Bank Failures Visualized

https://observablehq.com/@mbostock/bank-failures

by A_D_E_P_T • 3 years ago

819 317 3 years ago

That needs to go back to the 80s to capture the SnL crisis. It dwarfs 08 in bank failures. It better indicates the conglomeration of the many banks into the few we have today.

FormerBandmate3• 3 years ago

08 was artificially low because many banks got merged at a fire sale. Wachovia, Merrill Lynch, Bear Stearns, and National City stick out. Other financial institutions got essentially nationalized and stock became mostly worthless like Citi and AIG, although the government sold most of their stock in 2011

Credit Suisse is about the same size as SVB, Signature Bank, and First Republic combined but it got “acquired” by UBS at a price 60% below its last trading price in a deal where $17 billion of debt was wiped out so it doesn’t count here

btilly2• 3 years ago

Credit Suisse is not included because it is a Swiss bank, not an FDIC insured US bank.

Lehman Brothers is also not included because, even though it was a US bank, it was an investment bank with no FDIC insured deposits. It was around the size of all of this year's failures, combined.

As you note, bank bailouts that were not FDIC bankruptcies are also not included.

rrrrrrrrrrrryan2• 3 years ago

It depends on what we're trying to visualize. From an investor's perspective, a bank whose assets get sold for pennies on the dollar in a fire sale is essentially a failure. Lehmann Brothers was also a massive (investment) bank failure with huge second order effects on the economy.

This graphic seems to be modeling things from a taxpayer perspective. These banks failed and the government needed to step in to do something to ensure people could get their deposits.

themagician6• 3 years ago

I have faith that we can top S&L. We have the technology. We have the talent. There are six banks with over a trillion in assets in the US. I have faith that one of them has been doing some wild book cooking. I'd place a bet on Citibank, followed by Wells Fargo. There's an old saying in Tennessee — I know it's in Texas, probably in Tennessee — that says, fool me once, shame on… shame on you. Fool me… you can't get fooled again.

I'm just hoping this time it's something absolutely outrageous just for the lulz of it all. Like, let's get some FTX-style absurdity. All the absurdity happens in crypto right now but I still have faith in regular banking. Some people still like the challenge of regulated markets.

yieldcrv14• 3 years ago

If you like absurdity, FTX has recovered 7.3 billion out of the 8.6 billion hole and plans on relaunching the exchange to make the last billions back in fees

Most noteworthy is that this quick 8 month turnaround is partially thanks to the blockchain, and under no new laws being passed

edrxty• 3 years ago

This may be my favorite all time HN comment.

I totally buy it too.

bboygravity2• 3 years ago

JP Morgan is the biggest one. It can do ANYTHING it wants and get away with it.

It can make 10 billion USD spoofing gold prices for a decade and get away with a 1 billion USD fine (and keep doing it) for example.

The CEO can go on trips with Jeffrey Epstein, be friends with him and do business with him and get away with it.

It made tons of money off of the Madoff ponzi by providing Madoff with a bank account and not reporting the (from their perspective) extemely obvious ponzi that was going on for 15 years. Nobody went to jail and JP Morgan's fine was probably lower than what they made from the ponzi.

There are 100's of other examples of quite outrageous FTX-style crime. This is just what I happened to read about and remember. And that's only the publicly known stuff.

Let's turn it around: why would JP Morgan (and other big banks) NOT be engaged in extreme levels of crime that could be described as "financial terrorism"? If JP Morgan blows up it would be the end of the US and they know it and the US govt knows it. I repeat: they can get away with ANYTHING.

I think you will get your lulz.

RoddaWallPro• 3 years ago

Fool me... you can't get fooled again.

George W gaffes have become so hilarious to me, now that they're 15+ years in the past. I laughed out loud reading this one.

derbOac• 3 years ago

Yeah these sorts of figures always frustrate me because the historical context for these things is so much broader, and it seems obvious to me to go back to the 80s, if not earlier.

There's a better figure here I think:

https://www.pewresearch.org/short-reads/2023/04/11/most-u-s-...

Also some nice figures here:

https://www.bankingstrategist.com/history-of-us-bank-failure...

kibibyte• 3 years ago

TIL there was a different "First Republic Bank" that failed in 1988. This name carries a curse now.

DonsDiscountGas5• 3 years ago

Going by Wikipedia, in 2021 dollars I count:

1980s S&L crisis: $654 Billion (summed 1984-1992 failures) across 23 banks

2008 crisis: $733 Billion (summed 2008-2011 failures) across 61 banks

2023 so far (it's only May): $556 Billion (Signature + SVB + FRC) across 3 banks.

It looks like 2008-2011 is the "winner", although other commenters have mentioned forced mergers etc. may not be counted.

https://en.wikipedia.org/wiki/List_of_largest_bank_failures_...

boringg2• 3 years ago

That on 2022 dollars?

dnissley• 3 years ago

Aren't there some relevant details missing from this kind of analysis? Banks failing just means that the value of the banks assets fall below the value of their deposits, right? In which case the degree to which that happens seems to be highly relevant to this kind of comparison. E.g. the value of assets falling to 50% of deposits in bank failures in financial crisis A vs 90% in financial crisis B

haldujai8• 3 years ago

At the very least that’s missing Fannie, Freddie, Bear, Merrill, Lehman, TARP and arguably AIG for another 1.2T+, granted a lot of this was eventually repaid as the FDIC will be as well.

rr8082• 3 years ago

conglomeration of the many banks into the few we have today

There are still thousands of banks in the USA. I dont really see why there should be more than 100. Canada has 5 big ones and a few dozen tiny ones. Same in UK and Australia.

peter3032• 3 years ago

Deposit banks were banned from crossing state lines and investing in equities markets after the Great Depression. These limitations were removed in the 1990s.

dalyons• 3 years ago

It’s not all roses - in Australia and Canada the few major banks operate in pseudo-cartel fashion. There’s few enough that they can effectively collude without doing it illegally.

peter3032• 3 years ago

Round 2 of the 2023 bank crises might be commercial real estate loans like the 1980s crises. Some downtowns are still 40% vacant. This more a consequence of covid than bad bank behavior.

fragsworth• 3 years ago

You are correct. Additionally, the size of the bank(s) are not really what matters.

