Guest Post by Milan Adams
The notification arrives at 4:47 on a Friday afternoon, from an app you rarely open. It says that one of the big banks, a name you have used for most of your adult life, has been seized by the government. You read it four times before the words settle. Regulators tend to move late like this, after the branches have closed and the staff have gone home, on the theory that a weekend between the announcement and Monday morning gives everyone room to breathe.
Your thumb goes to the banking app out of habit, the way it might reach for a wallet to check it is still there. A small circle turns, then turns again. A gray, polite message appears, one you have never seen before, something about high traffic and service interruptions. Nobody is shouting. No sirens, no man on television waving his arms. Just a spinning circle and one flat sentence, which together announce that your money has stopped answering the phone.
Over on the family group chat a second signal lands. Someone has screenshotted a balance. Someone else reports that a rent transfer is pending, a word that will do a lot of heavy lifting over the next three weeks. Then the cousin who is always slightly ahead of everyone types the line that will define the mood for millions of people at once: get cash tonight.
Multiply that scene by five million kitchens, cars and office bathrooms, because all of it is happening at the same hour. Banks do not fail the way houses catch fire, suddenly and in plain view. They fail slowly, from the inside, after years of small pressures nobody bothered to watch.
By the time the public learns the name, the decision is already made and the paperwork already signed. All that remains is to manage the panic the paperwork creates, and that part never makes it into the textbooks. The failure itself is technical. The panic it causes is another matter entirely.
Most people never quite absorb how much of modern life runs through a handful of buildings. Direct deposits, payroll, mortgage payments, card settlements, the automatic withdrawal for the electric bill, the small transfer that keeps a shop’s lights on one more week. None of it sits in a vault. It is a number in a ledger that one bank keeps and that other banks agree to believe. Let one ledger stop updating and the agreement wobbles, and the wobble travels outward at the speed of the internet.
At this level, money is mostly a belief system, and a very well advertised one. You have almost certainly never seen the actual dollars behind your checking account, because in any physical sense they do not exist. What exists is a promise, and a chain of other promises, and a shared willingness to keep acting as though the promises are solid. That willingness does all the work. When enough people stop believing at once, the whole thing coughs, and the cough is what we call a bank run.
Ordinary life runs on a margin of days, not months. A paycheck arrives, bills go out, groceries land on a card, and the buffer at the end stays thin. Seventy-two hours of frozen accounts is an annoyance. Two weeks is a crisis, because a mortgage does not care about your bank’s paperwork and a grocery store does not take a sympathetic shrug.
Underneath sits a darker layer, and it has little to do with whichever bank is on your phone. Since the spring of 2023, when three of the four largest failures in American history landed inside eight weeks and then the noise simply stopped, the system has been unusually quiet. It is comfortable, and it is also the kind of quiet in which complacency gets made. A system that has not been seriously tested in three years is full of people who have forgotten what a test feels like, and forgotten systems are fragile in ways that stay invisible until the moment they are not.
Whether a major bank can collapse is barely in doubt. Banks collapse all the time, dozens a decade, most of them small enough that the news never reaches you. The interesting case is the big one, connected to your payroll and your landlord’s mortgage and half the small businesses in your city. That scenario would play out over the first twelve hours, then the twelve days after, then a long gray year.
What follows is an honest attempt to walk through that scenario from the ground up, in the order it would actually unfold, using the real machinery that exists today and the real precedents we already survived. Parts of it have been rehearsed by the government, and parts by the banks. You have rehearsed almost none of it, and that is the gap worth closing.
Words break first, and they break on social media rather than television. Regulators seize banks late on a Friday precisely so a weekend can absorb the shock, but the modern world does not observe weekends. By the time the official statement lands, the rumor has been loose for hours, carried by screenshots and half-sentences and the particular tone financial people adopt when they are trying very hard to sound calm. Confidence is fragile, and the internet is very good at finding the thin spot.
The run itself is instant. In 1907 it meant a line of men in hats outside a marble lobby. In 2008 it meant customers queued on a California sidewalk with police keeping order, after IndyMac failed that July with thirty-two billion dollars in assets, and thousands of people learning in real time that a bank is only as strong as the crowd outside it. By 2023 the crowd never showed up at all. Forty-two billion dollars left Silicon Valley Bank in a single day, most of it from laptops, at a rate north of a million dollars a second, and the bank was dead by morning.
Keep that number in your head. Forty-two billion in a single day, with no lines, no lobby and no hats. A bank that had existed for four decades and held the money of half the technology industry was emptied in roughly the time it takes to watch two episodes of something. Branch staff learned about the run from the same headlines as everyone else.
Failure inside those first hours happens in a specific order, and the order tells you what to protect. Apps slow first, buckling under millions of simultaneous logins. Transfers between banks lag next, since the rails that move money between institutions were built for a jog, not a stampede. Then, if the damage is bad enough, some banks quietly cap daily outflows or freeze wires outright, buying time to count the wounds.
ATMs become the second front. Cash is physical, and physical things run out. A machine holding a few hundred thousand dollars can be drained in one anxious afternoon, and the vans that refill it run on a schedule built for a normal Tuesday. In a genuine panic the machines near the branch empty first, then those by the grocery stores, then the ones in neighborhoods where people already live paycheck to paycheck. Whoever has the least buffer reaches the machine last.
Branches do open, usually by Saturday or Monday, and lines do form, because some people need a human face and some businesses need a stamped document. Police have managed bank crowds before and would again. The crowd at the branch is only the visible edge. Most of the dying is happening somewhere else, in the app.
