The Ascent of Money: A Financial History of the World

By Niall Ferguson

11 min read

Why does it matter? Because the next financial crisis is already statistically overdue.

Most people treat a financial collapse the way they treat a car crash on the highway — a shocking interruption, something that happens to other people, a reminder to be more careful for a few weeks. Here's the problem with that framing: since 1870, the probability of any given country experiencing a major financial catastrophe in any given year has been roughly 3.6%. That's not a rare event. That's a structural feature. The question isn't whether the next crisis is coming — it's whether you'll recognize it in time, or whether you'll be the person buying Mississippi Company shares on borrowed money in 1720, convinced with impeccable logic that this time was different. Niall Ferguson's argument is simple and devastating: we keep getting destroyed not because we're stupid, but because we're human — and those turn out to be surprisingly difficult to tell apart.

Financial Catastrophe Is Not a Rare Anomaly — It Has a 3.6% Annual Probability

Financial crises are a permanent statistical feature of economic life, embedded in the system as reliably as interest payments and quarterly reports. The question is not whether one is coming. The question is when.

Since 1870, there have been 148 instances in which a country's GDP fell by at least 10 percent, and 87 in which household consumption suffered a comparable collapse. Work through the full sample and you get roughly 3.6 percent annual probability of financial disaster, per country, per year. That is not a once-in-a-generation rarity. It is the kind of odds that, applied to your workplace, would mean expecting a career-upending shock every few decades, not every few centuries. Budget for it accordingly.

The number is startling for what it reveals about our expectations. We treat each crisis as an aberration — a freak event, a product of greed or folly peculiar to this time and these people. Ferguson's argument is that this reaction is itself the pattern. The shock is never the crisis; it is always our shock at the crisis.

Part of why we keep getting ambushed is a distinction Frank Knight drew in 1921 between calculable risk and genuine uncertainty. A roll of dice is risk: the odds are knowable in advance. But the interest rate twenty years from now, or whether a liquidity crisis will materialize next summer, belongs to a different category entirely, one where, as Keynes put it bluntly in 1937, there is no scientific basis for forming any calculable probability at all. We simply do not know. The 3.6% tells you the weather is coming. It tells you nothing about which Tuesday.

Governments Always Find the Exit from Debt Discipline — It's Called Inflation

By the autumn of 1923, a German worker cashing his wages had to spend them immediately (within hours, if possible) before inflation consumed their value. The denominations were staggering: twenty-billion-mark notes for routine purchases. Nearly five hundred quintillion marks were in circulation by year's end. Prices had risen to 1.26 trillion times their pre-war level, and annual inflation peaked at 182 billion percent.

Ferguson's reading makes this more than a horror story. Germany's hyperinflation was a political act dressed as a monetary crisis. Weimar financial elites made a calculation: let the mark collapse deliberately, and the Allies would be forced to revise their Versailles reparations demands — hyperinflation would make payment visibly impossible and, in the meantime, cheapen German exports against British and American competition. The calculation was catastrophic. The French invaded the Ruhr instead of renegotiating. But the episode reveals something about the bond market's power over governments: that power is real, and it has a known escape hatch. It is called the printing press.

The discipline itself is genuine, which is what makes the escape so meaningful. Medieval Venice issued bonds that functioned as a live war-confidence index: they stood at 80 in 1497 and fell to 10 after the defeat at Agnadello in 1509, with investors repricing the probability of default in real time. Bond markets do punish reckless borrowers. They do cut governments off. But between the moment of borrowing and the moment of reckoning, there is always a gap — and in that gap, every major sovereign debtor eventually chose one of two exits: inflate the debt into worthlessness, or refuse to pay it outright. The bond market rules until the debtor decides it doesn't.

The Legend Says Rothschild Predicted Waterloo. He Nearly Went Bankrupt Because He Didn't.

On the evening of June 18, 1815, Nathan Rothschild received word from his couriers: Wellington had won. Napoleon was finished. The war was over.

Most people would have celebrated. Nathan's stomach sank.

The story everyone knows goes like this: a financial genius foresaw the outcome of Waterloo, raced the news to London, and bought British government bonds just before their price soared on the victory announcement. In 1940, Nazi propaganda minister Joseph Goebbels commissioned a film making this explicit: Nathan bribing French commanders, deliberately spreading false reports of British defeat to trigger a panic selloff, then sweeping in to buy consols at the bottom. Master manipulator. Financial Bonaparte.

None of it happened.

The actual story begins earlier. For most of the final years of fighting Napoleon, Nathan had been the British government's primary gold smuggler, buying up French and Dutch coins, packing them into hundreds of crates and barrels, and shipping them to Wellington's armies on the Peninsula. When Napoleon escaped Elba in 1815 and Europe lurched back toward war, Nathan made a confident bet: this would be a long fight, like all the others. He and his brothers furiously bought every ounce of gold they could find, advancing nearly ten million pounds to the British government. Commissions of two to six percent on such sums meant a fortune in the making.

Then Wellington routed Napoleon's army in a single afternoon.

Peace meant no more armies to fund, no more allies to subsidize, no more need for gold. The metal's price would fall. Nathan was sitting on an enormous pile of bullion that nobody wanted, bleeding value by the day. His brothers wrote desperate letters urging him to sell. He was staring at one of the largest potential losses in the history of private finance.

What he did next was not obvious. He converted the gold into British government bonds (consols) at prices depressed by years of wartime supply. His logic: with the war ended and government borrowing set to shrink, prices could only rise. He held for over a year while his brothers pleaded with him to take whatever small profits remained. In late 1817, with consol prices up more than forty percent, he sold. The profit, adjusted for today's purchasing power, was roughly six hundred million pounds.

A partner at Barings, the Rothschilds' chief rival, offered the plainest possible verdict: Nathan was, in money and markets, what Bonaparte had been on the battlefield.

The comparison is more precise than it sounds. Bonaparte was also famous for adapting to conditions that had already destroyed his original plan. The legend says Rothschild succeeded by predicting what happened. The truth is he succeeded by responding to what he'd gotten badly wrong.

John Law's 1720 Collapse and the Enron Implosion Follow the Same Five-Act Script

In September 1719, the narrow rue Quincampoix had become something Paris had never seen. British embassy clerks reported that princes, dukes, and duchesses were arriving before dawn and staying until dark, pawning jewelry and selling country estates to buy shares in the Mississippi Company. The word millionaire (invented in France that year) was coined here, for people who had made fortunes on a stock that had climbed from 500 livres to over 10,000 in less than eighteen months. John Law, a Scottish-born convicted murderer turned financier, was by then the most powerful man in France: he controlled all indirect taxes, the national debt, the royal mints, and the colony of Louisiana. "L'état, c'est moi" ("The state is me"), Louis XIV had once declared. Law's version was "L'économie, c'est moi."

He held it for about eight months. By December 1720 the shares were back to 1,000 livres and still falling, Law had fled France, and the monarchy's last clear chance to reform its finances before the Revolution had vanished.

The collapse produced a template: five stages that have repeated with remarkable fidelity ever since. The first is displacement: some genuine opportunity that makes higher prices plausible. This triggers euphoria, as rising prices attract more buyers, which raises prices further. Then comes mania, when easy-money rumors pull in first-time investors and the fraudsters eager to relieve them of it. Distress follows when insiders recognize the gap between price and any conceivable future earnings and start quietly selling. Finally, revulsion: outsiders notice the insiders leaving and stampede for the exit at once.

Three structural features keep appearing underneath: insiders with far better information than the buyers they're selling to; cross-border money flows that bring in sophisticated speculators who time it right and naive latecomers who don't; and, almost always, a central bank that kept credit cheap long enough for the mania to build.

The Enron story runs the same five acts so faithfully it reads like homage. Ken Lay, a Missouri preacher's son who built a gas-trading firm into America's fourth-largest corporation, promised investors that derivatives could manufacture profit from thin air — gold from gas, where Law had promised gold from paper. His executives said openly that loyalty was bought with money and the real goal was "multiples" of your base pay. The Houston Astros renamed their stadium after the company. Then the insiders unloaded shares while assuring analysts the stock would hit a hundred dollars. Then a senior employee named Sherron Watkins wrote a memo predicting implosion. Then, two weeks after Alan Greenspan collected the Enron Prize for Distinguished Public Service, the company filed for bankruptcy.

The pattern isn't invisible. It's just that by Stage 2, everyone around you is getting rich, and the cost of acting on the pattern is paid immediately while the reward for patience arrives later — if at all.

Two Nobel Laureates Calculated a 1-in-10²⁴ Chance of Ruin — Then Proved Their Models Wrong in a Single Day

Financial crises follow recognizable patterns — the same five-stage script from displacement to revulsion, the same 3.6 percent annual probability of disaster. Why do even the smartest people keep getting destroyed by them? Long-Term Capital Management was finance's most plausible answer: two Nobel laureates, Myron Scholes and Robert Merton, alongside a former Federal Reserve vice-chairman and John Meriwether, the trader who had built Salomon Brothers' bond desk into the most profitable operation on Wall Street, all backed by mathematical models so precise they quantified catastrophe as a one-in-10²⁴ event, something that should occur roughly once every four million years.

