Jay Cooke built the retail machine that sold hundreds of millions of dollars of Union war bonds directly to farmers and mechanics. A decade later, his firm collapsed in a single morning, and touched off the Panic of 1873.
The bonds that broke him were meant to fund themselves: sell railroad bonds, use the proceeds to build track, earn land grants, sell the land, and use that cash to retire the bonds. A year into construction, the land sales came to $338.76.
"If we can only now manage to wriggle along until we reach the Missouri River, all will come out right."
Daniel Gamboa and Matt Harris co-host Backtest, a market-history podcast now in its second season; this episode is the fourth and final part of a series on the 19th-century railroad capital cycle that began with Cornelius Vanderbilt.
I listened to the full episode so you can skip it. 66 minutes of audio, 15 minutes of reading.
Here are the 8 takeaways that matter.
🎙️ Hosts: Daniel Gamboa and Matt Harris, co-hosts of Backtest, a podcast on defining moments in market history
📰 Published: 10 September 2026
🟣 Apple Podcasts | 🔗 Episode page | ⏱️ 1 hr 6 min | ✅ Time saved: 51 min
Key Takeaways
Jay Cooke built the retail machine that financed the Union's Civil War debt, selling directly to farmers and mechanics
He ran subscription offices out of corner drugstores at midnight, with coffee and donuts, and sold $830 million of one bond issue alone
After the war, his core business shrank 80% by 1869, and he passed three separate times on financing the Northern Pacific Railroad before finally saying yes
The Rothschilds, the world's biggest private bank, refused to touch the deal, calling the US "a blasted country"
The contract gave Cooke 12% commission, a board seat, custody of the railroad's funds, and about 60% of the company's equity
The flywheel depended on land sales funding bond buybacks; a year in, those sales totaled $338.76
His partner Harris Fahnestock warned him in writing, more than once, to stop funding the railroad before it broke the bank
A run on cash reserves closed the firm's doors within hours on September 18, 1873, closing the New York Stock Exchange for ten days
The Northern Pacific itself survived, reorganized, and was completed a decade later under JPMorgan and James J. Hill
Their closing lesson: durable customer problems and real bottlenecks matter more than the size of the capital raised
1. The Boy Broker of 1839
Gamboa and Harris open with Jay Cooke's start: a clerk's job in Philadelphia at 18, in the depths of a seven-year depression following the panic of 1837.
Cooke took a clerk's job at the brokerage E.W. Clark and Company in 1839, arriving from Ohio into a market where nobody trusted paper money. With state banknotes trading at uncertain values across the country, brokers who understood the discrepancies could make outsized spreads — "a grand time for brokerage and private banking," in Cooke's own words at the time
He moonlighted as a financial columnist for the Philadelphia Daily Chronicle, turning the firm's correspondence into nightly market reports. By September 1840 he was already diagnosing the crash correctly, attributing it to "wild speculations and imaginary prosperity"
Cooke's letters home reveal a man who saw money as something to spend, not hoard. "I look upon riches but as not more than the means whereby one can display his social and generous spirit." His firm would later give away 10% of its profits every year
He made partner at 21 and became the senior figure across E.W. Clark's loosely coordinated network of city offices — separate partnerships sharing one name, a structure that worked fine on commission business but became dangerous once the offices started carrying securities on credit. Cooke saw the risk and warned his partners to "snug up" in 1854; they kept riding the boom for three more years before the panic of 1857 took down Boston, then New York, then Philadelphia in a chain
Cooke turned some of his own property over to cover the firm's obligations when it dissolved, then spent three years as what Harris called an early angel investor before opening Jay Cooke and Company on January 1, 1861 — two months before Lincoln's inauguration, with roughly $5,000 to $10,000 of starting capital
2. Bonds for the Common Man
Pennsylvania needed war financing in 1861 with damaged credit from a prior default. Cooke's pitch to the state treasurer, and later to the Union Treasury, built the machine that made him famous.
