Author's Note
This study began with a question I could not answer.
I was looking at an exchange rate screen. One US dollar cost roughly ninety-five Indian rupees. I understood, in a vague way, that this was "the market". But then a second thought arrived and would not leave: who decided that? Nobody in Washington rings a bell each morning. Nobody in Mumbai signs a form. And yet every day, hundreds of millions of transactions across a hundred and eighty countries settle against a number that no single person sets and no single government controls — a number that quietly determines whether an Indian family can afford a laptop, whether an Indian airline can buy fuel, and whether the country's oil bill goes up by nine billion dollars this year.
The more I pulled at that thread, the more it turned out to be attached to everything else. Why is oil priced in dollars when America is not the largest importer of it? Why did the Soviet Union, a superpower with nuclear weapons and a continent of resources, never manage to make the ruble matter? Why can Europe — richer than the United States by population, with a single market and a single currency — not displace the dollar? Why did the world's central banks buy more gold in the last four years than in the previous fifteen? And what, if anything, can a country like India actually do about any of it?
Five months later, this document is what I have. I wrote it for the version of myself who sat down in March and did not know what a eurodollar was. So there is a rule running through every page: if a sentence cannot be explained in ordinary words, it does not belong in this report. There is no equation here that a school student cannot follow. Every technical term is defined at the moment it first appears and again in the glossary at the back. Where I use jargon, it is because you will meet that jargon in the newspapers, and you deserve to know what it means when you do.
Two warnings before you begin.
First, this is not a cheerleading document and it is not a doom document. The internet is full of confident people announcing that the dollar collapses next year, and equally confident people announcing that nothing will ever change. Both are selling something. The honest answer is that the dollar's position is eroding slowly at the edges while remaining unchallenged at the centre, and I have tried to show you exactly which edges and exactly which centre, with the numbers attached, so that you can form your own judgement.
Second, on India: I have deliberately resisted the temptation to write a flattering story. India in 2026 is the fastest-growing large economy on earth and its position is genuinely stronger than at any point in its history. It is also a country that runs a persistent trade deficit, whose currency is not fully convertible, whose bond market foreigners can barely access, and whose financial depth is a fraction of what a reserve-currency issuer requires. Both of those things are true simultaneously. A study that reports only the first half is a press release, not research.
Read it in order if you can. Each part is built on the one before it, and the last part — the one about whether the dollar's monopoly can be broken — makes very little sense unless you have first understood what money actually is.
Lovepreet Singh August 2026
How to Read This Report
This report is designed so that you can read it cover to cover, or dip into a single chapter and still follow it. Five recurring devices appear throughout. Learn them once and the rest of the document navigates itself.
Whenever a genuinely difficult idea appears — the Triffin Dilemma, the Impossible Trinity, Balassa–Samuelson — a yellow box like this one restates it in the plainest English available, usually with a household example. If you read nothing else on a page, read these.
Blue boxes carry the hard data: the actual figure, the actual date, the actual source. Numbers in this report are current to August 2026. Where a figure is contested or measured differently by different bodies, I say so rather than picking the flattering one.
Grey boxes tell a real story — a country, a year, a decision and its consequence. Theory is easy to argue with; 1991 is not.
Purple boxes take a claim that circulates widely — usually on the internet, sometimes in newspapers — and test it against the evidence. Several of the most popular claims about de-dollarisation do not survive contact with the data.
Red boxes are where I write the thing that is inconvenient for whichever argument I have just made. There is at least one in every major section. They are the most important boxes in the report.
A note on numbers
Money is measured in three different ways in this document and confusing them is the single most common error in public debate:
Nominal, at market exchange rates
Convert everything into dollars at today's rate. Best for measuring international purchasing power — buying oil, aircraft, chips, or paying foreign debt.
PPP — Purchasing Power Parity
Adjust for the fact that a haircut costs $2 in Patna and $40 in Boston. Best for measuring the real size of an economy and living standards. Explained fully in Chapter 24.
Real, or inflation-adjusted
Strip out price changes over time so that 1970 rupees and 2026 rupees can be compared.
Unless stated otherwise, "$" means US dollars, "bn" means billion (a thousand million), "tn" means trillion (a thousand billion), and "₹" means Indian rupees. One crore is ten million; one lakh is one hundred thousand.
