When it breaks: crises and contagion

One mortgage on one street goes bad. A family that could pay last year can’t pay this year, and the loan they took out stops being paid back. That sounds like a small, private misfortune — their problem, their bank, their street. Follow it for a few hops, though, and the loss walks into the life of someone who never met that family, never lived on that street, and never made a single bad bet. To see how, you have to stop looking at the parts and look at the wiring between them.

The economy is a web, not a row of boxes

It is tempting to picture the economy as a set of separate boxes: this household, that bank, this firm, that country. Each minds its own business. If one box has a bad year, the others carry on.

That picture is wrong, and the wrongness is the whole lesson. The boxes are tied together. Banks don’t just hold your money — they lend it to other banks and borrow from them. Firms don’t stand alone — each buys from suppliers and sells to customers, so every firm is a link in a chain of firms. People hold each other’s debts: your savings are somebody else’s loan. And underneath all of it runs the thing from item 2 — money is trust. Every link is a promise that the other side will pay.

Those links are what make the economy work. They let a swap (item 1) happen between strangers across oceans. But a link carries a loss as easily as a payment. The same wiring that makes the system efficient lets it spread a failure. You cannot have one without the other.

How a loss travels: direct exposure

The simplest way damage spreads is the most obvious once you see it: A is owed money by B, and B can’t pay.

Your bank lent money to a borrower. The borrower defaults — stops paying. That isn’t only the borrower’s loss. It is now the bank’s loss too, because the bank was counting on that money coming back. The bank is exposed to the borrower. When the borrower fails, the failure becomes the bank’s problem.

And the bank itself owes money — it borrowed from other banks. If enough of its borrowers default, it can’t pay the banks it owes, so now they take the loss. The failure hops along the wires of who-owes-whom. Each node passes its trouble to whoever was counting on it.

How a loss travels: the fire-sale

There is a second, sneakier route. Suppose a wobbling bank needs cash fast, so it sells the assets it holds — say, bundles of loans, or shares. To raise money quickly, it sells at whatever price it can get. A cheap, hurried sale is a fire-sale.

Here is the trap. The price of that asset is set by what it last sold for. When one holder dumps it cheap, the asset is now worth less — for everyone who holds it, not just the seller. A second bank that made no bad loans at all suddenly finds its holdings worth less on paper. If that loss makes it look shaky, it sells too — pushing the price down further, hitting a third. Everyone rushing for the same exit drops the floor out from under all of them. No direct loan connected them. The shared asset did.

How a loss travels: the run

The third route is the fastest, because it runs on fear. A bank only works because not everyone asks for their money at once. It lends most of the deposits out; it keeps a little on hand. That’s fine — as long as people trust it.

Now imagine a rumour: this bank is in trouble. Each saver thinks the same sensible thought — “if it’s failing, I want my money out first.” So everyone rushes to withdraw at once. But the bank lent the money out; it can’t return it all today. So it fails — because everyone feared it would. The fear made itself come true. This is a run, and it doesn’t need the bank to have actually been broken. Trust cracking is enough. Once savers have watched one bank fall, they look at the next and wonder. Fear spreads faster than facts.

A worked example: from one street to a stranger

Walk the hops, one node at a time.

1. The shock. Borrower B can’t repay a loan — say £1 million.
2. Direct exposure. Bank A lent B that money. A is now down £1 million it was counting on.
3. A pulls back. Shaken and short of cash, A stops making new loans to stay safe.
4. A firm is starved. Firm C had a loan from A coming up for renewal. A won’t renew it. C, perfectly healthy yesterday, now can’t fund its operations.
5. Workers lose jobs. C cuts costs and lays off staff. D, a worker at C, loses her wage — and she never borrowed a penny.
6. D stops spending. With no wage, D stops buying. The café E where she spent every morning loses a regular.
7. E’s supplier feels it. E orders less bread. The bakery F, an ocean away from B’s street, sells less, earns less, and lets someone go.

F never heard of B. F made no bad bet. Yet B’s bad mortgage reached F through six links F couldn’t see. The loss didn’t teleport — it travelled, hop by hop, down wires that were there all along.

The whole, and the humble part

Notice what makes this so hard to stop: no single seat can see the whole web. Bank A saw B. It did not see F. D saw her own lost job, not the mortgage six hops upstream that caused it. Each node sees only its neighbours. So the spread surprises almost everyone — not because they were foolish, but because the map is too big for any one person to hold.

That is the Protocol, stated plainly. The world is one connected system. The links that feed us are the same links that can carry harm to us, from places we’d swear were unconnected. And we are not standing above the web inspecting it — we are knots inside it, exposed through wires we never chose and mostly can’t see. Knowing that won’t let you trace every thread. But it might make you slower to call a distant failure “not my problem,” and a little humbler about how separate you really are.