Keep an eye on the other banks during these hours, because contagion is rarely rational and does not wait for facts. Depositors at perfectly healthy institutions start moving money too, unable to tell a bank with a problem from a bank that merely resembles the one with a problem. Signature Bank failed two days after Silicon Valley Bank in 2023, and First Republic followed in May. None of the three had the same disease, only the same symptom: a market that had suddenly stopped believing them.
Small businesses feel it before households, and they feel it hardest. A company with forty employees and a checking account at the failed bank cannot run payroll from an account it cannot reach. Owners call their bankers, then their accountants, then each other, hunting for someone who can promise that Friday’s checks will clear. Nobody can promise anything, because the machinery that would make the promise is itself in receivership. By Sunday night the local economy has developed a stutter.
The psychological moment arrives late the first night, once the adrenaline fades and the arithmetic sets in. How much is in the account, how much is insured, how much sits over the line, how long the household could last on whatever is in the drawer. Millions of families ran the same calculation in the autumn of 2008 and again in the spring of 2023. It produces the same cold feeling every time: your entire financial life is a number on a screen maintained by strangers you will never meet.
The cavalry does exist, it arrives quickly, and its name is the Federal Deposit Insurance Corporation. Here the story turns genuinely reassuring and stays genuinely real. Since the agency was created in 1933, no depositor has lost a penny of insured money in an American bank failure, a record that has survived depressions, wars, the savings and loan collapse, the dot-com bust and three of the largest failures in history. Few promises in modern life have been kept that faithfully for that long.
Coverage is precise. The standard ceiling is two hundred and fifty thousand dollars per depositor, per bank, per ownership category, and the FDIC moves to honor it within a few business days, transferring accounts to an acquiring bank, issuing checks, or both. For an ordinary saver the practical experience is often a strange administrative blur followed by a working account at a different bank, with a different logo and a different app to download.
The line that matters is the one above the ceiling. Anything past the insured limit is technically exposed, and the mood shifts there. A claim against the failed bank’s remaining assets arrives, a piece of paper for the uninsured portion, and then you wait. Loans, buildings and securities get auctioned off slowly, and whatever is recovered is distributed over months or years in fractions nobody can forecast. Going over the limit rarely means losing everything. It usually means losing control of the timeline.
Regulators have tools beyond insurance and will use them. Emergency lending windows at the Federal Reserve can flood the system with liquidity, which is a fancy way of saying the Fed lends cash to banks so the cash machines keep working and the cards keep clearing. Within days in 2023 the Fed stood up an entirely new facility, the Bank Term Funding Program, aimed at stopping a local problem from becoming a national one. It worked well enough that the panic cooled within weeks. Expect the same reflexes next time, only faster, because the next time will be watched by a public that has already seen the movie.
Treasury and the FDIC can stretch the rules when the situation demands. In 2023 they invoked the systemic risk exception, a clause that let them protect uninsured depositors at the two failed banks, on the argument that letting those depositors take losses would have triggered a wider collapse. The exception is not automatic. Deposits above the insured limit are covered only when regulators judge the wider system to be at risk, which is why Janet Yellen cautioned at the time that the guarantee would not apply to every depositor above the limit in every future failure. She later framed the stakes precisely, describing “the decisive actions that we took in March to protect depositors and provide additional liquidity to the system,” and crediting them with mitigating “the very serious risk of broader financial contagion in the banking system.”
The FDIC’s own leadership used blunter words that weekend. Chairman Martin Gruenberg said the agencies were taking “decisive actions to protect the U.S. economy by strengthening public confidence in our banking system.” Public confidence, in his phrasing, is what matters most. The money is real. What keeps it from running is confidence, and confidence is something a government can manufacture in an emergency, much the way a fire crew builds a firebreak.
None of that makes the days feel calm. Outages happen anyway, transactions run late, a bank’s website stays a mess for a week, and officials say more than they mean and less than they should. Markets lurch, which ripples straight into retirement accounts and makes millions of people feel poorer on paper before anything real has happened to them. The stock market is not the economy, but it does show how people feel, and people will feel terrible.
Contained problems burn out, and the second act becomes a slow exhale. Confidence returns in increments, the app works, payroll clears, and people stop checking their balance four times a day. Most failures end this way, absorbed and forgotten, a paragraph in a financial history nobody reads. Quiet failures are the norm, and the ones that make the news are the exception. That exception is what we are imagining.
Let it spread and the second act turns into a chain. Money starts leaving banks that merely resemble the failed one, and the resemblance need not run deep. A similar size will do, or a similar industry focus, or a similar pile of uninsured corporate accounts. Contagion in 2023 traveled along a shared profile rather than a shared balance sheet, and it toppled two more banks before the weekend ended and a third within two months. One failure, under the wrong conditions, becomes a verdict on an entire category of institution.
Credit is where an ordinary person feels the second wave, often without connecting it to the bank at all. Survivors turn cautious, since caution is how you survive the next scare. Lending tightens. A mortgage you were about to get becomes harder to qualify for. A small business line of credit that kept your cousin’s restaurant afloat gets trimmed without warning. Credit card limits shrink, sometimes overnight, and a household that was managing fine discovers its available buffer was never as wide as the statement implied.
Jobs follow credit with a lag of weeks or months. When businesses cannot borrow, expansion stops, then hiring stops, then trimming begins. A bank failure does not guarantee a recession, but a bad one is among the most reliable recession starters we know, because it attacks the exact tissue joining savers to borrowers. The 2008 crisis began as a mortgage problem, became a banking problem, and ended as a ten percent unemployment problem, in that order. The order matters.
