Starvard.
Starvard University · How Banks Work
The book central bankers still reach for when a crisis hits, explained in plain words. Walter Bagehot mapped the London money market in 1873: what a bank really does with your deposit, why trust holds the whole system up, why runs start, and the famous rule for stopping a panic that the world still follows.
Hello, my name is Finn, and I'm going to help you understand this course.
Here it is, explained like you are 10:
A bank does not keep your money in a box. It lends most of it to other people, keeps a little ready for anyone who asks, and hopes not everyone asks on the same day. Most days that works beautifully. On the rare bad day, it can all fall apart at once. This course explains how banks work, why they sometimes break, and the famous rule for saving them.
Walter Bagehot was the editor of The Economist, and in 1873 he published Lombard Street, named for the London street where the great banks stood. He was tired of banking being treated as a mystery. The money market, he wrote, is as concrete and real as anything else, and can be described in plain words, so he described it: the deposits, the reserves, the bills of exchange, and the one great reserve at the Bank of England on which the whole structure rested.
The world he described has changed around his book. Britain then was on the gold standard, there was no deposit insurance, and the Bank of England had not yet admitted that rescuing the market in a panic was its job. This course keeps the contrast in view throughout, from the improvised rescue of 1907 to the creation of the Federal Reserve, deposit insurance in the 1930s, the crisis of 2008, and the phone app bank run of 2023, in which Bagehot’s rule was quoted and used again.
The whole course in brief
What you will be able to do after this course
The seven lessons
Lesson 1
Here is the fact most people never quite learn: when you deposit money in a bank, the bank does not put your bills in a box with your name on it. In law and in practice, you have lent your money to the bank, and the bank owes it back to you on demand. The receipt in your account is a promise, not a storage ticket. Walter Bagehot, the English journalist who wrote Lombard Street in 1873, set out to explain the system his country had built on exactly this promise. He begins by insisting, “the Money Market is as concrete and real as anything else; that it can be described in as plain words.” This lesson takes him at his word.
So what does the bank do with your deposit? It keeps only a fraction of it as ready cash, its reserve, and lends the rest out: to a business buying stock, to a family buying a house, to a trader carrying goods across the sea. The bank earns the difference between the small interest it pays you and the larger interest it charges borrowers. That difference pays for the building, the clerks, the app on your phone, and the profit. Your deposit, in other words, is working twice at once. It sits in your account, spendable by you at any moment, and it is also out in the world, built into somebody's shop, roof, or harvest. Banking is the machine that lets the same money do both jobs.
The catch has a name: maturity mismatch. Your deposit can be demanded today, this minute. The loans it funds come back slowly, over months or thirty years, and cannot be called in this afternoon without ruining the borrower. On any normal day this is no problem at all, because only a small share of depositors want cash at once, and new deposits flow in as old ones flow out. The bank needs only enough ready money, its reserve, to meet the normal daily tide. Bagehot saw that in England nearly all banks kept that reserve in one place, as deposits at the Bank of England, so the whole country's safety rested on one final pile of cash. Hold that picture. Everything else in this course follows from it.
Lesson 2
A bank runs on something you cannot weigh. Bagehot writes of deposit banking, “Its essence is that a very large number of persons agree to trust a very few persons, or some one person.” Thousands of strangers hand their savings to a handful of bankers they have never met, in buildings they never inspect, against accounts they cannot audit, and then sleep soundly. That quiet confidence, multiplied across a whole country, is credit. It lets a young shopkeeper borrow against next year's sales, lets a railway be built with the savings of clerks and widows, and lets trade move at the speed of a signature instead of the speed of a gold shipment.
Bagehot adds the sentence every banker should have framed: “Credit is a power which may grow, but cannot be constructed.” Trust accumulates slowly, through years of promises kept in dull times, and it cannot be ordered into existence by a law, a speech, or a new building with marble pillars. England's credit system, he argued, grew up gradually over two centuries, and its very age was its strength. People trusted the Bank of England and the great London houses partly because their grandfathers had trusted them. A new country, or a new kind of bank, cannot simply copy the furniture and expect the confidence to come with it.
The dark twin of this lesson is just as important. Because credit is belief, it can vanish like belief, suddenly and for reasons that have little to do with arithmetic. A rumor, a failure next door, a photograph of a queue outside another bank: any of these can turn trust into its opposite in a single morning, even at a bank whose loans are perfectly sound. That is why banking feels so calm for decades and so fragile in a crisis. The system is strong exactly as long as almost nobody tests it, and the rest of this course is about what happens, and what should happen, when everybody tests it at once.
