13 Comments

Matt’s email newsletter is free, but historical examples are gated if you don’t subscribe. I get them through one of the archiving services if I need an old one, and subscribe to not miss one now. Strong recommendation!

Expand full comment

I would argue that even the weird HTM accounting trick is a proximate cause at best. Most of this is downstream of the way banking (maturity transformation) and money (ex-nihilo loans) currently work. Both of these are deeply flawed, cannot easily be separated form each other, and must be fixed If we want a stable banking system. I wrote some more on that here: https://unfashionable.substack.com/p/maturity-mismatch

Expand full comment
Mar 16, 2023·edited Mar 16, 2023

> I have never understood venture debt.

Venture debt is how startups sell equity without "selling equity". I find the speculation that SVB used it this way, giving startups funding on favorable terms to get deposits, credible. SVB probably preferred it to buying equity because the optics for SVB were better, since lending money is more of an "oh sure banks do that" thing than startup investing.

The other major use case I've heard for venture debt is that the optics for the *startup* are better compared to the fraught expectations around "raising a round". For example, a startup needing cash and worried about a disappointing valuation might instead take on some venture debt at relatively *unfavorable* terms.

Expand full comment

If somebody needs a specific date's newsletter emailed to them, I've got all of them from the last several years.

Expand full comment
Mar 16, 2023·edited Mar 16, 2023

Far from a nuts-and-bolts explainer, I read Patrick's piece as an opinionated attempt to model the broader political economy of the situation. The not-even-that-Straussian reading is that he leans toward Balaji's view (minus the crypto boosterism). US political actors clowned first by pushing steep rate hikes to get egg prices down, then by trying to blame tech depositors for the fallout without understanding that those depositors were, transitively, ordinary Americans. The net money movement was a massive confiscation from bank shareholders for redistribution to consumers.

Expand full comment

Being pretty illiterate wrt financial [anything], the most interesting aspect of Silicon Valley Bodega I've been able to at least follow along with is the "sometimes transparency is bad, actually" claims. Lots of important gears undergirding modernity's sausage factory depend on people not looking too hard - especially if they're not qualified to understand what they see. A little knowledge can hurt someone, etc. There's the expected push to make this another front in Social Media Bad - which it is, but there are so many better ways to argue that position - plus the usual suspects complaining about [not enough] regulation. More grist for the culture war takes mill.

The linkage I haven't seen anyone explicitly make is between bank disclosure and C-SPAN. For awhile there was an interesting idea in The Discourse that a lot of the performative bits of politics we see today, the prioritization of The Symbolic Representation Of The Thing over The Thing, is because it's become both much easier and in some sense "expected" to try and make politics go viral. Suddenly, there's the potential for an Audience beyond who's physically present, and all the behavioral modifications that result. Obviously, that's not a reversed-stupidity argument for going all the way back to party machines and smoke-filled rooms...but it's a different set of incentives, so we reap what we sow by Making Government Transparent. Same with banking. A hard problem to solve at the current sanity waterline.

Expand full comment

While I wholeheartedly agree that the fed's interest rate estimates are not a promise, it would be interesting to evaluation them as *predictions*. Does the fed futures market out-predict the fed at these? Are there any other sources of prediction to compare to?

Expand full comment

look, this is a hard problem but what you want is unlimited insured deposits. The FDIC acted like that anyway. If this needs to be something where private companies need to come create a market, so be it, let them figure the moral hazards and be wiped out if they do not. Then and only then let the FDIC be the insurer of last resort, or maybe the FDIC insures up to 250k and banks need to buy insurance beyond, dunno

more regulations may be good or not, but gotta accept that they will be played. We learn one crisis at a time, so maybe the lesson this time is that the fdic should just insure everything or something along these lines

Expand full comment

Here's a take on the main indirect effect. Depositors will start to pay attention to what they're earning on demand deposits vs. the risk. When you can get a Vanguard T-bill fund with no interest rate risk that yields 4.5%, why would rational actors accept bupkes on excess demand deposits?

Because people are so used to zero interest rates and because of inertia. SVB changes that. And the ease of moving money around via the internet will accelerate that, too.

Long term this should be bad for every bank's earnings power as their net interest margin should compress.

Expand full comment

If everybody did proper cash management (i.e., split it up among multiple banks) then we get the exact world we're currently in but with unlimited FDIC insurance.

Simplified math: Imagine that everyone has $25M. After SIVB they split it up among 100 different banks.

Before: Person1 keeps $25M in Bank1, Person2 keeps $25M in Bank2, ..., Person100 keeps $25M in Bank100. 100 banks each have $25M in deposits from a single depositor, 1% of which is FDIC insured.

After: 100 people each keep $250K in Banks 1-100. 100 banks each have $25M in deposits from 100 depositors, **100% of which is FDIC insured.**

So why not just have FDIC insure everything in one spot? If people are smart enough (or if a startup provides the necessary automation to make it trivial) then 100% FDIC backing is going to happen anyway.

Expand full comment