1The World Before Money
Every textbook opens with barter: the man with a cow who wants shoes, the cobbler who wants milk. It is a beautiful story. Anthropologists have looked for it for a hundred and fifty years and never found a society that worked that way.
Start with the problem money solves, because everything else follows from it.
Imagine a village with no money. You are a farmer with a sack of wheat. You want a pair of sandals. You walk to the sandal-maker. He says: I have sandals, but I do not need wheat this week; I need a goat. So now you must find someone who has a goat and wants wheat, trade with him, then walk back to the sandal-maker. If the goat-owner wants fish, not wheat, you need a fourth person. Every additional person multiplies the difficulty.
Economists call this the double coincidence of wants: for a trade to happen directly, I must want what you have and you must want what I have, at the same moment, in the same place, in the same quantities. In a village of ten people that is inconvenient. In a city of a million it is impossible.
Barter needs a perfect match. Money removes the need for a match. Instead of finding someone who wants exactly what you have, you sell what you have to anyone, receive a token that everybody accepts, and use that token to buy what you want from anyone else. Money is not a thing. Money is a permission slip that splits one trade into two halves that no longer have to happen at the same time.
Three other problems barter cannot solve
The matching problem is the famous one, but three more are just as fatal.
Indivisibility. A cow is worth roughly forty chickens. What if you want three chickens? You cannot cut off three-fortieths of a live cow and keep the rest useful. Most valuable goods cannot be divided without destroying their value.
Storage. A fisherman's wealth rots in two days. He cannot save. Without a way to store value, nobody can accumulate, and without accumulation nobody can invest, and without investment there are no roads, no ships and no factories.
Accounting. In a barter world there is no common ruler. Is a plough worth more than a year of a labourer's work? You would need to memorise the exchange rate between every pair of goods. With 100 goods that is 4,950 separate prices. With money it is 100 prices.
Table 1.1 — Number of exchange rates you must know
| Goods in the economy | Prices needed under barter | Prices needed with money |
|---|---|---|
| 10 | 45 | 10 |
| 100 | 4,950 | 100 |
| 1,000 | 499,500 | 1,000 |
| 10,000 | 49,995,000 | 10,000 |
Barter prices = n(n−1)/2. This alone makes a complex economy arithmetically impossible without money.
What actually happened instead
Here is where the textbook story breaks down. The anthropologist Caroline Humphrey put it bluntly in 1985: "No example of a barter economy, pure and simple, has ever been described, let alone the emergence from it of money." What researchers find in societies without money is not barter. It is credit and obligation.
In a small community, everybody knows everybody. If I have surplus grain and you are hungry, I give you grain. Not as charity and not as a trade — as a debt held in memory. Next season, when my roof fails, you fix it. The village runs on a mental ledger of who owes what to whom. It works beautifully at the scale of a few hundred people who cannot escape each other.
It collapses the moment you meet a stranger. A stranger has no reputation with you, no fear of your community's disapproval, and may never return. That is when you need a settlement that is final — something handed over that closes the obligation on the spot. Barter, when it appears in history, is what happens between strangers and enemies. Credit is what happens between neighbours.
Money was not invented to replace barter between neighbours. It was invented to allow transactions between strangers, at scale, without trust. Every later development in this report — coins, banknotes, SWIFT, the dollar — is a solution to the same problem: how do two parties who have no reason to trust each other complete an exchange? Understanding this makes the dollar's dominance intelligible. The dollar is not popular because Americans are liked. It is dominant because it is the most widely accepted answer to "how do we settle between strangers?"
Commodity money: the first answer
Before coins, dozens of societies converged independently on the same solution: pick a commodity that everyone values and use it as the middle step. The choices are revealing.