Other institutions absorb the refugees. Credit unions, insured through a parallel system up to the same two hundred and fifty thousand dollar limit, see a rush of new members during a scare, because people want somewhere that feels smaller and more human. Gold and silver dealers report sold-out inventories within days as savers reach for things no ledger controls. Those who had already spread their money, kept cash in the house and owned a few hard assets discover they are not frightened, and the difference between them and their neighbors has little to do with intelligence. It comes down to preparation, mostly.
There is a social dimension the economics papers skip. A banking crisis is a trust crisis, and trust is what societies run on when everything else fails. Neighborhood chats reorganize around who has cash and who can front a bill. Landlords get asked for a week’s grace and sometimes give it, because the landlord’s own account is frozen too. The informal economy thickens, and people rediscover that a favor owed is a form of savings.
How quickly it can end is the odd part. Once the guarantees are believed, once the acquiring banks are named, once the first Monday passes without catastrophe, the herd turns and walks back in. People who emptied accounts on Friday are redepositing a month later, sometimes into the very institution that scared them. Panics have a short half-life when managed well and a long one when they are not, and the distance between those outcomes is usually a matter of hours and a matter of words.
History is the only laboratory available, and it has run this experiment twice in living memory with two very different results. The 2008 crisis was long and deep, a mortgage contagion that metastasized over eighteen months, met by a response that was enormous, controversial and slow to arrive. The 2023 crisis was short and sharp, a deposit contagion that burned through three banks in two months and then went out, met by a response that was nearly instantaneous. Speed, more than anything else, decided how each one ended.
Speed matters because panic compounds. A run allowed to continue for a week becomes a run that cannot be stopped, because by then the rumor has hardened into a belief, and beliefs about money are self-fulfilling. Washington Mutual, the largest failure in American history at three hundred and seven billion dollars in assets, spent most of 2008 sliding, leaking deposits and losing the argument, and by the time the government seized it in late September the institution was already hollow. It had been dying for months. Silicon Valley Bank, by contrast, was healthy on a Wednesday and dead by Friday morning, because forty-two billion dollars left in one day and nothing had time to be gradual.
The second lesson is uglier and gets discussed less. The 2008 playbook spent taxpayer money to rescue banks, the public hated it, and the hatred produced a political backlash that reshaped a decade. Out of that came a newer tool, the bail-in. Rather than the government writing a check, the bank’s own large depositors and bondholders absorb losses, their claims converted into shares or written down so the bank can be recapitalized from inside. The template dates to Cyprus in 2013, where large depositors in two failing banks had a chunk of their uninsured money frozen and partly converted into equity.
A bail-in changes the arithmetic for anyone holding real money in one institution. Above the insured limit you stop being only a customer and become a creditor, and creditors take haircuts. The money does not necessarily vanish. It gets restructured, converted, frozen, repriced, and handed back later in a form you did not choose. For a wealthy saver, the bail-in is the quiet nightmare tucked inside the word rescue, because it means the rescued help pay for the rescue.
Two things stay constant across both eras. Insured deposits have always been protected, in every crisis, without exception, and that protection is not a courtesy withdrawn when things get bad. The surrounding economy enjoys no such cover. Jobs, credit, prices, rents, the odds of getting a car loan or a mortgage all sit outside the net, and all of them can be damaged even when your account is perfectly safe.
Jamie Dimon, who runs the largest bank in the country and has watched every crisis since the savings and loan era, wrote a line in his 2023 shareholder letter that has aged into prophecy. “The current crisis is not yet over,” he wrote, “and even when it is behind us, there will be repercussions from it for years to come.” He was right in the narrow sense, and the narrower truth is that the repercussions of any major failure outlive the headlines by a decade, surfacing in tighter credit, slower growth and a generation of savers who trust institutions a little less.
Recent years have been almost eerily calm, which is its own kind of data. Two failures in 2024, two in 2025, and two more so far this year, all of them small institutions with a few hundred million in assets at most, absorbed and forgotten within a week. The big names have held. Quiet, as any veteran of finance will tell you with a slightly nervous smile, is exactly the condition in which the next problem quietly assembles itself.
Preparation in banking is not complicated. It is boring, it is cheap, and it is the difference between watching a crisis and living through one. You are not trying to predict the next failure. You are trying to be indifferent to it, so that whichever bank goes down, and whenever, the event is an inconvenience rather than a catastrophe.
Spread the money across institutions. No single account should carry more than the two hundred and fifty thousand dollar insured limit, and if household savings exceed that, the excess belongs in a second bank, a third bank, or a different ownership category at the same bank. Joint accounts, retirement accounts and trust accounts each carry separate coverage, which means a married couple can legitimately protect far more than a quarter of a million at one institution by using the right account types. The FDIC publishes the rules in plain language, and an hour spent reading them beats a year of financial news.
Use both big banks and small ones. Size is no guarantee of safety, and smallness is no guarantee of virtue, but the two behave differently under stress. Giants tend to get rescued because letting them fall breaks too much. Small local banks tend to be more conservative, closer to their borrowers, less exposed to the speculative fashions that sink large institutions. Splitting money between the two types buys the political protection of the big and the prudence of the small, and it costs almost nothing to arrange.
Keep real cash in the house. A few weeks of expenses, in small bills, somewhere dry and boring. Small bills matter because in a crisis nobody can break a hundred, and the person selling you bread at the corner will be grateful for a five. Cash is the only money that keeps working when the power flickers, the network drops and the app refuses to load. It earns no interest, so treat it as insurance you hope never to use rather than as savings.
Own a little precious metal. Gold and silver sit outside every banking system on earth, which is exactly why they get bought in a panic. You do not need to become a coin collector. A modest holding of silver, cheap enough to be practical and divisible enough to be useful, gives you an asset no ledger controls and no regulator can freeze. It pays nothing and can sit flat for years, so it belongs in the same category as the cash. Learn the spot price now, while you can be calm about it, so that you are not guessing at value in a crowd.