I want to see the scale (sum) of what was actually lost when they went bankrupt, and how much we (the public) have to put up to keep the system from collapsing. Does anyone have an actual visualization of how much we ponied up to keep our banking system from collapsing?

Did the public just provide a reasonable interest rate loan for a few months to a year? Or was it a sweet 0% loan for...ever? The important details are lost in the media reports and it would be nice to get a sense for what really happened.

than3• 3 years ago

Generally speaking, it turns into free assets. The people involved usually don't acquire the assets/associated debt without some kind of guarantee.

OneWest Bank for example after 2008 had a guarantee where if the assets defaulted above a certain amount they would receive full value of the loans in a payout from the government. They were actively foreclosing on people to justify catastrophic losses to get the bailout. Not sure how that ended up since I was only marginally aware of the start of that and everything went silent once the news got wind of the perverse incentives.

In terms of trends, the bailout game has been played consistently since the the dollar went off the gold standard (1971 iirc).

The ponzi is starting to unwind now that inflationary pressures are out of control. I expect concentration to eventually lead to nationalization followed by a new currency which will fail because they lost all credibility from their mismanagement as a private entity.

That's what's happened historically with every country that debases its store of value above the point macro effects become noticable which are around 3:1 ratio).

TheSpiceIsLife• 3 years ago

There's argument to be made that approximately any cost to keep the system from collapsing is a trade-off worth making if the alternative is the system collapsing.

jeremyjh3• 3 years ago

There were winners and losers, but the government made money on TARP. That doesn't fit anyone's narrative very well, so you don't hear that much about it, but its a fact. So far not a public dollar has been lost in the current crisis. FDIC, like other insurance, is paid for by the insured. Every FDIC bank in the country is paying the cost of this.

https://en.wikipedia.org/wiki/Troubled_Asset_Relief_Program

pranshum• 3 years ago

I wrote a post about this. The frequency of failures was much higher but the individual failures were smaller: https://yarn.pranshum.com/banks2

a3w2• 3 years ago

saturday night life crisis?

cyclecount2• 3 years ago

Savings and loan crisis https://en.wikipedia.org/wiki/Savings_and_loan_crisis

lisasays• 3 years ago

A veritable Mandelbrot Set of failure

5kg2• 3 years ago

<del> r should be sqrt(assets). </del>

It is.

loeg• 3 years ago

Are the assets inflation-adjusted, also?

DonsDiscountGas5• 3 years ago

Or else replaced with a fixed-width rectangle and scale the height by assets. Would fit more easily on a timeline and humans are much better at comparing heights than we are areas.

mbo2• 3 years ago

your wish is my command: https://observablehq.com/@mjbo/bank-failures-as-a-stacked-ba...

jeffbee• 3 years ago

It is! That's the default radius type for the dot mark on Observable Plot. To play with it, add `type: "linear"` to the dict on line 6.

5kg2• 3 years ago

Thanks. I guess rename 'r' to 'area' will make it easier to read.

PheonixPharts• 3 years ago

I generally love Mike Bostock's work, and this is a beautiful visualization, but I think this comment points out why circles are just a bad way to represent anything visual other than pizzas: it's a leaky abstraction (which is a real sin in visualization)

You have to know how the chart is generated in order to correctly interpret what you're seeing. The value can be the radius, diameter (how pizzas are measured) or the area itself! All of these choices lead to different interpretations of what you're seeing.

Then, on top of that, we're just not great and visually comparing areas. Pizzas are a great example of this. A 12" pie and a 16" pie don't look that different but you're talking about nearly twice the amount of pizza! A good example in this visualization is Signature vs SVB

Then on top of that we have the dimensions of the circle forcing the visual to overlap on the x-axis without this meaning anything concrete.

All that you can meaningfully take away from this visually is "2008 had one big and a ton of small failures, and currently we're seeing 3 big, but not quite as big failures", and unfortunately there's not a lot more you can get out of this. Because of the x-axis problem it's hard to even tell if WaMu preceded some of the others or not.

frgtpsswrdlame• 3 years ago

Hm? Where's Lehman?

DonsDiscountGas5• 3 years ago

I think this is just consumer (ie FDIC) insured banks, Lehman was an investment bank.

riskneutral• 3 years ago

This needs to be re-expressed as a % of GDP to be useful. And as already pointed out, the largest of them (Lehman) is missing

FormerBandmate3• 3 years ago

Lehman Brothers isn’t technically a bank failure. It was an investment bank that went bankrupt, it had zero FDIC involvement

Findeton• 3 years ago

And Credit Suisse although it's not on the US.

WatchDog• 3 years ago

Constantly impressed by Mike Bostock's work.

csense• 3 years ago

I feel like it should include earlier time periods (80's? Depression?), and have some adjustment for different times. I'm not enough of an expert to guess whether to use inflation adjusted dollar amounts (using what inflation metrics?), or as a % of GDP, or a % of the banking system..

rr8082• 3 years ago

Total number of FDIC banks over time is listed here: https://banks.data.fdic.gov/explore/historical?displayFields...

steve_adams_86• 3 years ago

Are there no small failures this time around because of ongoing consolidation? This seems to be a dominant trend in economics across the board. Does anyone have a good answer why?

krona• 3 years ago

The trend is a form of preferential attachment. https://en.wikipedia.org/wiki/Preferential_attachment

rrrrrrrrrrrryan2• 3 years ago

Some midsize banks got a massive inflow of deposits during the pandemic and they didn't know what to do with the money, so they parked it in "safe" investments like mortgage backed securities.

Then when the pandemic subsided, people started burning down their savings accounts, and interest rates started skyrocketing. All the banks' money was tied up in all these 30-year mortgages that nobody wants anymore because they're paying like 2.5% and inflation is 8% now, but they're still forced to sell them for terrible prices just to keep their ATMs full. Eventually word gets out about what's happening, there's a bank run, and the bank starts death spiraling.

