9.2 — Banks, Credit and Interest

Here is the fact that most people find hardest to believe about the modern economy.

When a bank makes a loan, it does not lend out money that somebody else deposited. It creates new money by typing a number into an account.

This is not a fringe claim. The Bank of England published a paper in 2014 stating it plainly: "whenever a bank makes a loan, it simultaneously creates a matching deposit in the borrower's bank account, thereby creating new money." The textbook story — that banks are intermediaries channelling savers' money to borrowers — is, as a description of the mechanics, wrong.

Roughly 95 percent of the money in a modern economy is bank deposits created this way. Physical currency is the small remainder.

This chapter explains how that came about, why it is not a scandal, and where the danger actually lies.

How banking started

The goldsmith story is the standard illustration and it is approximately right.

Goldsmiths had strong vaults, so people deposited coin with them for safekeeping and received a receipt. The receipts began circulating — it is easier to hand over a receipt than to collect the gold and carry it. So the receipt became money.

The goldsmith then noticed that only a small fraction of depositors ever came for their gold at any one time, since the receipts were circulating instead. So he could issue more receipts as loans than he held gold, and collect interest on them.

That is fractional reserve banking, and it is the basis of the entire system.

The real history is messier. Deposit banking existed in Mesopotamian temples, in Greek temples, and in Roman argentarii. Medieval Italian merchant banks — the Bardi, Peruzzi and then the Medici — combined deposit-taking, currency exchange, transfers between branches and lending. The Bank of Amsterdam from 1609 and the Bank of England from 1694 are the institutional ancestors of central banking.

What banks actually do

Three functions, and mixing them up is where confusion comes from.

Maturity transformation. Depositors want their money available immediately. Borrowers want to repay over years. A bank stands between: it borrows short and lends long. This is genuinely useful and it is inherently fragile, because if all the depositors want their money at once the bank cannot produce it — not because it is dishonest but because the money is lent out for twenty years.

Risk assessment. Deciding who is creditworthy is a real service that requires information and expertise.

Money creation, as above.

Money creation, step by step

The bank does not need a prior deposit.

A bank grants you a loan of ten lakh rupees. It writes an asset on its books — your promise to repay — and a liability — a deposit of ten lakh in your account. Both sides of its balance sheet grow. Nobody's account was reduced. The money supply is now ten lakh larger.

You spend it. The money moves to the seller's account, possibly at another bank. The banks settle the difference between themselves at the end of the day, using reserves held at the central bank — and only the net difference needs settling, which is far smaller than the gross flow.

When you repay the loan, that money is destroyed — the deposit and the debt cancel each other. The money supply shrinks.

So the money supply expands when net lending grows and contracts when it shrinks, and this is why credit cycles drive economic cycles so directly (Chapter 9.9).

What actually limits bank lending is not the deposits it holds. It is capital requirements — the bank must fund a portion of its assets with its own shareholders' money so it can absorb losses — the availability of creditworthy borrowers, and the central bank's interest rate, which sets the cost of the reserves the bank needs for settlement (Chapter 9.7).

The medieval instruments

Merchants faced a specific set of problems and each solution is still in use.

Moving value without moving metal. Carrying gold across Europe is slow and it attracts bandits. The bill of exchange solved it: a merchant in Florence pays a banker in florins and receives a written instruction to the banker's correspondent in Bruges to pay the bearer in local currency on a stated date. No metal moves.

And it evaded the usury prohibition (Chapter 9.1), which is why it is important. The bill was denominated in a different currency at each end, and the exchange rate was set to include what was in fact interest. Canon lawyers accepted this because exchange risk is real risk and charging for risk was permitted. The whole of European commercial finance was built in the space created by that argument.

Bills also became negotiable — transferable to a third party by endorsement — which means they circulate, which means they are money.

Double-entry bookkeeping, described systematically by Luca Pacioli in 1494 and in use in Italy long before. Every transaction is entered twice, as a debit and a credit, so the books must balance. It makes error detectable, makes fraud harder, and — the significant part — it makes it possible to calculate profit, which means a business can be evaluated as a continuing enterprise rather than as a series of ventures.

Marine insurance, developed in Genoa and Venice, spreading the risk of a voyage across many underwriters. Lloyd's of London began in a coffee house in the 1680s where shipowners and underwriters met.

And the Knights Templar ran an early international transfer network for pilgrims (Chapter 7.5).

Public debt, and why it decided who won

This is the most consequential financial innovation in this volume and it deserves the space.

The problem. A war costs more than a year's tax revenue. A ruler who cannot borrow must either tax at ruinous rates immediately, debase the currency, or lose.

The medieval answer was to borrow personally from bankers, at very high rates, and default. Edward III of England defaulted in the 1340s and destroyed the Bardi and Peruzzi banks of Florence. The Spanish crown, despite American silver, defaulted repeatedly — 1557, 1560, 1575, 1596 and more (Chapter 8.5). A borrower who cannot be sued pays a high rate or is refused.

The solution was to make the debt the state's rather than the ruler's, and to make repayment credible.

The Dutch Republic did it first. The provinces borrowed against specified tax revenues, with the debt authorised and guaranteed by the assemblies of the same merchants who held the bonds. The creditors and the taxing authority were the same people, so default meant defaulting on yourself. Dutch interest rates fell to around 4 percent when other states paid double or more.

England adopted it after 1688 (Chapter 10.1). Parliament, not the king, guaranteed the debt; the Bank of England was founded in 1694 specifically to manage it; and the tax revenues pledged to service it were voted by the same body that held much of the debt.

The result is startling. Britain, with roughly a third of France's population, consistently outspent France militarily throughout the eighteenth century, because it could borrow more cheaply and therefore more. Its national debt reached extraordinary multiples of annual revenue and it never defaulted, which is precisely why it could keep borrowing.

This is a major part of the answer to Chapter 6.15's question about India, and to Chapter 8.10's about the gunpowder empires. A state that can credibly promise to repay can mobilise the future. One that cannot is limited to this year's harvest.

The mechanism generalises. Credible commitment is the scarce resource — the ability to make a promise that others believe because breaking it would be costly to the promiser. Chapter 13.1 makes this the foundation of what a functioning state is.

Bank runs

The fragility is structural, not a failure of management.

A bank is solvent — its assets exceed its liabilities — and illiquid, because its assets are twenty-year loans and its liabilities are payable today. If enough depositors demand cash at once, the bank must sell assets at fire-sale prices, which makes it actually insolvent.

And the run is rational for each individual. If you think others will withdraw, you should withdraw first. Everyone reasoning correctly produces a collapse that nobody wanted. This is a self-fulfilling equilibrium, and the Diamond-Dybvig model of 1983 is the formal statement of it.

Three institutions were built to stop it.

The lender of last resort. Walter Bagehot's rule, from 1873: in a panic, the central bank should lend freely, at a penalty rate, against good collateral. Freely, so the panic stops; at a penalty, so banks do not rely on it; against good collateral, so the central bank is rescuing the illiquid rather than the insolvent.

Deposit insurance. Introduced in the United States in 1933 after thousands of bank failures. If your deposit is guaranteed, you have no reason to run. India's scheme currently guarantees deposits up to five lakh rupees per depositor per bank.

And capital and liquidity regulation — the Basel framework, requiring banks to hold minimum capital against risk-weighted assets.

The problem this creates is moral hazard. A bank whose depositors are insured and which expects rescue has an incentive to take more risk, because the upside is private and the downside is public. Chapter 12.8 shows exactly that happening in 2008.

Islamic finance

Worth covering because it is a live sector and it illustrates the underlying question.

The prohibition on riba rules out a guaranteed return on a loan. The solutions used are risk-sharing structures: mudarabah, where the financier provides capital and shares in profit and loss; musharakah, a partnership; and murabaha, where the bank buys an asset and resells it to the customer at a marked-up price payable in instalments.

The honest assessment. Some instruments are genuine risk-sharing and structurally different from a loan. Others are economically equivalent to interest with a different legal form, and this criticism is made most forcefully by Islamic scholars themselves. The sector is real — it holds trillions of dollars in assets — and the debate about substance versus form is unresolved within it.

Where this shows up in your life

Your bank deposit is not your money sitting in a vault. It is a debt the bank owes you, and you are an unsecured creditor of a leveraged institution, protected by deposit insurance up to a limit.

Your home loan created money. The deposit that appeared in the seller's account did not come out of anyone else's.

Interest rates set by the central bank change what you pay on that loan, and Chapter 9.7 explains how a committee moves a number and the whole economy responds.

And the distinction from Chapter 9.1 matters for policy. Productive lending — to a business that will generate a return — and distress lending — to someone who cannot eat — are different activities with different consequences, and the same word covers both. Indian farm indebtedness, where borrowing at high rates from informal lenders against land is a documented factor in farmer suicides, is the distress case, and microfinance's mixed record is an argument about which of the two it actually delivers.

What the next page covers

A voyage to Asia could return several times its cost or lose everything. Chapter 9.3 covers the invention that made such risks bearable — the joint-stock company with transferable shares and limited liability, the world's first stock exchange in Amsterdam, and the first two great speculative bubbles, which happened within a few years of each other in London and Paris and which established every feature of every bubble since.