Put some wealth into things you can touch. Land, tools, a generator, a chest freezer, decent hand equipment, a bicycle that works. Tangible assets cannot be devalued by a central bank or frozen by a receiver, and several of them improve an ordinary Tuesday. You are not trying to become a homesteader. The idea is to hold a slice of your net worth that does not depend on an institution staying solvent.
Watch the health of your own bank. Stock price slides, repeated quarterly losses, a sudden change of chief executive, a heavy concentration in one industry, an unusually high share of uninsured corporate deposits. These are the warning lights, and they usually blink for months before anything breaks. Read your bank’s annual report once a year, or at least skim the headlines. The customers who got hurt in 2023 were mostly the ones who never asked whether their bank was doing anything risky.
Set up your backups before you need them. Open the second account now, not on the Friday afternoon when the app is dead. Verify the transfer links now, not during the panic. Keep a written list of account numbers somewhere physical, because the app that stores them may be the app that is broken. Preparation is cheap when nothing is wrong and worth a great deal when something is.
The card in your wallet is a small miracle that stops working the moment the chain behind it breaks. Every swipe is a message traveling from a terminal to a bank to a network and back, and any link can go dark. Alternatives are not paranoia. They are basic financial hygiene, and they cost almost nothing to build.
All of these options rest on the same principle, which is redundancy. One payment method is a single point of failure. Three or four overlapping methods mean that whatever breaks, something else still works, and the difference between a bad week and a disaster is usually just that.
There is a version of a banking crisis where the shelves are fine and the trucks keep running, and a version where the disruption lasts long enough that ordinary goods become the thing people trade. Which one you get is unknowable in advance, so you prepare for the second and enjoy the first. A stockpile is not really about hoarding, but about having enough surplus to trade without touching your own essentials.
Two rules govern the whole thing. Only trade what you can spare, so a bad trade never leaves you short of something you need. And never give away your last of anything, because the last of anything is the most expensive version of it. Build the network now, while everyone is calm and generous, so that when the pressure arrives you are trading from strength rather than desperation.
A major bank collapsing tomorrow would not end the world, and it would not be gentle. It would be fast, ugly and deeply personal for millions of people, most of whom did nothing wrong and simply kept their money where their parents kept theirs. Confusion in the first twelve hours. Paperwork and rumor in the first twelve days. Tighter credit, slower hiring, and a quieter, more cautious economy over the following year that nobody could quite explain.
The reassuring part is that the containment machinery is real, tested and fast, and it has an almost perfect record on the one promise an ordinary household cares about most. Insured money has always come back. Since 1933, without exception, across every disaster the country has survived, no depositor has lost a single insured dollar. That record is not an accident. It comes from institutions built to hold the line, and they would hold it again.
Everything insurance does not cover is where the discomfort lies. Your job, your rent, the loan you were planning to take, the price of what you buy, the general mood of the economy around you. Those carry no FDIC sticker, and they are where a bank failure does its lasting damage. Protecting them means protecting your flexibility, which means not having all your money in one place, not having all your options in one system, and not being the person who finds the flaw in the plan on the worst possible day.
Preparing for this is mostly a matter of being boring in advance. A second bank account, some cash in a drawer, a few ounces of silver, a stocked pantry, a neighbor who owes you a favor and to whom you owe one back. None of it is dramatic or expensive, and together it turns a potential catastrophe into an inconvenience you can shrug off over breakfast.
Sooner or later there will be another failure. Which name, and how many hours the app spins before someone tells you the truth, are the only parts still unknown. The people who come through it well will not be the smartest or the luckiest. They will be the ones who, on some ordinary quiet weekend long before the news broke, bothered to move a little money somewhere else and buy a little cash they never thought they would need.
Tweet
Continue reading...
[ H/T The Burning Platform ]
The notification arrives at 4:47 on a Friday afternoon, from an app you rarely open. It says that one of the big banks, a name you have used for most of your adult life, has been seized by the government. You read it four times before the words settle. Regulators tend to move late like this, after the branches have closed and the staff have gone home, on the theory that a weekend between the announcement and Monday morning gives everyone room to breathe.
Your thumb goes to the banking app out of habit, the way it might reach for a wallet to check it is still there. A small circle turns, then turns again. A gray, polite message appears, one you have never seen before, something about high traffic and service interruptions. Nobody is shouting. No sirens, no man on television waving his arms. Just a spinning circle and one flat sentence, which together announce that your money has stopped answering the phone.
Over on the family group chat a second signal lands. Someone has screenshotted a balance. Someone else reports that a rent transfer is pending, a word that will do a lot of heavy lifting over the next three weeks. Then the cousin who is always slightly ahead of everyone types the line that will define the mood for millions of people at once: get cash tonight.
Multiply that scene by five million kitchens, cars and office bathrooms, because all of it is happening at the same hour. Banks do not fail the way houses catch fire, suddenly and in plain view. They fail slowly, from the inside, after years of small pressures nobody bothered to watch.
By the time the public learns the name, the decision is already made and the paperwork already signed. All that remains is to manage the panic the paperwork creates, and that part never makes it into the textbooks. The failure itself is technical. The panic it causes is another matter entirely.
Most people never quite absorb how much of modern life runs through a handful of buildings. Direct deposits, payroll, mortgage payments, card settlements, the automatic withdrawal for the electric bill, the small transfer that keeps a shop’s lights on one more week. None of it sits in a vault. It is a number in a ledger that one bank keeps and that other banks agree to believe. Let one ledger stop updating and the agreement wobbles, and the wobble travels outward at the speed of the internet.