Massive banks have nothing to worry about because if the government allows them to fail they'll take the entire economy down with them. And (presumably) smaller banks weren't under the same pressure from investors to generate returns on idle capital, so they didn't commit the same sins as the midsize banks.

tickerticker• 3 years ago

SVB's fail was because of radical speed in deposit inflow, which was invested in fixed rate bonds just at the moment that the Fed was on a rising rate ratchet. Because of an affluent clientele, they had unusually high % of uninsured deposits. Very unlikely that a small bank would have had the need to layoff so many deposits in such a short time span. Very unlikely that a smaller bank would have had the high % of uninsured deposits. These are the funds that jump ship immediately.

First Republic fail was because of large pool of fixed rate assets....(jumbo low-rate mortgages and Treasury Notes) in a rising rate context. They also had an affluent clientele and thus a high % of uninsured deposits. The fixed rate assets lost so much value while rates rose, that they had no tangible equity. Thus, the hot money deposits raced out of the bank.

Smaller banks typically don't have such a concentration of affluent customers, which means that more of their clientele would be under the threshold for FDIC insurance. Smaller banks probably have a good book of commercial loans which are commonly priced at a variable rate.

I am baffled that SVB and FRB did not hedge their fixed rate portfolios with interest rate swaps....Maybe their mind was on staying abreast of the white-hot tech sector instead of wringing their hands about the next Fed rate decision.

ausudhz• 3 years ago

"too big to fail"

photochemsyn4• 3 years ago

Community-scale banks (aka credit unions) are a better idea for local-regional communities (aka cities and towns and agricultural regions) because their managers have to live with their clients.

Take a community of 100,000 families, in an economic system where they're all collecting income and paying bills and so on. The idea behind a bank is that they hold the community's money securely while making their own money by providing loans at a greater interest rate than the interest rate they pay out to their depositors. Fundamentally, this means the bank is dependent on the fact that customers don't try to withdraw all their money at once (as the above dynamic means the bank is never able to meet this demand).

FDIC guarantees by governments are intended to bolster this customer trust, as if everything goes sour the deposits (to a certain limit) are guaranteed. The important point is that the regional bank has a profit model based on giving successful loans to individuals and small businesses for things like houses, cars, business expansion, etc. The bank will of course want to know a fair amount about the people they are loaning to (drug addicts are not good bets, etc.) in order to ensure they get paid back.

I hear people laughing in the background about the naivete of this picture in the modern American financial system. Since this post is already a bit long, go watch the movies "Margin Call" and "The Big Short" to get an idea of what's actually been going on.

Night_Thastus• 3 years ago

I've always used a credit union, and I love it!

I can easily get a real person in front of me, there's not a million layers of bureaucracy, and most importantly - no bullshit fees. I don't constantly get my bank drained just for having an account in the first place.

Scoundreller• 3 years ago

To join mine, I bought a $5 share (as a minor) and continue to get a $5 dividend many many years later (but I gotta close the account to get the accrued dividends...).

I gotta go to the AGM one day. I hear it's catered.

cperciva2• 3 years ago

The model you describe is exactly why S&L happened. Small banks with highly correlated deposits holding their own loan books is a recipe for maximizing vulnerability to economic shocks. Interest rates go up and the loan book loses value; the local housing market drops and the loans get foreclosed and lose value; a major local employer goes out of business and depositors all start pulling their money out instead of reliably depositing paychecks.

Banks don't work that way any more because it's a really bad way to run a bank.

photochemsyn4• 3 years ago

This is a rather skewed perspective that ignores the fact that if you have FDIC insurance for depositors then that can be gamed unless you have strict regulation of the banks:

"The roots of the S&L crisis lay in excessive lending, speculation, and risk-taking driven by the moral hazard created by deregulation and taxpayer bailout guarantees."

https://www.investopedia.com/terms/s/sl-crisis.asp

This is why a lot of people worry that Silicon Valley Bank, First Republic, etc. might be the tip of an iceberg. If they've all leveraged themselves on risky speculation bets in the hopes that they'll get bailouts if it all goes sour (even though depositors greatly exceeded FDIC insurance limits) then you could have a domino situation.

Note also that it's a perfectly good way to run a bank as long as you don't get greedy and go for big risky bets, but the only way to ensure bank managers don't get the Las Vegas bug is to enforce the banking regulations in a fairly strict manner.

CPLX• 3 years ago

Might be a bad way to run a bank, might not, but it's definitely not the cause of the S&L crisis. That was a story of deregulation leading to massive moral hazard, rampant speculation, and multiple high profile cases of outright fraud.

Economic shocks and rising interest rates aren't usually the reason that financial institutions engaging in risky or fraudulent behavior get caught, they usually just make it so the clock runs out on their scheme to roll over the losses or otherwise avoid getting caught.

The liquidity crisis of 2008 didn't cause Bernie Madoff's fund to collapse, the fact that it was a complete fraud from inception did. It just made it impossible to ignore.

llanowarelves• 3 years ago

Evidently, the best way to run a bank is a 0% cash reserve minimum (as was granted to US banks starting during COVID), so you can make money off of literally every last penny, and then discourage and prevent customers from taking their money out, then [externalizing all costs]/[socialize losses] onto government (taxpayer) via bailout or customer via bail-in, while privatizing all profits in the meantime. And charge poor people $30 every time they dip under $0 even by a penny.

Doesn't mean it's a good idea for the rest of us non-banks.

NoMoreNicksLeft3• 3 years ago

Banks don't work that way any more because it's a really bad way to run a bank.

As compared to what we have now? With dubious financial instruments so opaque I'd have to spend 30 years lurking underneath desks on Wall Street eavesdropping on conversations just to have any clue at all how the fuck those work?

I'm almost comforted when there's a Bernie Madoff, because at least I can wrap my head around how a Ponzi scheme works (ignoring the ethics, obviously). I'm (irrationally?) worried that some of what's been going on the last few years actually makes plain Ponzi schemes look legitimate by comparison.

Not 25 years ago, we used to laugh about the stupid books with titles like "Dow Jones 100,000!" and whatnot, I think it was a Slashdot post way back when. And while we haven't quite reached that pinnacle of absurdity, it did hit 37,000 not so long ago.

None of us may be able to pick the queen of hearts, but some of us have caught on to the fact that it's three card monte. When it all falls down, should we console ourselves with "well at least the banks weren't small and local and vulnerable to economic shocks"?

opportune• 3 years ago

It’s more complex than that. The ideal situation for a bank is when all its depositors are customers of each other. The deposits never leave the bank so the bank has very stable liabilities.