Table 1.2 — Commodity monies and why they worked (or failed)
| Money | Where and when | Why it worked | Why it failed |
|---|---|---|---|
| Cattle | Indo-European, Vedic India, Africa | Universally valued; reproduces itself | Cannot divide; needs feeding; dies |
| Barley & grain | Mesopotamia, 3000 BCE | Divisible; everyone eats | Rots; bulky; quality varies |
| Cowrie shells | China, India, West Africa, 1500 BCE–1900 CE | Durable, small, hard to fake, pretty | Europeans shipped them in by the boatload and destroyed the value |
| Salt | Rome, Sahara trade | Essential, divisible, storable | Dissolves; heavy relative to value |
| Silver & gold | Everywhere, from ~3000 BCE | Rare, durable, divisible, uniform, dense in value | Supply is an accident of geology, not of economic need |
| Cigarettes | POW camps and prisons, 1940s– today | Divisible, portable, demanded by users | Smokers consume the money supply |
Notice what the successful ones share: durable, portable, divisible, uniform, difficult to counterfeit, and scarce. Gold and silver won not because of mysticism but because they score well on all six at once. They also do something subtler: they are useless. Gold does not rot and almost nobody needs it for survival, which means using it as money does not deprive anyone of food or shelter. A money made of wheat is a money that starves people during a famine.
For nearly three thousand years cowrie shells were real money across Asia and Africa. They were still legal currency in parts of India under the East India Company and in West Africa into the 1890s. Then European traders realised something. Cowries came from the Maldives, and steamships could carry them by the tonne. Between 1850 and 1890, roughly 35,000 tonnes of shells were imported into West Africa. The value of the currency collapsed, wiping out the savings of an entire region.
This is the first documented case of what we now call monetary imperialism: one power destroying another's money supply simply by flooding it. Keep this in mind for Chapter 16. The mechanism changes; the logic does not.
2Debt Came First: Mesopotamia and the Birth of Accounting
The oldest writing humanity possesses is not poetry, law or scripture. It is a receipt. Money began as a number written down, not as a coin held in the hand.
Around 3300 BCE in the city of Uruk, in what is now southern Iraq, someone pressed a reed into wet clay to record that a quantity of barley had been received. Roughly ninety per cent of all surviving cuneiform tablets from the earliest period are administrative: inventories, rations, loans, debts. Literature comes eight centuries later. Writing was invented for accounting.
The silver shekel: a unit that nobody held
By around 3000 BCE the temple bureaucracies of Sumer had done something remarkable. They defined a standard unit of account: the shekel, equal to roughly 8.3 grams of silver, and also defined as equal to one gur (about 300 litres) of barley — a month's ration for a worker. Every debt, fine, rent and wage in the city could now be expressed in shekels.
The crucial detail: almost nobody used silver in daily transactions. Silver was in the temple. The peasant who owed three shekels paid in barley, in wool, in labour on the canals. The shekel was a measuring stick, not a payment. This is the single most important fact in the early history of money, and it took economists most of the twentieth century to accept it.
Think of a school with a points system. Good behaviour earns points; points can be spent on privileges. Nobody carries a physical "point". The points exist as numbers in the teacher's register. Sumerian money was exactly this: numbers in a ledger, measured in silver but rarely paid in silver. When you hear that "the dollar is just numbers in a computer", the correct response is not alarm. It is: yes, and money has worked that way since 3000 BCE. The coin era, roughly 600 BCE to 1971 CE, is the unusual chapter, not the normal one.
Interest, and the invention of the debt crisis
Mesopotamia gave us the world's first interest rates, and they were brutal. The standard rate on a silver loan was 20 per cent per year; on barley, 33 per cent. The Code of Hammurabi (c. 1750 BCE) capped them at exactly those levels, which tells you they were being exceeded.
Compound interest grows faster than agriculture. A farmer who borrowed in a bad year could not repay in a good one. The sequence was always the same: first he pledged his oxen, then his fields, then his children, then himself. Within a few generations a large share of the free population would be in bondage to creditors, the army could not be raised because soldiers were debt-slaves, and the state faced collapse.
The Mesopotamian solution was radical and worked for two thousand years: the clean slate. On the accession of a new king — and sometimes more often — all agricultural debts were cancelled by decree, bond-servants freed, pledged land returned. The Sumerian word was amargi, literally "return to the mother", and it is the first written word for "freedom" in any human language. Its original meaning was debt cancellation.
Recorded royal debt cancellations in Mesopotamia: at least 30 between 2400 and 1600 BCE. Standard silver interest: 20% per annum. Standard barley interest: 33% per annum. Hammurabi's Code, §117: a debt-servant must be freed after three years regardless of the sum owed.