At this level, money is mostly a belief system, and a very well advertised one. You have almost certainly never seen the actual dollars behind your checking account, because in any physical sense they do not exist. What exists is a promise, and a chain of other promises, and a shared willingness to keep acting as though the promises are solid. That willingness does all the work. When enough people stop believing at once, the whole thing coughs, and the cough is what we call a bank run.
Ordinary life runs on a margin of days, not months. A paycheck arrives, bills go out, groceries land on a card, and the buffer at the end stays thin. Seventy-two hours of frozen accounts is an annoyance. Two weeks is a crisis, because a mortgage does not care about your bank’s paperwork and a grocery store does not take a sympathetic shrug.
Underneath sits a darker layer, and it has little to do with whichever bank is on your phone. Since the spring of 2023, when three of the four largest failures in American history landed inside eight weeks and then the noise simply stopped, the system has been unusually quiet. It is comfortable, and it is also the kind of quiet in which complacency gets made. A system that has not been seriously tested in three years is full of people who have forgotten what a test feels like, and forgotten systems are fragile in ways that stay invisible until the moment they are not.
Whether a major bank can collapse is barely in doubt. Banks collapse all the time, dozens a decade, most of them small enough that the news never reaches you. The interesting case is the big one, connected to your payroll and your landlord’s mortgage and half the small businesses in your city. That scenario would play out over the first twelve hours, then the twelve days after, then a long gray year.
What follows is an honest attempt to walk through that scenario from the ground up, in the order it would actually unfold, using the real machinery that exists today and the real precedents we already survived. Parts of it have been rehearsed by the government, and parts by the banks. You have rehearsed almost none of it, and that is the gap worth closing.
The First Twelve Hours, When the Screens Go Quiet
Words break first, and they break on social media rather than television. Regulators seize banks late on a Friday precisely so a weekend can absorb the shock, but the modern world does not observe weekends. By the time the official statement lands, the rumor has been loose for hours, carried by screenshots and half-sentences and the particular tone financial people adopt when they are trying very hard to sound calm. Confidence is fragile, and the internet is very good at finding the thin spot.
The run itself is instant. In 1907 it meant a line of men in hats outside a marble lobby. In 2008 it meant customers queued on a California sidewalk with police keeping order, after IndyMac failed that July with thirty-two billion dollars in assets, and thousands of people learning in real time that a bank is only as strong as the crowd outside it. By 2023 the crowd never showed up at all. Forty-two billion dollars left Silicon Valley Bank in a single day, most of it from laptops, at a rate north of a million dollars a second, and the bank was dead by morning.
Keep that number in your head. Forty-two billion in a single day, with no lines, no lobby and no hats. A bank that had existed for four decades and held the money of half the technology industry was emptied in roughly the time it takes to watch two episodes of something. Branch staff learned about the run from the same headlines as everyone else.
Failure inside those first hours happens in a specific order, and the order tells you what to protect. Apps slow first, buckling under millions of simultaneous logins. Transfers between banks lag next, since the rails that move money between institutions were built for a jog, not a stampede. Then, if the damage is bad enough, some banks quietly cap daily outflows or freeze wires outright, buying time to count the wounds.
ATMs become the second front. Cash is physical, and physical things run out. A machine holding a few hundred thousand dollars can be drained in one anxious afternoon, and the vans that refill it run on a schedule built for a normal Tuesday. In a genuine panic the machines near the branch empty first, then those by the grocery stores, then the ones in neighborhoods where people already live paycheck to paycheck. Whoever has the least buffer reaches the machine last.
Branches do open, usually by Saturday or Monday, and lines do form, because some people need a human face and some businesses need a stamped document. Police have managed bank crowds before and would again. The crowd at the branch is only the visible edge. Most of the dying is happening somewhere else, in the app.
Keep an eye on the other banks during these hours, because contagion is rarely rational and does not wait for facts. Depositors at perfectly healthy institutions start moving money too, unable to tell a bank with a problem from a bank that merely resembles the one with a problem. Signature Bank failed two days after Silicon Valley Bank in 2023, and First Republic followed in May. None of the three had the same disease, only the same symptom: a market that had suddenly stopped believing them.
Small businesses feel it before households, and they feel it hardest. A company with forty employees and a checking account at the failed bank cannot run payroll from an account it cannot reach. Owners call their bankers, then their accountants, then each other, hunting for someone who can promise that Friday’s checks will clear. Nobody can promise anything, because the machinery that would make the promise is itself in receivership. By Sunday night the local economy has developed a stutter.
The psychological moment arrives late the first night, once the adrenaline fades and the arithmetic sets in. How much is in the account, how much is insured, how much sits over the line, how long the household could last on whatever is in the drawer. Millions of families ran the same calculation in the autumn of 2008 and again in the spring of 2023. It produces the same cold feeling every time: your entire financial life is a number on a screen maintained by strangers you will never meet.
When the Government Reaches In
The cavalry does exist, it arrives quickly, and its name is the Federal Deposit Insurance Corporation. Here the story turns genuinely reassuring and stays genuinely real. Since the agency was created in 1933, no depositor has lost a penny of insured money in an American bank failure, a record that has survived depressions, wars, the savings and loan collapse, the dot-com bust and three of the largest failures in history. Few promises in modern life have been kept that faithfully for that long.
Coverage is precise. The standard ceiling is two hundred and fifty thousand dollars per depositor, per bank, per ownership category, and the FDIC moves to honor it within a few business days, transferring accounts to an acquiring bank, issuing checks, or both. For an ordinary saver the practical experience is often a strange administrative blur followed by a working account at a different bank, with a different logo and a different app to download.
