a meeting at the weekend

how doughnuts nearly killed capitalism

I’ve told this story a few times, but to my surprise never publicly in writing. Toward the end of my time at Credit Suisse, I had a very interesting conversation with someone who was retiring from the bank’s US dollar funding desk, reminiscing about the 2008-9 crisis. At the time, the Eurocrisis was just beginning to gear up, and American money market funds were just beginning to get worried about their exposure (I’ve written a bit about the “Eurotrash trade” at Alphaville recently). My friend reminisced:

“You know what the sentence that nearly destroyed the whole of capitalism was? It was ‘we’re going to have a meeting at the weekend’.

“The short term interbank market for dollars is in London and New York. But a lot of the actual dollars come from funny savings banks all over the USA. They’re not so big you’d ever have heard of them, but they’re not exactly small either, and when you have a lot of them together, it’s a pretty big source of funds for the rest of the market.

“And the treasurers of these banks, they’re not paid to take risks, you know? They don’t want to lose any money, and they aren’t exactly in the flow of information.

“So what I kept hearing, over and over again, from Chuck at the Savings Bank of Stockton California and Darlene at the Bank of Boise – those aren’t real banks, but that’s the sort of thing they’re called – was someone sounding nice and apologetic, saying that they trusted me, they trusted the bank, they knew we were a good credit, one of the best names in the market[1], but gee … it’s all just getting crazy now, and the board are asking why they’re lending to European names. So … we’re going to have a big meeting at the weekend, and look at all the information, and make some decisions. I’m sure we’ll be back facing you on Monday. Tuesday at the latest.

“These were the weekend meetings that nearly killed the world. Because when you’ve got trillions of dollars of overnight funding to roll over every single business day, the one thing you absolutely cannot afford to do is to stand back for a couple of days and take stock”.

It’s stuck with me, as you can tell. As a description of how the global banking system works, I think it’s very important. Central bank swap lines are more institutionalised now than they used to be, and there’s a lot less use of overnight funding than there used to be, by the way, so it isn’t exactly how the next such crisis will happen, and it’s probably not even particularly close to the way that Silicon Valley Bank fell over.

But there’s a really important general principle in there that’s often left out of economic models. (But which, to be fair, is the basic driver of the Diamond-Dybvig model of bank runs which won the Nobel Prize a few years back). Some things, and not just in finance, have a time scale built into them. If a piece of information doesn’t arrive in time, then it might as well not have arrived at all.

In fact, information flows which aren’t properly synchronised to the control cycle can actually be worse than those which are completely lost, if they end up causing oscillations and destabilising feedback. Perhaps, when you consider how things went down in March of 2023, and particularly the reaction of the venture capital industry - those people ordering coffee, pastries and fruit platters for Saturday, maybe they weren’t making such a bad decision after all.

[1] this was true, at the time.