Lesson 3
Now combine the first two lessons and the run explains itself. Depositors can all demand their money today. The bank has lent most of it out for months and years. If a rumor makes depositors fear the bank might fail, each one faces a simple, brutal calculation: the first people in line get paid in full, from the reserve and from whatever can be sold quickly, and the last people in line may get nothing. So even a depositor who believes the bank is basically sound has a reason to run, because the danger is not the bank's loans. The danger is everybody else running first. A run is not mainly a panic of fools. It is a stampede of reasonable people trapped in a bad set of incentives.
Worse, a run can kill a bank that deserved to live. To raise cash in a hurry, the bank must sell its loans and securities at whatever price a frightened market will pay today, and sound assets sold in a fire sale fetch fire sale prices. Losses that would never have happened, had the loans simply run their course, become real. The bank that was solvent on Monday can be genuinely broken by Friday, not by bad lending but by the speed of its depositors. Bagehot watched this logic at work in the London money market, where the deposits were held not only by families but by other banks, each one able to demand its money the same day, each one watching the others.
This is also why runs spread. If one bank fails, depositors at the next bank ask the obvious question, are we sure ours is different, and the queue begins there too. Bankers call it contagion, and it moves through a financial system the way fear moves through a crowd, by sight and by rumor. Bagehot's London had no deposit insurance and no modern central bank playbook, so a run was a regular terror of commercial life. The question his book answers, and the next two lessons take up, is whether anyone in the system has the job, and the tools, to stand in the doorway and stop the stampede.
Lesson 4
Somebody has to be the banker's bank. In Bagehot's England, that somebody was already the Bank of England, though it was slow to admit the job was its. Country banks kept their reserves in London. London banks kept their reserves as deposits at the Bank of England. So, as Bagehot puts it, the banking reserve in the Bank of England was not one bank's reserve among many. It was the final reserve of the whole country, the one pile of cash that every other pile of promises ultimately pointed to. When a panic came, every road in the money market led to the same door in Threadneedle Street.
Here was the problem Bagehot wrote his book to fix. The Bank of England was a private, profit seeking company with shareholders who wanted dividends, and its directors liked to think of themselves as ordinary bankers, free to protect their own till like anyone else. But an ordinary banker in a panic does the safe thing for himself: he stops lending, calls in loans, and hoards cash. If the holder of the final reserve does that, there is nowhere left for anyone else to turn, and the panic feeds on itself. Bagehot argues, with some heat, that holding the country's reserve is a public duty, that the duty must be acknowledged openly in calm times, and that the reserve must be kept large enough, deliberately, expensively large enough, to be worth something when the storm arrives.
The modern name for this role is lender of last resort. When every private lender has retreated, the central bank lends, because it alone cannot run out of the thing everyone is demanding: it issues the currency itself. Bagehot did not have that modern machinery in mind. The Bank of England of 1873 worked under the gold standard, and its notes had to be backed in ways that limited it. But the logic he laid down is the logic every central bank now follows, and the next lesson states his famous rule for exactly how that lending should be done.
Lesson 5
Here is the most quoted rule in all of central banking, stated by Bagehot in Lombard Street in two parts. Part one is about price. In a panic, he writes, “these loans should only be made at a very high rate of interest,” a rate he calls a heavy fine on unreasonable timidity. The high rate does real work. It scares away borrowers who do not truly need the money and are only grabbing cash out of caution. It protects the precious reserve for those in genuine need. And it draws money in from abroad and from hoards at home, because high interest is a magnet. Cheap emergency money, by contrast, invites the whole market to queue for insurance it does not need.
Part two is about quantity and security. At that high rate, he continues, “these advances should be made on all good banking securities, and as largely as the public ask for them.” Lend freely, on anything that was considered good collateral in normal times, to any solvent borrower who can pledge it. Refuse only the genuinely bad paper. The reason is psychological as much as financial. A panic is a fear that money cannot be had. The cure is the visible, unmistakable fact that money can be had, at a price, by anyone with good security. As Bagehot says elsewhere in the book, the holders of the final reserve must “lend freely, boldly, and so that the public may feel you mean to go on lending.” Half hearted lending, a little help offered nervously and then withdrawn, is the worst policy of all, because it spends the reserve without buying the confidence.
Put the parts together and you have the rule as central bankers still recite it: in a crisis, lend freely, at a penalty rate, on good collateral. Each word carries a lesson. Freely, so the panic starves for lack of fear. Penalty rate, so the help is a bridge, not a gift, and so it ends when the crisis ends. Good collateral, so the central bank rescues the solvent and the illiquid, not the insolvent and the reckless. Bagehot worked this out from watching the panics of 1847, 1857, and 1866, in which the Bank of England had lurched between refusing and relenting. The rule sounds simple. Applying it, in a screaming market, with imperfect information about who is truly solvent, is the hardest job in finance, which is why the next lesson is a tour of the times it was, and was not, followed.