The line that matters is the one above the ceiling. Anything past the insured limit is technically exposed, and the mood shifts there. A claim against the failed bank’s remaining assets arrives, a piece of paper for the uninsured portion, and then you wait. Loans, buildings and securities get auctioned off slowly, and whatever is recovered is distributed over months or years in fractions nobody can forecast. Going over the limit rarely means losing everything. It usually means losing control of the timeline.
Regulators have tools beyond insurance and will use them. Emergency lending windows at the Federal Reserve can flood the system with liquidity, which is a fancy way of saying the Fed lends cash to banks so the cash machines keep working and the cards keep clearing. Within days in 2023 the Fed stood up an entirely new facility, the Bank Term Funding Program, aimed at stopping a local problem from becoming a national one. It worked well enough that the panic cooled within weeks. Expect the same reflexes next time, only faster, because the next time will be watched by a public that has already seen the movie.
Treasury and the FDIC can stretch the rules when the situation demands. In 2023 they invoked the systemic risk exception, a clause that let them protect uninsured depositors at the two failed banks, on the argument that letting those depositors take losses would have triggered a wider collapse. The exception is not automatic. Deposits above the insured limit are covered only when regulators judge the wider system to be at risk, which is why Janet Yellen cautioned at the time that the guarantee would not apply to every depositor above the limit in every future failure. She later framed the stakes precisely, describing “the decisive actions that we took in March to protect depositors and provide additional liquidity to the system,” and crediting them with mitigating “the very serious risk of broader financial contagion in the banking system.”
The FDIC’s own leadership used blunter words that weekend. Chairman Martin Gruenberg said the agencies were taking “decisive actions to protect the U.S. economy by strengthening public confidence in our banking system.” Public confidence, in his phrasing, is what matters most. The money is real. What keeps it from running is confidence, and confidence is something a government can manufacture in an emergency, much the way a fire crew builds a firebreak.
None of that makes the days feel calm. Outages happen anyway, transactions run late, a bank’s website stays a mess for a week, and officials say more than they mean and less than they should. Markets lurch, which ripples straight into retirement accounts and makes millions of people feel poorer on paper before anything real has happened to them. The stock market is not the economy, but it does show how people feel, and people will feel terrible.
The Weeks After, Where the Real Damage Spreads
Contained problems burn out, and the second act becomes a slow exhale. Confidence returns in increments, the app works, payroll clears, and people stop checking their balance four times a day. Most failures end this way, absorbed and forgotten, a paragraph in a financial history nobody reads. Quiet failures are the norm, and the ones that make the news are the exception. That exception is what we are imagining.
Let it spread and the second act turns into a chain. Money starts leaving banks that merely resemble the failed one, and the resemblance need not run deep. A similar size will do, or a similar industry focus, or a similar pile of uninsured corporate accounts. Contagion in 2023 traveled along a shared profile rather than a shared balance sheet, and it toppled two more banks before the weekend ended and a third within two months. One failure, under the wrong conditions, becomes a verdict on an entire category of institution.
Credit is where an ordinary person feels the second wave, often without connecting it to the bank at all. Survivors turn cautious, since caution is how you survive the next scare. Lending tightens. A mortgage you were about to get becomes harder to qualify for. A small business line of credit that kept your cousin’s restaurant afloat gets trimmed without warning. Credit card limits shrink, sometimes overnight, and a household that was managing fine discovers its available buffer was never as wide as the statement implied.
Jobs follow credit with a lag of weeks or months. When businesses cannot borrow, expansion stops, then hiring stops, then trimming begins. A bank failure does not guarantee a recession, but a bad one is among the most reliable recession starters we know, because it attacks the exact tissue joining savers to borrowers. The 2008 crisis began as a mortgage problem, became a banking problem, and ended as a ten percent unemployment problem, in that order. The order matters.
Other institutions absorb the refugees. Credit unions, insured through a parallel system up to the same two hundred and fifty thousand dollar limit, see a rush of new members during a scare, because people want somewhere that feels smaller and more human. Gold and silver dealers report sold-out inventories within days as savers reach for things no ledger controls. Those who had already spread their money, kept cash in the house and owned a few hard assets discover they are not frightened, and the difference between them and their neighbors has little to do with intelligence. It comes down to preparation, mostly.
There is a social dimension the economics papers skip. A banking crisis is a trust crisis, and trust is what societies run on when everything else fails. Neighborhood chats reorganize around who has cash and who can front a bill. Landlords get asked for a week’s grace and sometimes give it, because the landlord’s own account is frozen too. The informal economy thickens, and people rediscover that a favor owed is a form of savings.
How quickly it can end is the odd part. Once the guarantees are believed, once the acquiring banks are named, once the first Monday passes without catastrophe, the herd turns and walks back in. People who emptied accounts on Friday are redepositing a month later, sometimes into the very institution that scared them. Panics have a short half-life when managed well and a long one when they are not, and the distance between those outcomes is usually a matter of hours and a matter of words.
What 2008 and 2023 Already Taught Us
History is the only laboratory available, and it has run this experiment twice in living memory with two very different results. The 2008 crisis was long and deep, a mortgage contagion that metastasized over eighteen months, met by a response that was enormous, controversial and slow to arrive. The 2023 crisis was short and sharp, a deposit contagion that burned through three banks in two months and then went out, met by a response that was nearly instantaneous. Speed, more than anything else, decided how each one ended.