So a local bank will be great when a community transacts mostly with itself. SL happens right around when multinationals and corporate centralization started becoming more of a thing..

We live in the era of trillion dollar corporations, fintech, and the internet so a community bank ends up more like a small big bank. Its depositors are getting paid by corporations from way outside the community, and they’re spending money on the internet or at giant multinational retailers and thus shifting deposits outside the community. In this paradigm, it’s actually better to be huge and have a diversified customer base because the best way to keep your liabilities stable is to have the largest market share possible.

You’ll notice, banks usually have generous incentives for setting up direct deposit. Direct deposit is the best way for them to have a predictable measure of incoming deposits when there isn’t a high likelihood of one particular account transaction with another.

EGreg2• 3 years ago

Credit Unions are better also because they are literally libertarian socialism. Imagine Uber with no shareholder class - as a cooperative, and so on. No one to extract rents from the two sides of the marketplace, the people themselves own the network.

So are housing cooperatives. They have low rent because there is no landlord class. Look at the Mitchell-Lama program in NYC, 30 years later they the most desirable and awesome apartments for the price, all around the city. Meanwhile next door the capitalist landlord-owned buildings are 3x as much for worse amenities.

https://en.wikipedia.org/wiki/Mitchell%E2%80%93Lama_Housing_...

The democratically-run city tries to fix this with rent control, but it's only some housing, and just a few mins ago we learned this:

https://www.nydailynews.com/new-york/ny-rgb-preliminary-vote...

Plus the government is notoriously bad at getting even the simplest stuff done:

https://therealdeal.com/magazine/national-july-2022/that-emp...

So yeah, I'm a libertarian socialist. People in the US hear socialism and they think of a big, oppressive government, but actually government is just as much if not more on the side of the capitalist, the industrialist... bailing out large banks and injecting trillions into corporate equity to prop up markets. In the 19th century they actually violently put down strikers, evicted people, operated debtor's prisons etc. which is why most libertarians back then were socialist. It was cool. Today that lefty libertarian stuff is mostly in Europe.

https://en.wikipedia.org/wiki/The_Soul_of_Man_Under_Socialis...

There are very few famous lefty libertarians today, people like George Carlin and Russel Brand maybe.

JPws_Prntr_Fngr• 3 years ago

That article at therealdeal has me pretty stunned.

The headline story is that landlords are keeping rent controlled units vacant rather than renovating them which would be unprofitable, and daring the toothless regulators to do something about it. Ok, makes sense.

Then as a side note, the article mentions that they’re also deferring maintenance on whole buildings: roofs, boilers, cameras, etc. The fact that these landlords are incentivized not just to let a few units turn into roach hotels, but to let their whole buildings crumble, without the remaining tenants leaving to a better building - that’s a fundamentally broken market. Way beyond a few rent controlled units.

tadfisher• 3 years ago

Credit unions aren't insured by the FDIC, they have NCUA insurance.

candiddevmike2• 3 years ago

Why are drug addicts not good bets? Coffee drinkers seem pretty successful?

judge2020• 3 years ago

Sure, "Coffee is a drug" just as much as Caffeine and sugar are. OP likely meant that those who abuse scheduled drugs (outside of weed) are more likely to develop a habit of putting off work to feed that addiction, or to be fired for misconduct in connection with their use, resulting in more late payments and eventual repossessions/foreclosures on loans. It's all in terms of managing risk, but I'm sure chronic alcoholics are just as risky, if not more risky, as meth heads, since they blend in and have a good credit score but are more likely to lose their job for being intoxicated at work or go to jail for something like drunk driving.

kortilla2• 3 years ago

This doesn’t work. The very failures we are seeing today are why this doesn’t work.

Why would someone leave their money in the regional bank offering 0.3% that all of their issued loans average barely above when people can take it to a bank offering 3% or get CDs offering that?

Community scale banks aren’t a solution for this.

zamnos6• 3 years ago

When Chase, BofA, and Citigroup are all offering approximately 0.0000% APY on their CDs, and it's credit unions that are offering 3+% (you should be shooting for at least 4% right now), it's not clear that big banks are actually better than a smaller, community-focused bank or credit union.

https://online.citi.com/US/ag/current-interest-rates/cd

https://www.bankofamerica.com/deposits/bank-cds/cd-accounts/

https://www.chase.com/personal/savings/bank-cd

I double checked, and it turns out they all have a couple of products offering closer to 4.5% APY, but the tables on those linked pages still have far too many numbers like 0.05% APY on them to be taken seriously.

WeylandYutani2• 3 years ago

This would make the bank extremely vulnerable if the community is hit by a recession. You need to diversify the risks not bet it all on a single hand.

WeylandYutani2• 3 years ago

Some people want smaller banks because "too big to fail". While those same people are already freaking out when two banks fail.

I'm taking the downvotes: just open an account at JPMORGAN. Banking is serious business your artisanal cornershop credit union can't hack it.

stall84• 3 years ago

That is quite discouraging. Thank you. (no that is an awesome UI.. but dang..)

attentive• 3 years ago

Is this inflation-adjusted to compare to 2008 dollars?

djbusby2• 3 years ago

Would be cool to compare the dollar-of-fail as % of GDP at the time. Finance is all about ratio after all.

momofuku• 3 years ago

For the life of me, I'm not able to understand why none of these lists include Lehman Brothers?

A commenter below posted:

Lehman Brothers is also not included because, even though it was a US bank, it was an investment bank with no FDIC insured deposits. It was around the size of all of this year's failures, combined.

So does this mean, that since Lehman had no customer deposits, it doesn't count?

SeanAnderson• 3 years ago

Well, they're visualizing a dataset from the FDIC. So, it's more like it doesn't count due to a technicality. It's not clear to me if there's an "official" dataset that includes all non-FDIC institutions.

djbusby2• 3 years ago

For this dataset, yes. It's only got deposit banks (afaict)

mywittyname2• 3 years ago

Lehman had no FDIC-insured accounts. So it shouldn't be included in a list of failures managed by the FDIC, because it wasn't.