Cooke talked Pennsylvania's treasurer into pricing a $3 million war loan at par, on patriotism alone, when the expectation was 75 cents on the dollar. He built a distribution network from scratch — newspaper ads, county agents, bonds sold in units as small as $50 — that oversubscribed the issue with banks, railroads, insurers and, above all, individual farmers and merchants
The Philadelphia treasurer later revealed that his rival, Anthony Drexel, had warned him off Cooke's firm entirely. Drexel told him not to bank with Cooke "for the reason that you had no cash capital" and the risk was too great — advice the treasurer ignored, and which Gamboa and Harris both note was accurate
Treasury Secretary Salmon Chase named Cooke one of 148 regional bond agents in 1861; Cooke outspent his $150 advertising budget by 20 times and outsold everyone. His brother Henry, a Washington insider close to Chase, helped land him the job as sole agent for the five-twenty bonds in October 1862, a mandate of a kind Harris said was fraught with corruption at the time. "Jay Cooke is executing this with, as far as we could tell, basically zero shenanigans."
He scaled the machine to 2,500 agents and ran it by telegraph, with hourly sales reports. A headline he wrote and placed everywhere read, in part, "a solemn duty to perform to your government." By January 1864 he had placed $500 million
The 1865 seven-thirty bond campaign pushed the same model further — night subscription offices in drugstores and groceries, coffee and donuts included. He sold $830 million of them; one Vermont agent wrote that farmers in log cabins had thousands of dollars out on loan to the government
The Union entered the war essentially debt-free and exited it $2.8 billion in debt, more than half of it raised through Cooke's firm. Historian Henrietta Larson wrote that Cooke's confidence rested on "great self-confidence and a never wavering faith" in his own righteousness, and Cooke himself later wrote that he had been, "Like Moses and Washington and Lincoln and Grant," God's chosen instrument in saving the Union
3. Hunting the Next Big Thing
With Treasury refinancing work turning competitive and unprofitable, Cooke spent the late 1860s searching for a new business to replace it.
Cooke's firm became a quasi-venture-capital shop, and mostly struck out. An overheating steam wagon, a self-lighting match, gold mines, a Central American canal — he passed on all of them, sticking to his rule of only backing things already proven profitable. An anthracite coal mine, an ironworks and a federally chartered life insurance company all underperformed, partly because bankers make poor operators
The Union Pacific and Central Pacific had already locked up the best relationships by the time Cooke looked seriously at railroads. The Northern Pacific approached him three separate times in the late 1860s; he turned it down each time as unproven and unprofitable
In 1868 he did take on one small line, the Lake Superior and Mississippi, and it went well — the bonds sold out in about a week, with demand for ten times more. A trip to Duluth left him convinced: "I felt sure that vast cities would grow up at Duluth and Superior"
His government bond business fell 80% in 1869, just as the golden spike ceremony connected the Union Pacific and Central Pacific in Utah — the country's first transcontinental line, cutting a New York-to-California trip from months to about a week. The bankers who financed it made a fortune. Cooke had passed
The government's incentive structure for that first transcontinental was land plus bonds: alternating square-mile sections along the route, and federally backed 30-year bonds paid out per mile of completed track — a subsidy tied to genuine strategic stakes, since a coast-to-coast railroad secured US territorial claims against Britain, Russia and France and advanced the policy of westward expansion
4. The Deal Nobody Else Wanted
Six days after the golden spike ceremony, Cooke signed on to finance the Northern Pacific — a project every sophisticated investor in two countries would go on to refuse.
Cooke's diligence began with his own survey rather than the promoters' numbers, over the objection of one Northern Pacific director; the other backed him. The results were encouraging: an estimated $85 million construction cost against the $100 million he planned to raise, plus a marketing report describing "the valley of inestimable value" along the route
He sent a partner to London to pitch the Rothschilds, then the largest private banking empire in the world, and was turned away flat. The size of the ask and the absence of a single completed mile of track killed it; the Rothschilds had already written America off as, in their words, a "blasted country"
The final contract, signed January 1, 1870, was the largest single corporate financing in the world to that point. Cooke committed to raising $5 million within 30 days himself and up to $95 million more on a 12% commission, with no personal obligation to fund the larger amount
The land-and-bonds structure was designed as a self-funding flywheel: build track, earn land, sell land, use the proceeds to retire bonds early, and let bondholders convert directly into land if they preferred. Cooke separately secured board seats, custody of the railroad's deposits, roughly 60% of the company's equity, and half of a new land-sale company
Every sophisticated banker who looked at the deal passed. The 7.3% coupon wasn't enough to compensate for financing "a railroad basically from nowhere to nowhere," as Vanderbilt reportedly put it later — and almost nobody believed the land story. Cooke hit his first $5 million target by January 24, 1870, but his final buyer list was, in the hosts' words, essentially all his friends
5. A Flywheel That Never Turned
Construction started in February 1870. The land sales that were supposed to fund it never materialized, and Cooke kept advancing cash anyway.