Speed matters because panic compounds. A run allowed to continue for a week becomes a run that cannot be stopped, because by then the rumor has hardened into a belief, and beliefs about money are self-fulfilling. Washington Mutual, the largest failure in American history at three hundred and seven billion dollars in assets, spent most of 2008 sliding, leaking deposits and losing the argument, and by the time the government seized it in late September the institution was already hollow. It had been dying for months. Silicon Valley Bank, by contrast, was healthy on a Wednesday and dead by Friday morning, because forty-two billion dollars left in one day and nothing had time to be gradual.
The second lesson is uglier and gets discussed less. The 2008 playbook spent taxpayer money to rescue banks, the public hated it, and the hatred produced a political backlash that reshaped a decade. Out of that came a newer tool, the bail-in. Rather than the government writing a check, the bank’s own large depositors and bondholders absorb losses, their claims converted into shares or written down so the bank can be recapitalized from inside. The template dates to Cyprus in 2013, where large depositors in two failing banks had a chunk of their uninsured money frozen and partly converted into equity.
A bail-in changes the arithmetic for anyone holding real money in one institution. Above the insured limit you stop being only a customer and become a creditor, and creditors take haircuts. The money does not necessarily vanish. It gets restructured, converted, frozen, repriced, and handed back later in a form you did not choose. For a wealthy saver, the bail-in is the quiet nightmare tucked inside the word rescue, because it means the rescued help pay for the rescue.
Two things stay constant across both eras. Insured deposits have always been protected, in every crisis, without exception, and that protection is not a courtesy withdrawn when things get bad. The surrounding economy enjoys no such cover. Jobs, credit, prices, rents, the odds of getting a car loan or a mortgage all sit outside the net, and all of them can be damaged even when your account is perfectly safe.
Jamie Dimon, who runs the largest bank in the country and has watched every crisis since the savings and loan era, wrote a line in his 2023 shareholder letter that has aged into prophecy. “The current crisis is not yet over,” he wrote, “and even when it is behind us, there will be repercussions from it for years to come.” He was right in the narrow sense, and the narrower truth is that the repercussions of any major failure outlive the headlines by a decade, surfacing in tighter credit, slower growth and a generation of savers who trust institutions a little less.
Recent years have been almost eerily calm, which is its own kind of data. Two failures in 2024, two in 2025, and two more so far this year, all of them small institutions with a few hundred million in assets at most, absorbed and forgotten within a week. The big names have held. Quiet, as any veteran of finance will tell you with a slightly nervous smile, is exactly the condition in which the next problem quietly assembles itself.
How to Armor Your Own Savings Before Monday
Preparation in banking is not complicated. It is boring, it is cheap, and it is the difference between watching a crisis and living through one. You are not trying to predict the next failure. You are trying to be indifferent to it, so that whichever bank goes down, and whenever, the event is an inconvenience rather than a catastrophe.
Spread the money across institutions. No single account should carry more than the two hundred and fifty thousand dollar insured limit, and if household savings exceed that, the excess belongs in a second bank, a third bank, or a different ownership category at the same bank. Joint accounts, retirement accounts and trust accounts each carry separate coverage, which means a married couple can legitimately protect far more than a quarter of a million at one institution by using the right account types. The FDIC publishes the rules in plain language, and an hour spent reading them beats a year of financial news.
Use both big banks and small ones. Size is no guarantee of safety, and smallness is no guarantee of virtue, but the two behave differently under stress. Giants tend to get rescued because letting them fall breaks too much. Small local banks tend to be more conservative, closer to their borrowers, less exposed to the speculative fashions that sink large institutions. Splitting money between the two types buys the political protection of the big and the prudence of the small, and it costs almost nothing to arrange.
Keep real cash in the house. A few weeks of expenses, in small bills, somewhere dry and boring. Small bills matter because in a crisis nobody can break a hundred, and the person selling you bread at the corner will be grateful for a five. Cash is the only money that keeps working when the power flickers, the network drops and the app refuses to load. It earns no interest, so treat it as insurance you hope never to use rather than as savings.
Own a little precious metal. Gold and silver sit outside every banking system on earth, which is exactly why they get bought in a panic. You do not need to become a coin collector. A modest holding of silver, cheap enough to be practical and divisible enough to be useful, gives you an asset no ledger controls and no regulator can freeze. It pays nothing and can sit flat for years, so it belongs in the same category as the cash. Learn the spot price now, while you can be calm about it, so that you are not guessing at value in a crowd.
Put some wealth into things you can touch. Land, tools, a generator, a chest freezer, decent hand equipment, a bicycle that works. Tangible assets cannot be devalued by a central bank or frozen by a receiver, and several of them improve an ordinary Tuesday. You are not trying to become a homesteader. The idea is to hold a slice of your net worth that does not depend on an institution staying solvent.
Watch the health of your own bank. Stock price slides, repeated quarterly losses, a sudden change of chief executive, a heavy concentration in one industry, an unusually high share of uninsured corporate deposits. These are the warning lights, and they usually blink for months before anything breaks. Read your bank’s annual report once a year, or at least skim the headlines. The customers who got hurt in 2023 were mostly the ones who never asked whether their bank was doing anything risky.
Set up your backups before you need them. Open the second account now, not on the Friday afternoon when the app is dead. Verify the transfer links now, not during the panic. Keep a written list of account numbers somewhere physical, because the app that stores them may be the app that is broken. Preparation is cheap when nothing is wrong and worth a great deal when something is.
Paying for Things When the Card Machines Die
The card in your wallet is a small miracle that stops working the moment the chain behind it breaks. Every swipe is a message traveling from a terminal to a bank to a network and back, and any link can go dark. Alternatives are not paranoia. They are basic financial hygiene, and they cost almost nothing to build.