The liquidation of Lehman was handled by the Treasury department (and a bankruptcy court), not the FDIC, and Congressional involvement was necessary to pass a bill to finance the resulting "bailout". The FDIC maintains its own fund to pay back depositors.

It might sound like a pointless technicality, but it's really not. It would be like asking why an NHSTA report doesn't include plane crashes.

EGreg2• 3 years ago

So basically, a ton of more failures in teh past, now it's slowed down but we are losing our stuff over it?

rout395742• 3 years ago

I'd say, "Our few failures now, which we think of as plausibly just the leading edge, are already comparable in size to the _entire_ 2008 cascade."

Losing stuff now seems... premature, but not insane.

haldujai8• 3 years ago

This is not even remotely close to 2008… This figure conveniently excludes the largest 6 failures of 2008, TARP, and major international banks and insurers simultaneously affected.

As a reference the prospect of AIG failing without a bailout was so significant it became an international issue threatening both US and EU economic stability. SVB and FNB are borderline irrelevant in comparison.

The amounts reflected in SVB and FNB are also exaggerated as the money isn’t “lost” as if parent can obtain liquidity elsewhere (via government or merger) the asset will return principal + interest at maturity.

yalogin• 3 years ago

Weird part about this whole thing is, we have been repeatedly told that banks are good and they learned a lesson in 08. Now both tech and banking are in trouble again.

echelon• 3 years ago

we have been repeatedly told that banks are good and they learned a lesson in 08

This is a wholly different lesson.

In 2008, banks were making bad investments.

In 2023, the changing interest rate environment caused good investments to become worth less than their original value. If held to term, things would be fine, but liquidity issues put stress on the system.

These are not the same, and we have better means of dealing with this problem. SVB and First Republic were handled appropriately. Investors written down to zero, depositors made whole.

The future banking system will be even more robust after having learned this new lesson. I've no doubt that we'll have regulations that demand a better mix of investments that are regularly audited and stress tested.

We as a society also have to stop thinking of banks as a means to earn interest on deposits. Chasing the best rate has led to this problem. Deposits are liabilities.

candiddevmike2• 3 years ago

What lesson are they going to learn? Go big or go home? I don't foresee any kind of legislature becoming law over this, especially with our divided government.

Mistletoe4• 3 years ago

I strongly disagree, the current mess is largely caused because some banks didn’t anticipate interest rates going up from historic lows. That’s profoundly bad risk management. A child could have done better.

https://fred.stlouisfed.org/series/FEDFUNDS

anecdotal1• 3 years ago

In 2023, the changing interest rate environment caused good investments to become worth less than their original value.

They were not good investments

brvsft• 3 years ago

Every time someone says "stress test", I will point out that the stress tests did not model for interest rates rising like they did, which triggered the current problems we're having (or, really, interest rates being so low for so long was the actual problem IMO, but that's a whole separate conversation). Stress tests as currently implemented would not resolve the issue, and even if they were so wise to reactively throw in more aggressive changes to interest rates to these stress tests, would that be enough to catch the next problem?

https://www.federalreserve.gov/supervisionreg/dfa-stress-tes...

See Tables 2.A and 3.A in 2022 Stress Test Scenarios (PDF), covering a 3-year period starting in Q1 2022. Look at the 3-month Treasury rate. In 2.A, rates go up to 1.5 by Q1 2024 and sit there. In 3.A, rates sit at 0.1 for three years.

Rates are not the only variables, but the point remains that the interest rate changes that triggered these problems do not appear in the stress tests.

lottin9• 3 years ago

A bank is a business and businesses fail sometimes. This is inevitable.

kavalg2• 3 years ago

However, this time it looks like the government failed, not the banks. If you look at the US government (bonds) as just another business the bank can invest in, you can't help but notice that the business had not been very well managed. It promised that bonds will keep their value, but they did not (currently at market price and in the future, due to their yield suffering from inflation). We still don't have a clear message on the situation with government debt as well, but we are probably headed towards more inflation and/or some (hopefully soft) default(s). And as always, fewer people will have more assets/control and the rest of us will pay the bill.

beginnings3• 3 years ago

National governments should control their own money, scrap interest entirely. Let the market decide what each currency is worth for international trading.

The status quo international banking system is designed to enslave, and governments can't do anything about it because the banks are above them.

How anyone can defend the current system is beyond me. It's not remotely close to being the best we can do, it's a scam at every level, way more complicated that it needs to be, by design.

highwayman47• 3 years ago

What is the alternative?

lottin9• 3 years ago

National governments should control their own money, scrap interest entirely. Let the market decide what each currency is worth for international trading.

Isn't this how the current system already works? Except for the "scrap interest entirely" part, which I'm not sure know what it means.

beginnings3• 3 years ago

No, governments don't control the money printing machines.

Scrapping interest means that when the money is loaned after being printed, there isn't a debt attached to it.

realjhol• 3 years ago

It's not remotely close to being the best we can do

"We" can't do anything at all. We are modern day peasants, and we will accept what the regime tells us to accept

kypro• 3 years ago

Would be slightly more insightful if it was inflation adjusted. The circles on the right should be ~30% smaller.

smashem• 3 years ago

And how about relative to M1 money supply?

littlestymaar2• 3 years ago

Comparing to M1 doesn't make much sense, but “total bank assets” roughly doubled in the meantime: https://fred.stlouisfed.org/series/TLAACBW027SBOG

_heimdall5• 3 years ago

The definition of M1 and M2 money was changed in May 2020, conveniently right as the government started printing money for their pandemic response.

vukadinovic• 3 years ago

I was thinking the same!

bobbylarrybobby3• 3 years ago

And if the circles’ areas, not their radii, were the failed assets. As it is, a circle’s apparent size (we judge 2D shapes by their area, not any one linear dimension — although this is difficult to do in practice, and linear marks are generally superior) is the failure size squared, which distorts the data a boatload.

Here is a, er, more faithful representation of the data. The recent failures don't look quite so crazy anymore!

https://observablehq.com/d/d85ce0c639bc4df7

mbo2• 3 years ago

I have also built an alternative as a stacked bar chart https://observablehq.com/@mjbo/bank-failures-as-a-stacked-ba...