A renewed push into London and Europe failed again in 1870, this time compounded by the outbreak of the Franco-Prussian War. By year-end the initial $5 million was spent on construction, marketing and commissions, without finishing the track it was meant to cover
Cooke redirected his war-bond sales machine at US retail investors instead, placing the pitch in more than 1,300 newspapers — and it barely moved. Meanwhile 1871 was the single biggest year for railroad construction in American history as a share of GDP, with roughly 6,500 miles laid nationwide; almost none of that capital found its way to Northern Pacific bonds
Construction problems compounded the funding shortfall. Frozen ground in Minnesota buckled and sank once it thawed, forcing costly rework, while Northern Pacific president Gregory Smith ran the railroad from a distance while also running the Vermont Central, rarely visiting the site and ignoring the expectation that he consult Cooke before major spending. One overdraft prompted the sardonic internal joke that "there will be the usual growl from Philadelphia"
Cooke's wife died in July 1871, in what both hosts flagged as the hardest personal stretch of his life to that point
By early 1872, a US senator publicly said the Northern Pacific's land wasn't worth five cents an acre, against the $2.50 the railroad was advertising. Partner Harris Fahnestock, watching New York deposits drained by Philadelphia's advances to the railroad, began writing Cooke directly: replace the management, and stop the advances before they sink the firm. By the time Fahnestock wrote that letter, land sales for the year totaled $338.76 — not $338,000, three hundred and thirty-eight dollars
Cooke showed up to a Northern Pacific board meeting for the first time in years that August to walk directors through his own firm's deteriorating finances. The company still closed 1872 as one of its most profitable years on paper, even as it sat on the edge of running out of cash
6. The Panic of 1873
Two shocks hit railroad confidence in the fall of 1872, and the following September, Cooke's firm ran out of road.
The Crédit Mobilier scandal broke in September 1872, revealing that a construction company tied to Union Pacific executives had systematically overcharged the railroad to bribe members of Congress, including the vice president and the Speaker of the House. Railroad stocks and bonds became radioactive across the market
Days later, US Army fighting with Native American tribes near the Northern Pacific's construction crews left dozens dead and wounded, with newspapers estimating 20,000 hostile forces in the area — a second body blow to investor confidence in the project
Fahnestock's final written appeal in the fall of 1872 warned Cooke that funding the railroad with depositors' money was unjust to the farmers and merchants who trusted the firm. Cooke's reply was to tell the London office to sell what it could and "have a little faith and confidence in the partners here." Advances to the Northern Pacific, contractually capped at $500,000, had already passed $1.5 million
The firm survived an October 1872 liquidity crisis only because a friend at First National arranged a loan after the Comptroller of the Currency backdated the firm's bank statement by a day. Fahnestock wrote that night, "we are in a perfectly helpless position"
By January 1873, Cooke wrote what the hosts called maybe the most important sentence of his life: "If we can only now manage to wriggle along until we reach the Missouri River, all will come out right." He sold assets, begged agents to hold bonds with a buyback promise, and tried the Treasury market again — only to be forced to share a refinancing deal he'd hoped to run alone, shrinking it from a possible $300 million placement to $50 million
The railhead reached the Missouri River on June 3, 1873, and freight and passenger revenue finally appeared — just over half a million dollars by month's end. But the firm had advanced nearly $7 million against it, mostly through deposits rather than earned capital, with land sales still minimal
On September 18, 1873, Harris Fahnestock told three New York bank presidents the firm needed $1 million by 10 a.m. to avoid failure. No one would lend against nothing left to pledge. Just before 11 a.m. he closed the New York office and delivered the suspension notice to the stock exchange floor by hand; Philadelphia closed by 11 a.m. the same day
The failure triggered a full-blown panic — fire sales, deposit runs, called loans — and the New York Stock Exchange closed for ten days, the longest shutdown in its history apart from World War One
7. The Vision, Decades Later
The Northern Pacific outlived the firm that built its first stretch of track.