- Cash, in small denominations. This is the foundation and nothing substitutes for it. Fifties and twenties are hard to break in a crisis, so stack fives, tens and singles. Keep enough for groceries, fuel and a week of small emergencies, and keep it in more than one hiding spot in case of fire, flood, or a very determined teenager.
- Peer-to-peer apps, as a bridge. PayPal, Venmo and Cash App can keep working while the internet is up, because they do not depend on your bank’s app being functional. Keep small balances across more than one, and set the accounts up in advance, since verifying a new account during a panic is miserable. Do not treat them as a bank. Treat them as a pipe that might still flow when your bank’s pipe is clogged.
- Precious metals for larger trades. Silver and gold carry value everyone recognizes, and inside the community of people who prepare for these things they function as money. Learn the weight and purity of what you hold, and learn roughly what it buys, so you are not the person at the swap meet getting quietly fleeced because you cannot tell whether an ounce is worth fifty or five hundred.
- A barter network of actual humans. Least glamorous, most powerful. Know your neighbors. Know who fixes things, who grows things, who can drive, who owns a truck, who can watch children, who can cook for a crowd. When money stutters, the people around you become the real infrastructure, and relationships built in calm weather become the currency you spend in a bad one. Traditional-skills books cataloguing old household knowledge, the goods and crafts that carried communities through hard decades, turn into genuinely useful reference material.
- Prepaid and gift cards, used carefully. A card to a major retailer or fuel chain is a crude but real store of value, and it can bridge a gap while the system sorts itself out. Know the limits. They are not insured, they can be devalued by the retailer’s own troubles, and fees and inflation eat them. Keep a small stack as a bridge, never as a vault.
All of these options rest on the same principle, which is redundancy. One payment method is a single point of failure. Three or four overlapping methods mean that whatever breaks, something else still works, and the difference between a bad week and a disaster is usually just that.
The Stockpile That Becomes a Currency
There is a version of a banking crisis where the shelves are fine and the trucks keep running, and a version where the disruption lasts long enough that ordinary goods become the thing people trade. Which one you get is unknowable in advance, so you prepare for the second and enjoy the first. A stockpile is not really about hoarding, but about having enough surplus to trade without touching your own essentials.
- Food and water. Canned goods, rice, dried beans, pasta, coffee, salt, and a serious water filter. Coffee earns its own line because it is the most reliable morale good in any crisis, and a world without it is a world people will pay to escape.
- Medical supplies. Over-the-counter painkillers, antiseptics, bandages, blister care, oral rehydration salts, and enough knowledge to use them. A good first-aid reference book beats a fancy kit you do not understand.
- Ammunition and defensive basics. Common calibers trade well because everyone has a weapon that takes them. Buy boring, popular sizes, and buy a little more than you shoot.
- Fuel and light. Propane cylinders, a small camp stove, batteries in every size, candles, matches, a solar lantern or two. Light and heat are the first things people miss and the first things they will trade for.
- Hygiene and household goods. Toilet paper, soap, toothpaste, feminine products, laundry detergent, trash bags. Unglamorous, universally needed, quietly valuable.
- Tools and practical skills. Hand tools, duct tape, rope, a good knife, fishing gear, sewing supplies, and the ability to actually use them. Someone who can repair things becomes valuable in a way money cannot buy.
- Comfort and small luxuries. Alcohol, tobacco, chocolate, coffee, playing cards, books. These make a hard month bearable, and they trade at a premium precisely because they are not strictly necessary.
Two rules govern the whole thing. Only trade what you can spare, so a bad trade never leaves you short of something you need. And never give away your last of anything, because the last of anything is the most expensive version of it. Build the network now, while everyone is calm and generous, so that when the pressure arrives you are trading from strength rather than desperation.
The Long Shadow, and the Case for Calm
A major bank collapsing tomorrow would not end the world, and it would not be gentle. It would be fast, ugly and deeply personal for millions of people, most of whom did nothing wrong and simply kept their money where their parents kept theirs. Confusion in the first twelve hours. Paperwork and rumor in the first twelve days. Tighter credit, slower hiring, and a quieter, more cautious economy over the following year that nobody could quite explain.
The reassuring part is that the containment machinery is real, tested and fast, and it has an almost perfect record on the one promise an ordinary household cares about most. Insured money has always come back. Since 1933, without exception, across every disaster the country has survived, no depositor has lost a single insured dollar. That record is not an accident. It comes from institutions built to hold the line, and they would hold it again.
Everything insurance does not cover is where the discomfort lies. Your job, your rent, the loan you were planning to take, the price of what you buy, the general mood of the economy around you. Those carry no FDIC sticker, and they are where a bank failure does its lasting damage. Protecting them means protecting your flexibility, which means not having all your money in one place, not having all your options in one system, and not being the person who finds the flaw in the plan on the worst possible day.
Preparing for this is mostly a matter of being boring in advance. A second bank account, some cash in a drawer, a few ounces of silver, a stocked pantry, a neighbor who owes you a favor and to whom you owe one back. None of it is dramatic or expensive, and together it turns a potential catastrophe into an inconvenience you can shrug off over breakfast.
Sooner or later there will be another failure. Which name, and how many hours the app spins before someone tells you the truth, are the only parts still unknown. The people who come through it well will not be the smartest or the luckiest. They will be the ones who, on some ordinary quiet weekend long before the news broke, bothered to move a little money somewhere else and buy a little cash they never thought they would need.
Tweet
Continue reading...
[ H/T The Burning Platform ]
:max_bytes(150000):strip_icc()/GettyImages-2030238083-7e6963ad50074c1b89e0d0930cc71d91.jpg)