The clean slate worked because the creditors were mostly the palace and the temple — the state was cancelling debts owed to itself. Once creditors became a private class with political power, cancellation became impossible. Rome tried and failed; the resulting land concentration and debt bondage contributed to the collapse of the Republic. Every modern debt crisis in Part VI of this report — Argentina, Sri Lanka, Zambia — is the same problem: debts that arithmetically cannot be paid, and a creditor class with enough power to prevent them being written off. Five thousand years, and we have not solved it.
What Mesopotamia proves
- Money is a unit of account before it is anything else. The measuring function came first; the payment function came later.
- Credit is older than currency. Debt tablets predate coins by 2,700 years.
- Money is a creature of the state. Temples and palaces set the standard, fixed the equivalences, and enforced the contracts. Money did not bubble up from the market; it was imposed from above and then adopted by the market.
- Debt has a political limit. Beyond a certain point it must be written off or it destroys the society carrying it.
3The Invention of the Coin
Around 600 BCE, three civilisations independently hit on the same idea within about a century of each other: stamp a lump of metal with a mark and let the mark do the work of trust. It changed war, empire and everyday life.
The problem with using silver as payment is verification. Is this lump pure? Does it weigh what you claim? Every transaction required a scale and a sceptic. The coin solved this with a piece of psychological engineering: a stamp from an authority that vouched for weight and purity, so that the buyer could count instead of weigh.
Three births
Lydia, western Turkey, c. 630–600 BCE. King Alyattes and later his son Croesus struck lumps of electrum — a natural gold-silver alloy from the Pactolus river — stamped with a lion's head. Croesus later issued the first pure gold and pure silver coinage, which is where "rich as Croesus" comes from.
India, c. 600–500 BCE. The Mahajanapadas, the sixteen great kingdoms of the Gangetic plain, produced punch-marked coins called karshapana: silver pieces cut to a standard weight and struck with multiple small symbols — sun, wheel, animals, hills. They circulated from Gandhara to Magadha. The Arthashastra, attributed to Kautilya around 300 BCE, describes a full state mint with a lakshanadhyaksha (superintendent of coinage) and an examiner of counterfeits.
China, c. 600–500 BCE. Bronze objects shaped like miniature spades and knives — representations of the tools they replaced — eventually standardised as the round coin with a square hole, which survived essentially unchanged for over two thousand years.
A coin is a lump of metal with a government's signature on it. The signature says: "I certify this is 8 grams of silver, and if you doubt me, you are calling the king a liar." You no longer need scales or a chemist. You count. That single shortcut is why coins spread across the entire Old World in about three hundred years.
What coins were really for: paying soldiers
Coinage did not appear in the most sophisticated commercial societies. Mesopotamia had banking, interest, contracts and international trade for 2,500 years without coins. Coins appeared in societies with something else: professional armies.
An army on the move is the ultimate stranger-economy. Thousands of armed men arrive somewhere, need food, and have no relationship with the locals. The state's solution is elegant and brutal: mint coins from captured bullion, pay the soldiers in coins, then demand that the local population pay taxes in those same coins. The population must now sell food to the soldiers to obtain coins to pay the tax. The army is provisioned without the state organising a single supply convoy.
This is the engine of all state money, then and now. A government creates a token, then creates demand for that token by requiring it in payment of taxes. You accept rupees not primarily because you trust the RBI but because the Income Tax Department will accept nothing else. This single mechanism — economists call it chartalism — explains why a piece of paper with no backing has value, and it is the missing half of the argument when people ask "but what is the dollar actually backed by?" Answer: by the fact that 340 million people must obtain dollars every April or go to prison.
Debasement: the first inflation
Every coin-issuing state eventually discovered the same trick. If a coin's value comes from the stamp, why put full silver in it? Melt down 100 coins, add cheap copper, restrike 110. The state gains 10 coins from nothing. This gap between a coin's face value and its metal cost is called seigniorage, and it is the ancestor of every modern government's ability to finance itself by printing.
Table 3.1 — The debasement of the Roman silver denarius
| Emperor | Period | Silver content | Consequence |
|---|---|---|---|
| Augustus | 27 BCE – 14 CE | ~98% | Stable prices, expanding empire |
| Nero | 54 – 68 | ~93% | First deliberate debasement, to fund rebuilding |
| Marcus Aurelius | 161 – 180 | ~75% | Plague and German wars |
| Septimius Severus | 193 – 211 | ~50% | Army pay raised 33% |
| Gallienus | 253 – 268 | ~5% | Hyperinflation; prices up ~1,000% in a generation |
| Diocletian | 284 – 305 | reformed | Price Edict of 301 fixes 1,200 prices; death penalty for breaches; fails within years |
Diocletian's Edict is the first recorded national price-control programme. It did not work. Chapter 27 explains why price controls fail so reliably.
