Silicon Valley learns how to kill a bank

The demise of tech's favourite bank in simple economics

By now you’ve probably heard of Silicon Valley Bank, which appears to have joined Northern Rock, Bear Stearns, and Landsbanki in the list of famous bank run victims. Exactly how this happened looks to be the result of two forces: the unwinding of ultra-loose monetary policy, and the traditional bank run panic.

How banks die

People tend to think of banks as a place to store their money. If they’re looking to buy a house, they might think of them as a place to get a loan. What a bank actually is is a machine for moving money around through a process called ‘maturity transformation’. The way this works is that when you open an account and pay into it, the bank takes your money and promises you can have it back the moment you want to spend it.

This doesn’t mean that it puts it in a vault. Money sitting doing nothing is essentially money wasted, and the bank has staff to pay and shareholders to reward. What it will usually do is take the money, lend it out, and have the borrower agree to pay it back at some point in the future, plus a little extra.

What our hypothetical bank has just done is acquire a set of short term liabilities that are essentially unconditional. Anyone, at any point, can walk up to their bank and say they want their money back, and it will have to hand it over to them. Money which is now sitting in someone else’s account in the form of a loan, which they don’t have to repay for ages, because the bank has also acquired long term illiquid assets in the form of the promised repayments. This is what we mean by maturity transformation; turning short-term liabilities into long-term assets. It’s also one of the most infamous ways that banks die.

You may have already spotted the core problem here. If the bank has lent out or invested the money it took in, what happens if the depositors decide they want their money back? If, for instance, the idea gets around that the bank is insolvent?

What happens then is a bank run. People start queueing in the streets to withdraw their money, and by the time the bank’s chairman can make a media appearance to point that “no, really, everything is fine”, it isn’t and the bank is dead.

This makes banks strangely fragile. So long as everyone thinks they’re fine, then they’re probably fine. But if people end up believing that something is wrong, the sensible thing to do is take your money out, just in case it is. And if everyone wants to take their money out, then the bank starts selling assets at a loss to meet cash demand, creating an actual hole in its balance sheet, and now there really is a problem and before long if you’re still in the queue you aren’t seeing a penny of that last payslip.

This is what makes bank runs so weirdly powerful; they’re self-fulfilling prophecies, and if you think there’s even the slightest chance of one happening, the best move is to act now because you can be sure everyone else hearing the news will act the same way.

Generally, banks and regulators are pretty aware that this can happen. Banks go out of their way to make sure that they’re seen as solid, and regulators offer schemes that make sure that deposits are protected up to a given value. If you’re not at risk of losing your money, you don’t need to go queue in the rain to get it back, the bank run dynamic never develops, and banks which are solvent stay that way.

So what on earth happened to Silicon Valley Bank?

Back to the valley

Robert Armstrong has a fascinating¹ overview of what went wrong here, but the short version appears to be that Silicon Valley Bank was flooded with cash, which eventually caused it to implode. Wait, what?

The obvious question here is ‘in what world does people shovelling money into your bank cause you problems’, to which the answer appears to be ‘when you need to put it somewhere’. As Armstrong writes, the bank “did not have the capacity, or possibly the inclination, to make loans… at the rate deposits were rolling in”.

So let’s go back to Silicon Valley Bank. You’ve got a balance sheet that on one side is filled with deposits that people can access at any moment, and on the other side you need something better than just keeping the money in vaults, because you’d quite like to turn a profit this year.

What appears to have happened is that Silicon Valley Bank piled its money into US government bonds. What could go wrong there? Bonds are safe! The government is good for the money! That’s the sensible thing to do!

Well, almost. The bonds the Bank bought were long term, paying off well into the future. If you hold them until they pay off, you get the full value. But what if you need the money now? What if your customers are eyeing competitors offering higher interest on balances?

Well then you have to sell them, and the amount you get for them is dictated by the current interest rate. Briefly, when interest rates are high, the price of bonds is low. And if interest rates were low last year, and then suddenly shot up, and you’d bought quite a lot of bonds, you’ve just watched the current value of your investments plummet.

This is where the bank run dynamic kicks in. Because the Bank’s customers are businesses, they tend to hold quite a lot of money in their accounts, well above the limits of the guarantees offered by the government. So if the Bank went down with your money in it, you might not get that back.

As Marc Rubinstein writes, Silicon Valley Bank’s customers “all know each other”. Some withdrew their money, or advised others to do so. And once you know other people are going to pull their money out, the race is on to get yours to safety.

Aftermath

A few things will happen now. Silicon Valley Bank will probably be sold. The depositors may well be made whole, if they can make it through that process.

The problem is that isn’t guaranteed, and the startup ecosystem — the companies that used the bank — may take an absolute beating. They need that money now. They have staff to pay, commitments to meet, and payments infrastructure that relied on services that are now closed.

1. Admittedly, fascinating here is used in the sense of “as fascinating as content about bank balance sheets can be” and look it takes all kinds to make a world