Panic of 1873: The railway bubble that hit one of America's greatest financiers

ELI5/TLDR

In the 1870s, America went mad for railways. Companies borrowed staggering amounts of money to lay track across the continent, and ordinary people bought the IOUs. The most respected financier in the country, a man named Jay Cooke, bet everything on one risky railway line, ran out of cash, and his bank collapsed in a single morning. That collapse triggered the Panic of 1873, shut the stock exchange for ten days, and dragged the US economy into six years of misery. The hosts keep one eye on the present: the same shape of story, they argue, is playing out right now with AI.

The Full Story

This is an episode of the FT’s Story of Money podcast, hosted by Gillian Tett and Robin Wigglesworth. Wigglesworth is writing a book on the history of the bond market, and the railway boom is his favourite chapter — “a glorious, demented period of finance.”

The technology that physically built a country

First, the scale. Railways did to the 19th-century US what the internet did to the modern world, except in steel and dirt instead of cyberspace. Before the Civil War, America had about 25,000 miles of track, almost all of it east of the Mississippi. After the war, the country decided to stitch itself together coast to coast.

The difference from Europe matters. In Europe, railways connected towns that already existed. In America, railways created the towns — track went down first, and settlements grew up around it. That made each line less like infrastructure and more like a venture-capital bet. Hugely expensive, physically lethal (thousands of labourers, many of them Irish and Chinese immigrants, died building them), and financially risky. The payoff, when it worked, was a country transformed: suddenly you could eat Maine lobster in Los Angeles and Californian oranges in New York. An agricultural nation became an industrial one.

“There’s more glory in the rush of a single railway track than there was in the sacking of Troy.”

Bonds, not stocks, were the engine

Here is the part that gets forgotten. The railway boom didn’t run on the stock market — it ran on debt. A bond is just a loan you can buy and sell: you hand over money, you get paid interest, you get your money back at the end. Towns and railway companies issued these by the truckload.

The numbers are dizzying. US railways had $416 million in bonds outstanding by 1867; $2.2 billion by 1874; $5 billion by 1890. Adjusted for inflation that’s roughly $180 billion — which sounds modest until you realise the entire US economy in 1890 was only about $15 billion. So that bond pile was a third of GDP. Translated to today, Wigglesworth reckons it’s the equivalent of a $10 trillion capital-spending splurge in a single industry. For comparison, he pegs today’s AI data-centre build-out at around 1.2% of GDP; the railways were 10 to 15%. In relative terms, the railway mania dwarfs the AI boom.

Jay Cooke, the man who saved the Union and then broke the bank

The central character is a financier almost nobody remembers now: Jay Cooke. Calling him “the J.P. Morgan of his day,” the hosts say, undersells him — at his peak he was arguably the most powerful financier in the world, more powerful than the Rothschilds.

Cooke was a bundle of contradictions. He looked, in Wigglesworth’s words, “like Santa Claus if Santa Claus looked really mean.” An abolitionist who wouldn’t let people of colour ride his tram lines. A man who admired Native Americans, named his estate after a tribal chief, and then financed railways that tore through Native lands. He hated corruption and was very good at it.

His genius was selling bonds. During the Civil War he came out of early retirement to finance the North, and his innovation was mass selling — total war applied to finance. Bonds had always been sold to banks, insurers, and the wealthy. Cooke sliced them into tiny $50 denominations, ran newspaper ads, handed out leaflets and Q&As, and got ordinary citizens to fund the war. (The US copied this almost verbatim in WWI’s Liberty Loan program.) Ulysses Grant credited Cooke’s labours with keeping “the continued life of the nation.” Cooke, not a humble man, agreed:

“Like Moses and Washington and Lincoln and Grant, I have been, I firmly believe, God’s chosen instrument, especially in the financial work of saving the Union.”

That Civil War push had a long tail: it primed an entire population to buy bonds. People had been educated about them, had made money on them, and wanted more. So when the railway visions arrived, the buyers were ready — and the instruments multiplied into a carnival. Historian Richard White, quoted at length, describes investors moving “from government bonds to government secured railway bonds, to convertible bonds, to mortgage bonds… and from there potentially into the drink.” First mortgage bonds, second mortgage bonds, land grant bonds, income bonds — “bonds on anything and everything.” The bankers, he wrote, were “so many carnival barkers.”

The fatal mistake

Cooke, to his credit, stayed out of the transcontinental mania at first. He distinguished speculation from investment, and the second-transcontinental scheme — a northern line through Montana to Seattle that had already failed to raise money — was pure speculation. He said no twice.

Then his ego pulled him back in. He’d been the favourite to become Grant’s Treasury Secretary (the Hank Paulson move), and didn’t get the job. Stung, wanting to prove he was worth even more than they thought, he threw himself into the Northern Pacific Railway — which Wigglesworth says was barely a company at all, “a shell… an idea.”

The deal: Cooke would sell $100 million of the company’s bonds at 7.3% interest, buying them at a 12% discount to pocket the spread, plus advance some of his own cash, in return for an equity stake. The structure looked survivable — he’d only promised to try to sell the bonds and to advance a few hundred thousand dollars. But two things went wrong. The bonds wouldn’t sell (they languished, unsold, on his bank’s balance sheet). And the $500,000 advance ballooned to $7 million as he doubled down. The railway itself was a sinkhole: brutal terrain, incompetent executives he didn’t control despite his stake, and running battles with Native American tribes defending their land. He’d bet the house on black, and the wheel kept landing red.