Claim: "Inflation is a modern problem caused by leaving the gold standard."
Reality: Rome inflated its currency by more than 99 per cent while on a pure metal standard. Medieval England, Ming China, and sixteenth-century Spain all suffered severe inflation with gold and silver money. Metal restrains a government only if the government chooses to be restrained; if it does not, it debases. What actually changed after 1971 is not the existence of inflation but its mechanism — and, in the good cases, the arrival of independent central banks that made low inflation a stated legal objective. Chapter 28 has the full argument.
When a government debases, people are not fooled for long. They spend the new, low-silver coins and hoard the old, high-silver ones. Within a few years, only bad coins circulate. This is Gresham's Law: bad money drives out good, when the law forces both to be accepted at the same face value.
India lived this in reverse in November 2016. When ₹500 and ₹1,000 notes were demonetised, they instantly became "bad money" — and were dumped into the banking system as fast as people could manage, while the new ₹2,000 note was hoarded. Same law, running in the direction the state chose.
4Paper: China's Thousand-Year Head Start
China invented banknotes around 1000 CE, ran the world's first national paper currency, produced the world's first hyperinflation, and then abandoned paper money entirely for four hundred years. Europe repeated every stage of the experiment, several centuries late.
The origin is mundane. Sichuan province in the Song dynasty used iron coins, which were heavy: a purchase of one bolt of silk required about 40 kilograms of currency. Merchants began depositing coins with trusted shops and carrying the receipt instead. Around 1023 the Song government took over the practice and issued the jiaozi — the world's first government-issued paper money.
It worked because the state made it work: the notes were printed with multiple colours and intricate seals to defeat forgery, they were redeemable for coin, they were issued in limited series with expiry, and they were accepted for taxes. The tax loop again.
Paper money went through three stages, and every country repeats them.
Stage 1 — the warehouse receipt. You leave gold with a goldsmith; he gives you a paper that says "the bearer may collect 10 grams of gold". The paper is as good as gold, and lighter. This is fully backed.
Stage 2 — the partly-backed note. The goldsmith notices that only a few people come to collect on any given day. So he issues more paper than he has gold. This is where money starts being created.
Stage 3 — fiat. The state drops the promise entirely. The paper is worth something because the state says it is and demands it back in taxes. Kublai Khan did this in 1260. The United States did it in 1971. Every country in the world is at Stage 3 today.
Paper money has one unforgiving property: the cost of producing more of it is essentially zero. Under a metal standard, a government wanting more money had to find a mine or win a war. Under paper, it needs a decision. Every single hyperinflation in recorded history — Yuan China, revolutionary France, Weimar Germany, Hungary 1946, Zimbabwe 2008, Venezuela 2018 — happened under paper or near-paper money, and every one followed the same trigger: a government with obligations it could not meet from taxation, and a printing press.
This is not an argument for gold. It is an argument for institutions. The question that matters is not "what is the money made of?" but "who can create it, and what stops them?" Hold that question. It is the real subject of this entire report.
Why Europe caught up and overtook
China invented paper money and abandoned it. Europe adopted it late and built the modern world on it. The difference was institutional.
When the Bank of England was created in 1694, it was not an arm of the King. It was a private company of subscribers who lent to the Crown, and critically, Parliament — not the monarch — pledged specific tax revenues to service the debt. A king could default; a Parliament that represented the creditors would not. This is the origin of the phrase "the full faith and credit of the government", and it is the reason Britain could borrow at 3 per cent while France, an absolute monarchy with a larger economy, paid 6 per cent and eventually went bankrupt into a revolution.
Remember this mechanism. In Chapter 13 it reappears as the single most underrated pillar of American monetary power: the credibility that comes from constraints on your own government.
5Goldsmiths, Banks and the Money Banks Invent
Around 97 per cent of the money in a modern economy was not printed by any government. It was typed into existence by commercial banks when they made loans. This is the least understood fact in economics, and it is not controversial — the Bank of England published a paper saying so in 2014.