The fund opened in 1994 and returned 43 percent in year one, 41 percent in year two. By late 1997, $6.7 billion in capital supported $126 billion in borrowed assets (19-to-1 leverage), spread across over a hundred uncorrelated strategies and 7,600 positions. The partners were so confident they returned $2.7 billion to outside investors, preferring to trade their own money.

Then on 21 August 1998, Russia defaulted on its domestic bonds, and LTCM lost $550 million in a single day (fifteen percent of all capital), driving leverage to 42-to-1. Every bet went wrong at once. A hundred supposedly independent strategies correlated simultaneously. As one observer told Meriwether afterward: "John, you were the correlation."

The post-mortem was simple and devastating. The models ran on five years of data. Eleven years would have caught the 1987 crash. Eighty years would have caught Russia's last default, after the 1917 Revolution. The laureates had built models for how markets behave in theory and mistaken that for how they behave in history.

But the problem doesn't stop at data. It runs through human cognition in ways no dataset can correct. Psychologists Kahneman and Tversky demonstrated this with a simple experiment. Give two groups 1,000 Israeli pounds each. Ask the first: a guaranteed 500 more, or a fifty-fifty shot at 1,000? Eighty-four percent take the certainty. Tell the second group they're starting from 2,000: a guaranteed loss of 500, or a fifty-fifty chance of losing 1,000? Sixty-nine percent gamble. The expected values are mathematically identical. Framing the choice as a loss reversed the majority. Losses hit roughly two and a half times harder than equivalent gains.

That's loss aversion. Here's a different failure mode, more basic still, stripped of any financial context: a bat and ball together cost £1.10; the bat costs £1 more than the ball; how much is the ball? About half of people answer ten pence. The correct answer is five pence. Not a money problem, not a probability problem — arithmetic. The error is hardwired, not trained in.

Ferguson catalogs ten such biases. Two show the pattern. Availability bias: we overweight the crisis we just lived through, which is why every post-crash regulation fights the last war. Hindsight bias: once a crash happens it looks inevitable, which makes us confident we'll spot the next one. LTCM's founders were the most mathematically sophisticated investors in history. They made the same category of error as someone who answered ten pence. Better tools don't fix the brain holding them.

The Last Time Finance Was This Globally Interconnected, a Pistol Shot in Sarajevo Closed Every Stock Exchange

Imagine you are a financial economist in 1910. You can point to fifty-seven sovereign governments whose bonds trade in London, emerging-market spreads under two percentage points above British consols, and forty million pounds flowing annually into Chinese securities. You write a book arguing that this web of interdependence has made great-power war economically irrational, that the "delicate interdependence of credit-built finance" effectively guarantees peace. Prominent people praise it. Your name is Norman Angell, your book is The Great Illusion, and within four years you are proven catastrophically wrong.

On June 28, 1914, Gavrilo Princip shot an archduke in Sarajevo. By July 30, panic had seized every major financial market. Vienna closed first. Continental Europe followed. London and New York shut for five months. A generation of capital integration, built painstakingly over thirty years, was destroyed in days.

Ferguson's lesson: deep interdependence amplifies shocks rather than absorbing them. The tighter the web, the more catastrophically it tears.

The Chimerica of 2007 ran on the same logic. China was lending the United States more than four billion dollars every working day, buying Treasury bonds to suppress its own currency and keep exports cheap, flooding US mortgage markets with enough capital to make hundred-percent loans to borrowers with no income a routine product. Ferguson named this codependency Chimerica — and noted, without elaboration, that the word also means chimera.

The reason history keeps delivering these surprises to people who should know better is arithmetic. The average Wall Street CEO's career spans twenty-five years, meaning institutional memory at the top of American finance reaches back only to 1983 — a decade after the 1970s oil shock, decades before the last time a genuinely globalized system tore apart without warning. Every generation inherits stability that looks permanent because the last failure is just far enough away to feel theoretical. The history exists. The people who most need it have already stopped reading it.

[NOTE FOR CLOSING REVISION: The closing paragraph above — beginning "The reason history keeps delivering these surprises..." — appears nearly verbatim in the closing section. The closing should not repeat it; remove or rework that passage there.]

The Mirror Is Always Accurate — The Problem Is Who Looks Away

The irony Ferguson leaves you with is almost too neat: every crash described in this book — Law's millionaires reduced to paupers, Nathan Rothschild's near-ruin financing Wellington's campaign, LTCM's Nobel-calibrated certainty vaporized in a single afternoon — is documented, analyzed, and sitting in any decent library. The knowledge doesn't vanish between disasters. It simply stops being consulted by the people making the largest bets. Incentives run quarterly. History runs in centuries. Ferguson's image for this is a mirror — financial markets reflecting, every working day, how we value ourselves and the world's resources back at us. It keeps showing the same face. We just keep mistaking the reflection for something new.