Lesson 6
Start where Bagehot started, with 1866. The great London discount house Overend, Gurney and Company failed that May, and the panic that followed, the last great panic of his lifetime, convinced him the Bank of England's duty had to be written down plainly. The Bank, this time, lent boldly at very high rates, and the panic burned itself out. His book, published seven years later, is the report of a man who had watched the machine nearly break and wanted the repair manual on record. In 1907 the same drama played in the United States, which had no central bank at all. When New York trust companies faced runs, the rescue had to be improvised by the elderly banker J. P. Morgan, who locked his fellow bankers in his library until they agreed to fund it. The spectacle of a nation's finances depending on one rich man's dining room shamed Congress into creating the Federal Reserve in 1913.
Then came the great failure to follow the rule. In the early 1930s, thousands of American banks faced runs, and the young Federal Reserve, bound by the gold standard and by its own hesitations, did not lend freely enough to stop them. Roughly nine thousand banks failed, depositors lost their savings in queues, and the Great Depression deepened into catastrophe. The political answer changed banking forever: federal deposit insurance, beginning in 1934, which took the ordinary family's reason to run off the table entirely. In 2008 the run came back in modern dress, not as queues at tellers but as institutions refusing overnight funding to other institutions, and central banks around the world answered in Bagehot's exact words, lending freely, at scale, against collateral that a frightened market would no longer accept. In 2023, Silicon Valley Bank showed the new speed of the old disease: depositors, many holding balances far above the insurance limit, withdrew tens of billions in a single day by phone app, and regulators guaranteed deposits and opened a lending facility over one frantic weekend.
Notice what has changed since 1873 and what has not. The gold standard is gone, so central banks are no longer rationing a metal reserve. Deposit insurance exists, so insured depositors need never queue. Central banks now state the lender of last resort role openly, the very acknowledgment Bagehot demanded. What has not changed is the underlying machine of Lesson 1: money you can demand today, funding assets that pay back slowly, held together by confidence. Every generation rediscovers that machine in a new costume. Bagehot's rule survives because the machine does.
Lesson 7
Bagehot wrote for bankers, but his book ends up teaching depositors something too: most of your safety comes from arrangements made long before any panic, not from clever moves during one. Here is the practical version, in plain terms. This is general education, not advice about your own situation. First, know whether your money sits in an insured deposit. In the United States, deposits at FDIC insured banks are covered up to 250,000 dollars per depositor, per bank, for each account ownership category, and credit unions carry similar federal insurance. Money inside that cover does not need a queue, a rumor check, or a fast phone app. If the bank fails, the insurance pays, and in modern resolutions insured depositors have usually had access within days, sometimes with no interruption at all.
Second, notice where your money is not a deposit. Balances far above the insurance limit at a single bank, money in investment accounts, and money in products with bank like names but no insurance behind them all carry different rules, and a panic is the wrong time to learn them. Spreading larger sums across more than one insured bank, or across ownership categories, is the dull, effective classic. Third, keep a modest emergency cushion in an insured account you can reach, so that a banking scare is an interesting news story for you rather than a household emergency. The households hurt worst in the old panics were those whose every dollar stood in the queue.
Finally, practice panic behavior in advance, because in the moment everyone else will be practicing it against you. If headlines shout that a bank is failing, check two things before moving anything: is my balance insured, and is this a problem at my bank or a problem in the news? An insured deposit is the one financial asset you genuinely do not have to rescue. Bagehot's London depositors had no such floor under them, which is why their fear was rational and their runs were deadly. You live in the system his book helped design: insurance for the small saver, a central bank expected to lend freely in a storm, and rules meant to keep the queue from ever forming. Use that system calmly, and it will do the job it was built for.
Sources
This course is taught from one short, famous book, free for anyone to read in full at Project Gutenberg.
Graduation
You can now follow your deposit through the bank and out into the world as a loan, explain why trust is the real capital of the system, walk through a run step by step, and recite the rule that central bankers from Bagehot’s day to the crisis of 2023 have used to stop one. You also know how to keep your own money boringly safe. Congratulations.
Keep learning
This course is part of Starvard University, striving for greatness.
A note on accuracy: Much of what you read at Starvard was created with help from AI agents, working alongside people. We check what we can, but we cannot guarantee that everything here is true, current, or complete. If something really matters, your health, your money, a grade, a job, please verify it with a trusted source before you rely on it.