Return to the goldsmith. He holds gold for a hundred customers and has issued a hundred receipts. He observes that on a typical day only three or four people come to withdraw. So he has, sitting in his vault, gold that nobody is asking for.
He makes a loan. But he does not hand over gold. He writes the borrower a new receipt. There are now 110 receipts against 100 units of gold. The borrower spends his receipt; the shopkeeper who receives it does not rush to the vault either. Nobody has been defrauded and nothing has broken. The goldsmith has created ten units of money out of a judgement about human behaviour.
You go to a bank for a ₹30 lakh home loan. You imagine the bank takes ₹30 lakh from depositors' savings and hands it to you. That is not what happens.
The bank types two numbers into its computer. On one side it writes: the customer owes us ₹30 lakh (an asset). On the other side it writes: we owe the customer ₹30 lakh (a deposit — a liability). Your account balance now reads ₹30,00,000. That balance is money. You can spend it anywhere. It did not exist ninety seconds ago and it came from no depositor.
When you repay the loan, the reverse happens: the deposit is cancelled against the debt, and the money is destroyed. Money in a modern economy is not a fixed pile that circulates. It is a level that rises when people borrow and falls when they repay.
What stops banks from creating infinite money
Four things, and none of them is "the amount of deposits they have".
1. Capital requirements
A bank must hold shareholders' own money — capital — against its loans, typically 8–13% of risk-weighted assets under the Basel rules. Each new loan consumes capital. When capital runs out, lending stops. This is the real brake.
2. Reserve and liquidity requirements
Indian banks must park a share of deposits with the RBI as the Cash Reserve Ratio (CRR, 4% in 2026) and hold government securities as the Statutory Liquidity Ratio (SLR, 18%). If your borrower pays someone at another bank, you must settle in central bank reserves — which you must buy.
3. Profitability and risk
A loan that is not repaid destroys the bank's capital. Banks lend only where they expect repayment; in a recession they stop lending precisely when the economy needs it most. This is why recessions feed themselves.
4. The central bank's interest rate
The price of the reserves banks must obtain to settle payments. Raise it and lending becomes expensive; lower it and lending becomes cheap. This is the main lever of monetary policy and the subject of Chapter 29.
Of India's broad money supply (M3), physical currency in circulation is roughly 12–13 per cent. The remaining 87–88 per cent exists only as bank deposits — entries in databases. In the UK the figure is about 3 per cent notes and coins; in the US, roughly 8–10 per cent depending on the measure. Most money has no physical existence anywhere.
The three kinds of money you actually use
Table 5.1 — The hierarchy of money
| Type | Who creates it | Who can hold it | Risk if issuer fails |
|---|---|---|---|
| Central bank reserves | RBI / Federal Reserve | Only commercial banks and government | None — the issuer cannot run out of its own currency |
| Cash (notes & coins) | Central bank / treasury | Everyone | None, except inflation |
| Bank deposits | Commercial banks, by lending | Everyone | You lose it — unless insured (in India, up to ₹5 lakh per depositor per bank) |
| E-money / wallets / UPI balances | Payment firms, on top of bank deposits | Everyone | Depends on where the float is held |
| CBDC (e₹, digital yuan) | Central bank directly | Everyone (pilot stage) | None — this is the point of it |
If 97 per cent of money is bank deposits, and banks hold only a fraction in reserves, then no bank on earth can repay all its depositors at once. Not one. This is not a flaw that regulation has fixed; it is the design. Banking works because everyone does not ask at once.
When enough people do ask at once — Northern Rock in 2007, Silicon Valley Bank in March 2023, Yes Bank in India in 2020 — the bank fails even if its loans were sound, because the loans are long and the deposits are short. Deposit insurance and central bank lending exist purely to stop the panic from starting. The entire modern financial system rests on a confidence trick that is honest about being one.
In September 2019 the RBI capped withdrawals at Punjab and Maharashtra Cooperative Bank at ₹1,000 after discovering that over 70 per cent of its loan book had gone to a single bankrupt property developer, concealed through thousands of fake accounts. Roughly 900,000 depositors were frozen out; several died waiting. Deposit insurance was then ₹1 lakh; it was raised to ₹5 lakh in February 2020 partly in response.