JBiserkov2• 3 years ago

They still look pretty crazy to me! https://www.fdic.gov/bank/historical/bank/index.html

Total assets of banks failed in 2008 + 2009 => 373 + 171 = 544 Billion

Total assets of banks failed in 2023 (so far, still 2/3 to go) 548 Billion

ghayes• 3 years ago

FWIW, I think the size of the insolvency is more telling of the failure than the AUM. E.g. First Republic was about ~$13B in the hole when it was taken over. Otherwise a large bank with a large hole looks way worse than a small bank with a huge hole.

Agent-Cooper• 3 years ago

The circles' areas are showing the failed assets. Observable Plot defaults to using a square root scale when encoding a quantity with the radius of a circle.

https://observablehq.com/plot/marks/dot#dot-options

p-e-w• 3 years ago

we judge 2D shapes by their area

I've seen this claim a few times in discussions about such charts, but it's far from obvious to me that this is really true. My best guess based on my own perception is that we estimate quantities represented by circles as proportional to something between the diameter and the area.

navane• 3 years ago

You ploted the size on a log scale and claim "the recent failures dont look quite so crazy anymore?" Is this a joke?

HPsquared• 3 years ago

That's a log scale...

cduzz• 3 years ago

I would also have a 2nd plot that has all banks / banking / bank assets with the failures in a different color, so we could identify the relative size of the failed banks, and also see the actual size of the big players in banking.

IceHegel• 3 years ago

This uses a log y axis. The circles are actually better for visualizing.

roflyear3• 3 years ago

Thanks for that!

ajross2• 3 years ago

It's also leaving out non-FDIC bank failures like Lehman, Bear Stearns et. al. which would make the '08 crisis much (MUCH) larger.

Basically post-2008 the class of "investment banks" basically disappeared. But none of that is shown in this chart.

ivalm• 3 years ago

They didn't hold deposits. So it's kind of reasonable. They are called "bank" but they aren't the same kind of institution.

NoboruWataya• 3 years ago

I can understand why these banks were omitted but it completely changes the picture. Lehmans and Bear Stearns were the bank failures of 2008.

ekianjo• 3 years ago

Are you trying to say that the current crisis is nothing ? These are famous last words.

Buildstarted• 3 years ago

This person adjusted the values for inflation

https://twitter.com/viralhysteria/status/1653369995877908481

bitslayer• 3 years ago

If we are comparing the impact, shouldn't it be adjusted for the size of the economy? The U.S. GDP was under $15T in 2008 and is over $23T today.

miclill• 3 years ago

"More insightful" is too gentle. At best the chart is misleading. But the _mistake_ to not adjust money for inflation is very common unfortunately...

tempestn• 3 years ago

Looking at this, it seems even more odd that no tiny banks have failed along with these larger ones.

AYBABTME• 3 years ago

It's just a matter of visibility. The failed banks so far were all family of the tech industry. Since the tech industry is so far one of the most impacted by contracting conditions, its banks are being poked at first. Now that it's done, the other banks are going to get some love.

fwlr• 3 years ago

Highly suggest sorting by date rather than size. Putting the smallest failures at the top causes a little bit of an optical illusion making the first tower appear smaller than it really is. It also highlights that whatever is happening here is drastically different than what happened in 2008 - here we have three “big bank” failures while 2008 had only one, but we have zero “little bank” failures while 2008 had hundreds.

haldujai8• 3 years ago

The difference is this is missing the biggest failures of 2008 (Lehman, Bear, Merrill) as well as Fannie, Freddie, AIG and TARP.

jbverschoor• 3 years ago

Indeed. Many factors make me distrust the author about his intentions and bias (great mongering)

JoeAltmaier2• 3 years ago

Yes, yes, inflation changes things (a little) but the glaring result is, these bank failures are massive, significant and unprecedented.

We can find historical failures of adjacent institutions but clearly something new is happening?

haldujai8• 3 years ago

If two regional banks failing with little realized loss is massive and unprecedented how would you describe 2007-2008 when the entire world economy was brought to its knees with napkin math ~1.2T of bailouts/failures in the US let alone Europe with several banks nationalized (e.g. UBS and RBS) and nation states defaulting on debt?

JoeAltmaier2• 3 years ago

From the linked graph, recent bank failures of this kind rival that entire event. And we're probably not done.

JBiserkov2• 3 years ago

https://www.fdic.gov/bank/historical/bank/index.html

d--b• 3 years ago

Here is a forked version that includes Lehman Brothers:

https://observablehq.com/d/dfe4257c84256e26

a3w2• 3 years ago

Nice. I thought WhoTF is Washington Mutual?

adasound• 3 years ago

Jesus Tap-dancing Christ

dilippkumar• 3 years ago

I’m stealing this excellent phrase.

nologic013• 3 years ago

It makes perfect sense to focus on events defined in a uniform way across the dataset (in this instance US FDIC insured bank failures). Otherwise what you are depicting is an arbitrary collation of events that may have very different causes.

What I derive from the visualization (btw Bostock is a genious) are a couple of simple yet still tentative observations:

* A new cluster of failures might be forming. The pattern of correlated failure is not new and it may point to similar business models and/or exposure to the same factors. Measurable correlation may also imply some level of contagion. This is defined as one event increasing the chances of another (after conditioning on common factors). But it is anybody's case what the fully developed phenomenon will look like. Three events is still small number statistics. Dont run an AI model on this.

* There is at this point a remarkable absence of small bank failures. If this persist it hints that it is actually not the vanilla banking model (and poor risk management) that underlies what is happening, but more idiosyncratic factors that are specific to a few actors. In some sense that would be good news.

One thing is for sure. The crisis will be properly visualized.

theteapot2• 3 years ago

There is at this point a remarkable absence of small bank failures.

Not sure it's that remarkable? 1. not adjusted for inflation as others have commented. 2. There are less banks in the US, maybe up to 1/2 as many as in 2008 (https://www.statista.com/statistics/184536/number-of-fdic-in...). 3. You may only be seeing the leading edge of something similar to 2008-2010 range - too early to tell.

nologic013• 3 years ago

For a qualitative discussion of what might be happening inflation is a just a nuisance factor. There are other similar factors (e.g the size of the economy is not constant either) that we can ignore for this purpose.