The railroad reorganized within months, converting $30 million of bonds and unpaid interest into preferred stock and cutting expenses to the bone. It turned profitable by 1876 and was an attractive investment again by 1880 — at which point Drexel, JPMorgan and other investors who had passed at 83 cents on the dollar in 1873 came in to fund the rest of the build
JPMorgan backed the completion to Puget Sound in 1883 and Tacoma in 1888, and railroad operator James J. Hill eventually consolidated the line with his Great Northern, finishing what Cooke had drawn up
Hill succeeded where Cooke's operators failed partly because he ran the railroad himself, on-site, with land that had a decade of real settlement and traffic behind it — plus a healthier economy than the one that sank Cooke in 1873
Jay Cooke returned to the railroad as a guest in 1891, at age 70, riding the finish line into Tacoma. By then some 6 million people lived in areas the railroad had opened. He wrote in his journal that it was the fulfillment of prophecies he'd made decades earlier: "I felt that I was fully justified"
8. Optimism Outran Reality
Gamboa and Harris close by drawing the throughline from this series back to today's capital-intensive technology buildouts.
Harris's core takeaway is that optimism and reality move at different speeds when something is genuinely new. The Northern Pacific's flywheel — sell bonds, build track, earn land, sell land, repeat — was logically elegant but depended on flawless execution across three separate entities (the railroad, the land company and the bank), with no room for the delays that always show up in practice
He singles out bond pricing as the correctable mistake. Investors weren't only pricing in risk they'd never seen at this scale before; the 7.3% coupon itself was mispriced relative to market feedback, and Cooke never adjusted it
Gamboa's framework for today's AI buildout borrows directly from the Vanderbilt and Cooke stories: identify a durable customer problem, then find the real bottleneck. A permanent, acute problem — his example is Apple sourcing glass for iPhones — is worth capital; a temporary one isn't. Whoever survives the inevitable overbuilding and market panic with capital discipline intact gets to buy the wreckage at a discount afterward
Their bottom line is that the Northern Pacific's underlying vision was eventually vindicated, but Jay Cooke himself never got to finish building it — the gap between when a big infrastructure bet is right and when the market agrees is exactly what broke his firm.
Bonus Insights
Gamboa opened the episode comparing Cooke's story to Christopher Nolan's "Odyssey," not the more obvious Achilles-style glory story he used for Vanderbilt in the previous episode. His read: Cooke, like Odysseus, understood people and psychology and used that to engineer a huge win, only for his own decisions to go wrong for everyone around him
The hosts trace a straight line from 19th-century railroad financing to today's AI infrastructure buildout. Harris opens by crediting Ben Thompson's appearance on Invest Like the Best for the observation that Google built part of its infrastructure advantage by buying up "dark fiber" left over from WorldCom's collapse, deploying capital — some of it from Berkshire Hathaway, which now owns the Northern Pacific through BNSF — the same way earlier generations poured capital into rail ahead of demand
Harris compared Cooke's failed 1871 marketing blitz to basketball: "If Jay Cooke were a shooting guard in the NBA, his hand would be as cold as the ground in Minnesota in January 1871"
Gamboa compared contemporary skeptical commentary on Northern Pacific bonds to dot-com-era financial journalism, likening period warnings from the Commercial and Financial Chronicle to the Barron's articles that flagged the dot-com bubble decades later
Daniel Drew, Vanderbilt's frequent foe from the previous episode in this series, reappears as a footnote — his own Wall Street firm was one of two New York houses that failed in the week before Jay Cooke and Company went down
Books & Resources Mentioned
Jay Cooke, Private Banker – Henrietta Larson (The 1936 Harvard-published biography the hosts say unlocked this whole story for them)
1873: The Rothschilds, the First Great Depression, and the Making of the Modern World – Liaquat Ahamed (A friend of the show who answered the hosts' questions ahead of this episode; his book covers the same period)
Jay Cooke's Gamble: The Northern Pacific Railroad, the Sioux, and the Panic of 1873 – M. John Lubetkin
Panic on Wall Street: A History of America's Financial Disasters – Robert Sobel
Invest Like the Best with Patrick O'Shaughnessy (Where Ben Thompson of Stratechery made the Google/dark-fiber comparison the hosts open with)
If this was worth your time, send it to someone who has to have a view on this.
Get the latest market chatter as it happens: