenious solution to the scarcity of money or an institutionalized confidence scheme is, genuinely, a matter of perspective. Both descriptions are accurate.
It solved a real problem. Commercial economies require more money than physical commodity supplies can provide. Gold is finite; commerce is not. The ability to expand the money supply through credit — to create purchasing power against future productive activity — allowed economies to grow beyond the constraint of whatever gold happened to be in circulation. The industrial revolution was financed on credit that did not exist as physical gold. The modern global economy, which produces and trades trillions of dollars in goods and services annually, runs on money that is mostly numbers created by banks making loans.
It also created a system with structural instability at its foundation, moral hazard baked in at every level, and a persistent tendency to concentrate the benefits of money creation among those who have the most access to credit while distributing the risks of system failure to everyone. The goldsmith who discovered he could issue more receipts than he had gold made himself rich. The merchants who held receipts for gold that wasn't there bore the risk.
The Bank of England institutionalized that arrangement in 1694. Every central bank since has inherited it.
The goldsmith probably did not intend to invent the mechanism that would underpin global capitalism for the next four centuries. He probably just noticed that the vault had more gold in it than he needed to keep on hand, and that there was money to be made in the gap.
He was right. There was. There still is. That gap — between the claims a financial system issues and the assets it actually holds — is one of the most productive and one of the most dangerous spaces in all of economics.
It was not designed. It was discovered. And we have been building on it ever since.
Sources
- Code of Hammurabi (ca. 1754 BCE) — deposit, grain loan, and interest rate provisions; Avalon Project, Yale Law School; original stele in the Louvre, Paris
- Marco Polo, Il Milione (ca. 1298–1300; Rustichello da Pisa, ed.) — description of Yuan Dynasty paper currency (chao)
- King Charles I seizure of merchants' gold (1640) — approximately £130,000 held at the Royal Mint, Tower of London; see J.K. Horsefield, British Monetary Experiments, 1650–1710 (1960)
- Johan Palmstruch and Stockholms Banco (founded 1657; collapsed 1664) — see Lars Jonung, The Economics of Private Money (1989)
- Sveriges Riksbank founded 1668 — oldest continuously operating central bank; riksbank.se
- Bank of England founded 1694 — charter issued to lend £1.2 million to William III in exchange for note-issuing authority; J.H. Clapham, The Bank of England: A History (Cambridge University Press, 1944)
- Bank of England, "Money creation in the modern economy", Quarterly Bulletin 2014 Q1, by Michael McLeay, Amar Radia, and Ryland Thomas — bankofengland.co.uk; confirms that bank lending creates deposits rather than lending pre-existing deposits
- Walter Bagehot, Lombard Street: A Description of the Money Market (1873) — lender of last resort doctrine
- 97% of UK money as bank deposits: Bank of England, "Money in the modern economy: an introduction", Quarterly Bulletin 2014 Q1