The bank size distribution is more relevant but 50% of a large number is still a large number. You'd expect this to somehow show up in the statistics.

Timing is indeed a key aspect. The actors involved have quite a bit of discretion and options and it takes time for all these to play out. But if it does transpire that small banks are not as seriously affected as in the last crisis and its more of a "middle class" bank disease this would be a strange new thing.

0xDEF• 3 years ago

Both SVB and Signature Bank were atypical banks with atypical depositors.

SVB catered to risky tech startup and Signature Bank was involved in risky crypto shenanigans.

On SVB:

As a regional bank in the Bay Area, SVB offered services specifically designed to meet the needs of the tech industry, and soon became the largest bank by deposits in Silicon Valley and the preferred bank of almost half of all venture-backed tech startups.

On Signature Bank:

By 2021, cryptocurrency businesses had represented 30 percent of its deposits.
Banking officials in the state of New York closed the bank on March 12, 2023, two days after the failure of Silicon Valley Bank (SVB). After SVB failed and in light of the closure of the cryptocurrency-friendly Silvergate Bank earlier in the week, nervous customers withdrew more than $10 billion in deposits.

First Republic Bank was also an atypical bank that offered high interest rates for high-net-worth depositors by not insuring deposits and having a very high loan-to-deposit ratio:

Fitch Ratings and S&P Global Ratings downgraded First Republic's credit rating, citing "a high proportion of uninsured deposits" from wealthy customers who are more likely to move their money elsewhere and a loan-to-deposit ratio of 111%, meaning that it had lent out more money than it had in deposits from customers.

robocat• 3 years ago

Sure, outliers will fail first. That doesn’t say much about whether there are vanilla banks about to fail.

dan1234• 3 years ago

So is this just a problem with the US banking system, or are banks in other regions failing too?

kzrdude• 3 years ago

Credit Suisse failed recently too

mjfl• 3 years ago

what about lehman brothers?

alvis• 3 years ago

At times, it's not the size that matters, but rather the quantity that makes a chain reaction

misja111• 3 years ago

Where are Lehman Brothers, Bear Sterns? And where are all the non US banks, such as e.g. Credit Suisse? They would make the other numbers in this graph dwindle ..

Tempest1981• 3 years ago

FDIC data only, but yes

pjc502• 3 years ago

They're not in the set of US depositor banks. (Lehman was an investment bank)

cyclecount2• 3 years ago

Lehman Brothers and Bear Stearns were non-FDIC insured investment banks.

bilekas• 3 years ago

I’m not familar with all the recent banks that are failing but is it a result of consolidation during the 2007 crisis?

Seems really strange visually at least

MrQuincle• 3 years ago

Makes me think of firebreaks. What are the equivalents of firebreaks in the sector?

crimsoneer• 3 years ago

The Fed acting as a backstop, essentially (like they did after Lehman in 2008)

alex-moon• 3 years ago

Wonder what year that massive spike is. Oh no, can't zoom in on mobile! Guess I'll have to check later on my desktop, if I remember. Until then, it is a mystery.

HatchedLake721• 3 years ago

I can zoom in on an iPhone.

It’s 2008/2009

dandare2• 3 years ago

The "Bank failures" title is misleading, when only FDIC banks failures are included.

_heimdall5• 3 years ago

"investment banks" are also misleadingly named in my opinion. This dataset seems to stick to what the more common idea of a bank that accepts deposits and offers checking/savings accounts.

dandare2• 3 years ago

Well, yes :)

sam_goody3• 3 years ago

And now that the fed said they will only "fully" back depositors in large banks that go under, there will be even more consolidation, so the next crash can consist of just one huge circle that acted gregariously and irresponsibly before its collapse, but no one could do anything because of its size.

What makes this even worse, is that Yellen would have lost nothing had she said they would back every bank and not just large banks - small bank failures are just not that common, and the risk is low, (and can be walked back in two years when the tension lowers).

Even worse, well over half of the SVB bailout went to <20 accounts.

IMO, they should have raised the amount that FDIC is insuring, which is unquestionably too low. [They still should do that.] They then should have insured everyone equally up to that point. And if Circle and Roku would lose 90% of their value from such a move, that beats the whole banking system consolidating into one or two massive players that can act with impunity.

OK, I am not an economist, and admit that I probably don't know what I am talking about, but HN is full of smart people with domain knowledge. Please (gently) broaden my perspective.

penguinten• 3 years ago

Even worse, well over half of the SVB bailout went to <20 accounts.

Do you have a source for this?

sam_goody3• 3 years ago

I don't have time to research now, but there were some well detailed breakdowns of the large payments when the bailout happened.

The first relevant link in DDG is a statement by the Chairman of the FDIC. He claims that the top ten accounts held more than 13.3B between them, which is more than 10% of the total of all deposits in the bank, and more than half of the $20B that the government is expected to payout in total. [1][2]

The FDIC would otherwise have had to pay a scant 3M of that, so there's 2/3 of the bailout just in those 10 accounts. If we include the next ten accounts, it will be much, much more. (One of the articles at the time claimed more than 85% went to 15 accounts, but I didn't want to be extreme without time to research sources.)

[1] https://www.banking.senate.gov/imo/media/doc/Gruenberg%20Tes... [2]: https://www.fdic.gov/news/press-releases/2023/pr23023.html

Ilverin• 3 years ago

The large banks are more heavily regulated. Trump raised the limit at which a bank is called large from 50bil to 250bil. The small banks value their freedom from regulation more than they would value full FDIC coverage. The fed would love to make that trade:more regulation of small banks in exchange for full FDIC coverage for small banks.

Canada is dominated by five enormous banks, they are closely regulated, and Canada has had a better experience than USA with both fewer bank failures and higher real interest rates paid to customers.

theironhammer• 3 years ago

Sure the Canadian banks are more stable but they are EXTREMELY conservative when it comes to lending out money. That's why the USA has a robust venture capital milieu while Canada does not.

In fact a Canadian bank would rather loan money to an American business to buy out a Canadian business than to lend the same Canadian business to expand.

refurb• 3 years ago

Canada has higher real interest rates paid to customers? You sure? My Canadian bank account pays pretty much what my US one does, 0.00001% or something similar.

mywittyname2• 3 years ago

What makes this even worse, is that Yellen would have lost nothing had she said they would back every bank and not just large banks - small bank failures are just not that common, and the risk is low, (and can be walked back in two years when the tension lowers).

FDIC is supposed to protect smaller individual depositors, not gigantic companies with hundreds of millions of dollars in deposits. Gigantic companies have financial teams that should be tasked with protecting their operating capital.

FDIC is also a last-resort. The idea is that bank assets are sold off and all depositors are made whole. It's very difficult and time-consuming to do this with a mega bank though.

Maybe FDIC limits should be raised, but the problem with that is someone needs to pay for it. Perhaps companies with $100MM deposits should be paying for private insurance.

roflyear3• 3 years ago

I'm not sure I understand your argument, because it is ONLY larger banks that are failing! Not the uber-big but larger.

99112000• 3 years ago

Now plot it relative to the money supply :D

wufufufu• 3 years ago

Can someone explain why this is being downvoted?

daliz• 3 years ago

The world is changing. And that's good news.

alkonaut• 3 years ago

This honestly looks much less frightening that I would have thought. But then my experience of '08 was "that wasn't so bad?" which I understand is highly regional. Visited some colleagues in 2015 in the UK who kept referring to "the recession" when talking about struggling customers or abandoned buildings. I had to ask whether they had experienced another recession or whether they still meant the 2008 one...

itsthecourier• 3 years ago

Where is Lehmann Brothers there?

torcete2• 3 years ago

Ok, this must be a silly question but... What programming language/library is that?

I am used to R + ggplot2 and I am aware that are quite a variety of other tools out there. In this case I unable to identify which one he is using.

virtuallyvivek• 3 years ago

JavaScript and the library is Observable Plot

Link: https://observablehq.com/plot/

aketchum• 3 years ago

The person who wrote this blog also wrote the very popular and famous D3 visualization library for Javascript. It also looks like this is javascript from the code.

torcete2• 3 years ago

Ah, thanks. I can see it is also based on the grammar of graphic. Neat!

WillAdams• 3 years ago

I only saw a couple of entries which might have been credit unions --- were they intentionally excluded? Or are they that much less likely to fail?

herbcso• 3 years ago

Credit Unions aren't FDIC insured, which is what this data set is using.

everybodyknows• 3 years ago

Excellent point. To Joe Consumer, bank vs credit union accounts are ready alternatives. Where might one find data on failures? A quick scan of this site offered no clues.

https://ncua.gov/

Insurance blurb from NCUA:

https://ncua.gov/files/publications/guides-manuals/NCUAHowYo...

They do have access to a high-margin business, "temporaily":

https://ncua.gov/regulation-supervision/letters-credit-union...

Lending at up to 18% is permitted.

stuff4ben• 3 years ago

Is it me, or does this kind of illustrate that the banking industry doesn't know what it's doing and cannot be trusted in the future? I don't know what the answer is, but these "fat cats" running banks really aught not to.

conradev• 3 years ago

The graph seems to be in nominal dollars, but I feel like real dollars is the only way to get an accurate comparison of magnitudes

boringg2• 3 years ago

Way easier to diagnose & save 3 banks than the multitudes in 2008 from a time spent analysis perspective (wether or not we should is a different matter). The question is the ripple effect here - it seems to have stemmed 3 large breaches where as in 2009 it was death by a thousand cuts (with some large breaches).

chunk_waffle• 3 years ago

So we're supposed to just go about our day, anxiety free, after seeing this?

abeppu2• 3 years ago

Mike Bostock knows far more about data visualization than I do so I assume there are good reasons for choosing this presentation. But it strikes me as hard to think about, so I must ask: when is stacked+packed circles a good choice?

The drawbacks to me seem to be:

- At any given point on the x-axis, the associated height of the pile is partly taken up by circles but is partly space lost to circle packing. This proportion is hard to estimate visually.

- Because circles extend in the x-direction, they contribute to height over a _variable_ time range. E.g. WaMu is still the tallest circle at the 2010 line. It's a bit like a kernel density plot where the _kernel_ is data dependent.

Because of these effects taken together, though it's tempting to think of the profile of the pile as a smoothed average of the rate of bank-failures, but this is pretty misleading, which I suppose is why there's no labeled y-axis.

reylas• 3 years ago

The way I understand the graph is that is not the "number" of circles, but the area (money) that they consume. So those three bank failures are taking up the same space as all those other failures from 2008.

It is the total volume of money involved vs the count. The height of both stacks show just how much damage three did vs all the others in the left hand stack.

abeppu2• 3 years ago

Yeah I get the count vs area distinction, but I think the point would have been more cleanly communicated with e.g. a stacked barchart with failures by quarter. With circles, the filled area over any span of the x-axis is not solely due to failures which happened during the corresponding time period. So the area above a short span in early 2010 includes area contributed by WaMu, which happened in Sept 2008.

rout395742• 3 years ago

My read of it was that the circles are centered on the date of the failure, and the stacking was some flavor of minimum height. The result is you get an approximation of "area under the curve" of bank value that has failed, while retaining a sense of whether a particular peak was a few large actors, or a mound of tiny. I find that clear distinction valuable.

The nice thing about Mike having posted the code is that, if you think there's a better way to portray it, the canvas is spread out before you. :)

hattmall• 3 years ago

Why so many in Georgia? Roughly 25% of all bank failures were in a state with less than 3% of population? Subprime lending?

majormajor2• 3 years ago

I've seen a bit of discussion about concerns around office property mortgage and such (e.g. JP Morgan wanting to be insulated about losses on outstanding loans) for FRB which seem concerning as a post-Covid systemic risk.

I wonder if the consumer deposit side, though, has a lot to do with an unanticipated risk of the particular business model SVB and FRB were using: higher-income people. Sounds great - less risk of defaulting on loans - but then you are highly-concentrated in larger deposits for a smaller total population (larger meaning less insured, so easier to spook; smaller population means the panic doesn't have to spread as far)... so it works well for decades and is seen as something to aspire to... until it suddenly doesn't.

danlugo92• 3 years ago

Great, the more the system cracks the sooner the veil will come down.

Bank Failures Visualized

Related Stories