The lesson is not that banks are evil. It is that a deposit is not "your money in a vault". It is a loan you have made to a bank, and it is only as safe as the bank's loan book and the state's willingness to stand behind it. This distinction becomes decisive in Chapter 13, where we ask why the world trusts American bank deposits more than anyone else's.
6So What Is Money?
A working definition, the three jobs money does, the six properties good money needs, and the one question that determines whether any currency — rupee, dollar, yuan or bitcoin — can succeed.
After five chapters of history, here is the definition this report will use:
Unpack that. Socially and legally enforced: courts uphold it, taxes require it. Claim on real resources: money is not wealth, it is a ticket to wealth. Confident expectation that others will accept it: this is the whole ballgame. Money is a belief that is true because it is widely held. Nothing else about it matters as much.
The three jobs
Table 6.1 — The three functions of money
| Function | What it means | Everyday example | What breaks it |
|---|---|---|---|
| Medium of exchange | Everyone accepts it in payment | You buy tea with a ₹10 note | Hyperinflation — shops stop accepting |
| Unit of account | Prices and debts are measured in it | A car is "₹9 lakh" | High inflation — people start quoting prices in dollars |
| Store of value | It holds worth over time | Savings for a house | Inflation, devaluation, confiscation |
These fail in a specific order, and the order is diagnostic. Store of value fails first (people buy gold, dollars, property). Unit of account fails second (menus in dollars, rents in dollars). Medium of exchange fails last (shops refuse the currency outright). When you see a country pricing things in dollars internally — Zimbabwe 2008, Lebanon 2021, Argentina for decades — the currency is already in stage two of death.
The six properties
Durable — survives handling. Portable — high value in small bulk. Divisible — splits without loss. Uniform — one unit identical to another. Limited in supply — scarce enough to hold value. Acceptable — others take it.
The last one is not just another item on the list. Acceptability is the only property that cannot be engineered. You can design a perfectly durable, portable, divisible, uniform, scarce currency and have it fail completely because nobody uses it. This is precisely the wall that the BRICS currency proposals, the SDR and most cryptocurrencies run into.
The network effect: why money is a winner-take-most market
Money gets more useful the more people use it — exactly like a language or a telephone network. A currency used by 1 per cent of the world is nearly useless to you: you cannot pay for anything with it. A currency used by 60 per cent is enormously useful even if it has flaws, because it will be accepted everywhere you go.
This produces a brutal dynamic. Small advantages compound into total dominance, and dominance persists long after the original advantage has disappeared.
Why do people still use QWERTY keyboards when better layouts exist? Because everyone else does, so all keyboards are made that way, so everyone learns it, so everyone else does.
Why does the world still use dollars when America is 26 per cent of the world economy but the dollar is 57 per cent of reserves and 50 per cent of payments? Same reason. You hold dollars because your supplier wants dollars, because his supplier wants dollars, because the commodity he buys is priced in dollars. Everyone individually is behaving sensibly. Collectively the result is a monopoly that no one voted for and no single participant can escape alone.
This is the central fact of this entire report. The dollar's strength is not primarily American power. It is everybody else's inability to move at the same time.
Sources: IMF COFER 2026Q1; SWIFT Global Currency Tracker, June–July 2026; BIS Triennial Survey. The final figure is out of 200% because every FX trade involves two currencies.
| ~26% | 57.1% | 50.1% | 81.2% | ~88% |
|---|---|---|---|---|
| US share of world GDP (nominal, 2026) | Dollar share of global FX reserves (2026 Q1) | Dollar share of SWIFT payments (June 2026) | Dollar share of trade finance (June 2026) | Of all FX trades have USD on one side |
Because acceptability is self-reinforcing, currency dominance does not decline gently in proportion to the issuer's economic weight. It holds, and holds, and holds — and then, if a genuine alternative appears and a critical mass moves together, it goes quickly. Sterling was 87 per cent of reserves in 1947, eleven years after Britain had already ceased to be the largest economy. By 1973 it was 7 per cent. The transition took roughly twenty-five years and was triggered not by economics but by two political shocks: Suez in 1956 and the devaluation of 1967.
This cuts both ways in the argument. Optimists about de-dollarisation should note that nothing has yet reached critical mass. Pessimists should note that the fall, when it came for sterling, was faster than anyone predicted.