How it actually broke — and the Austrian butterfly

The trigger came from an unexpected place: Austria. Vienna was having its own boom (the Gründerzeit, the age of entrepreneurs) which crashed in May 1873 — the Gründerkrach. European investors who lost money there started liquidating everywhere, and European money was a huge presence in the American bond market. A butterfly flapping in Vienna, a hurricane on Wall Street.

Jay Cooke & Co., sitting on unsold Northern Pacific bonds, was running dangerously low on cash. Note the distinction the hosts draw: this was a liquidity crisis, not necessarily insolvency. The bank likely had enough value in its assets; it just couldn’t turn them into cash fast enough to meet withdrawals. (Wall Street’s cash cycle then still followed the farming calendar — money got pulled out at planting and harvest.) On 15 September 1873, President Grant stayed overnight at Cooke’s estate. Nobody knows what they discussed. Three days later, with no government rescue and no central bank to flood the system with cash, the New York office told its Wall Street counterparties: a million dollars by 10 a.m. or we’re dead. The money didn’t come. They shut the doors.

“This was like a J.P. Morgan or a Goldman Sachs out of the blue suddenly saying, ‘We’re dead.’”

Cooke, told by telegram, wept in front of his staff — “this great strong man, the pillar of a nation.” He went personally bankrupt and lost his absurd estate (fake Roman ruins, a personal telegraph station, its own theatre) and moved in with his daughter. He clawed back a modest fortune by 1880 via a Utah silver mine and died not poor.

The wreckage, and the bigger argument

The economy did not recover for six years. Over 120 railways with $550 million in bonds went bust by end of 1875. Half of US iron foundries closed. Corporate bankruptcies peaked at 10,000 in 1878. Deflation set in, and under the gold standard the government had almost no tools to fight it — a “libertarian’s dream” of hard-money creative destruction. It was called the Great Depression until the 1930s claimed that name; historians now call it the Long Depression.

The episode closes on the question Tett keeps circling: are bubbles bad? Drawing on Daniel Gross’s book Pop, the argument is that manias build infrastructure that outlives the investors who funded it — fibre-optic cable from the dot-com bust, railways from 1873. “The optimal number of manias is definitely not zero.” But the hosts are careful about AI. The railway track laid in the 1870s is still physically there 150 years later. Data centres may not be: chips have a half-life, and the whole sector is betting on one form of AI (large language models) when world models or neuro-symbolic approaches might supplant it — leaving “stranded assets,” a VHS-versus-Betamax outcome. Not all capex booms are created equal.

“The four most dangerous words in finance: this time is different.”

Key Takeaways

A bond is a tradable loan — you lend money, collect interest, get the principal back. The railway boom was financed by bonds, not stocks, which is the detail most histories skip.
Scale in relative terms beats absolute terms. $5 billion of railway bonds in 1890 was a third of a $15 billion economy. Adjusted to today, a ~$10 trillion single-industry splurge — roughly 10x the AI build-out as a share of GDP.
Jay Cooke’s Civil War innovation was mass-market bond selling — tiny $50 denominations, newspaper ads, leaflets — which primed an entire population to keep buying bonds afterward. (Reused as WWI’s Liberty Loans.)
Liquidity crisis ≠ insolvency. Cooke’s bank probably had enough underlying value; it just couldn’t raise cash fast enough to meet withdrawals. That distinction is what “bank run” actually means.
No Federal Reserve, no deposit insurance, gold standard. With no lender of last resort and no ability to cut rates or print money, a single bank failure could domino, and the government had almost no tools to soften the deflation that followed.
Contagion was already global in 1873. A crash in Vienna (the Gründerkrach) forced European investors to liquidate American bonds — the spark that lit the fuse.
The “good bubble” thesis: manias overbuild infrastructure that society keeps even after investors are wiped out (railways, fibre-optic cable). But durability varies — physical track lasts 150 years; data centres full of fast-depreciating chips betting on one AI architecture may not.

Claude’s Take

This is popular financial history done well — two knowledgeable hosts, a genuinely good central character, and a clean through-line from 1873 to the AI question without being heavy-handed about it. The Jay Cooke material is the strongest part: a real person, well-sourced (contemporary New York Times reporting, an early biography, Richard White’s scholarship), with the contradictions left intact rather than sanded smooth.

The 7/10 reflects what it is: a smart, well-told podcast episode rather than a rigorous analysis. The AI comparison is the weakest thread — provocative but unfalsifiable, and the hosts know it (“very hard to know how stupid they are… before the fact”). The relative-GDP framing (railways at 10-15% vs AI at 1.2%) is the single most useful number here and deserves more scrutiny than it gets — it’s doing a lot of rhetorical work. The liquidity-vs-solvency distinction is correctly drawn and is the kind of mechanism worth carrying forward; it’s the same fault line as every bank run since. No BS to flag — the hosts are appropriately hedged, and where they speculate (what Grant and Cooke discussed) they say so.

Panic of 1873: The railway bubble that hit one of America's greatest financiers | The Story of Money