The problem for these smaller banks isn't just that interest rates went up. It's that they had a flood of deposits when interest rates were extremely low, mainly due to the various stimulus programs. The smaller banks couldn't grow their loan portfolio at a comparable rate to the increase in deposits so they had to buy more bonds. Then with swollen bond portfolios, rates went up causing large unrealized losses and at the same time, deposits were shrinking as the excess cash was disappearing. SVB is an extreme version of this story but I've been hearing about community banks feeling this funding stress since last fall.
SVB bought lots of long duration bonds at close to the lowest interest rates in history. That was a conscious choice. They made a decision to chase yield ignoring the very high likelihood that the short and long-term interest rates would rise significantly in the next couple of years. At the time they were buying these bonds, we were scrambling to get through the bank approval process to get our mortgage refinance locked in at historically low fixed rates (sub-3% for 15 years) because I assumed those rates would not be available again for many years and I expected them to snap back to "normal" levels by late 2021. Apparently, we were on the opposite side of the "smart money."
At low interest rates like 2020-21, the only reason to buy long bonds is to hold them for a specific purpose for a time that matches their duration. That is why annuities cost a lot for a given amount of income in 2021 - the life insurance companies would need to fund them with low interest rate bonds matching the durations of the mortality period. Annuities are much better priced today.
SVB wanted to report decent earnings in 2021 and did nothing to manage the duration risk of what they had bought which is astonishing. Once the wealthy "sophisticated" investors with lots of uninsured money in the bank figured this out a couple of weeks ago, they ran for the door crushing the bank and regulators on the way. The bank, depositors, and regulators showed high levels of incompetence and obliviousness. FDIC appeared to perform well, but they are the ones that come out with the shovels and wheelbarrows after the elephants finish their show.
I think a key lesson for the regulators out of this is that banks with a high percentage of assets above FDIC limits will need to be regulated to have much higher levels of low duration assets that they can sell quickly without taking a significant loss. Or, just insure all bank deposits and charge higher FDIC premiums to the banks - that will not make the well-run banks happy as they will have to reduce their profits to cover the costs of the idiots.
Of all the risks in banking I've seen cataloged, I can't remember ever seeing 'too many deposits' included . I think typically banks with excess reserves just lend them on the overnight federal funds market.
I can't see how the FDIC can make a credible commitment not to bail out depositors in future (they didn't bail out Indymac, but that brief period of bravery ended with the GFC). So, it would be better to make the guarantee explicit, and regulate accordingly (roughly, turn them into public utilities).
SVB of course was involved in lobbying to de-regulate mid-size banks like itself, a few years back. This sort of failure was a predictable outcome, at the time, if people had spent even a few moments thinking about how, as Noah describes, their business model made them extra risky if interest rates ever went up. Hopefully we will hold the line on making their management and stockholders suffer losses. (There's even an argument for making non-depositor lenders, if they have any, take a haircut. If they had bonds, the folks under-writing those bonds ought to have assessed the riskiness of the business more clearly.)
There would still have been a fundamental blind spot for the regulators. since they were holding TBonds and government backed MBS, those bonds could be considered by the regulators as hold to maturity and not marked to market, so regarded as assets at par value. In reality, they ran into a cash flow crunch when the depositors started asking for money and then exposed themselves as selling assets at a loss impairing their capital. That is when the "sophisticated" VC investors bolted for the door and sucked out cash at an astonishing rate. There are probably very few banks that could withstand having 20% of their deposits transferred out in one business day without emergency lending from the Fed. If only Mary was there waving a handful of honeymoon cash to save the day....
So the relationship between HTM and MTM assets needs to be looked at very hard now that there is a real world example of a "stress test" of what a modern bank run looks like. My understanding is that this scenario is not part of the vaunted stress testing protocols in Dodd-Frank.
Yes- Holding company bond holders seem very likely to lose money (rightly so), possibly bank-level lenders and bond holders also- a reason why every publicly traded liability instrument in the banking sector below depositors in the retroactively changed waterfall has had its market price adjusted downward
“Without borrowing short and lending long, there really isn’t much of a reason for a bank to exist!”
The job of a bank is to manage the RISK of borrowing short to lend long. SVB appear to have done the thing that is the obvious banking option, but with:
- much longer dated long lending than is in any way sensible for an institution with 100% demand deposits - 1 year treasuries are still “lending longer” than overnight deposits but are much less exposed to rate moves; and
- zero risk management of the risk by swapping those longer rates to floating.
By lending long and swapping to floating could still have achieved additional margin due to the longer-term commitment, but would have had a gain on the derivative to offset the losses on the bond portfolio.
It really is banking risk management 101 and to have done none of it seems utter madness.
Summers' comment, as written, is indeed stupid. "Borrowing short and lending long" is in fact what banks DO. Of course they are supposed to manage their risks in doing so (something that SVB seems to have inadequately, to be charitable), but that is a different sort of issue.
Fair; I would give him the benefit of the doubt that (even extended) tweet enforced brevity meant the “(managing the risks of...) borrowing short and lending long” was elided, but on a strict read of his tweet I agree.
I think your charitable assessment of their risk management processes is, yes, charitable!!
It does also make an interesting discussion case for impact of regulation: Following Trump reforms banks of SVB’s size didn’t need to take the impact of unrealised losses to their regulatory capital, which is essentially their buffer for risk. By removing that risk from the regulatory capital, it definitely feels like SVB decided it was a risk they didn’t need to worry about at all, which seems foolish. Tricky game that regulators have to play!!
The Fed knew what SVB was doing and it wasn’t ordinary course of business. Was on an unprecedented scale for SVB or for any bank (increasing bond portfolio from 10 billion to 100 billion in two years) and I can assure you the Fed doesn’t need a regulation to throw the gauntlet down to a bank over unusual actions and ask them for the policies and oversight procedures that govern and control these actions. Macro prudential regulation is as much about consultation and questioning as it is about a rule book (and I don’t see that the 2018 stress testing rule changes would have made a huge difference). It doesn’t take a genius to note that fair value MTM losses (which occurred AFTER the bond portfolio had already been built up) exceeded tangible common equity. The Fed and FDIC allowed them to keep taking deposits regardless.
Surely preventing SVB taking deposits in virtually indistinguishable from just causing bank run and/or putting SVB into receivership directly? Or is that your point as to what the Fed should have done when SVB showed MTM insolvency?
1) The Fed should have stopped them from building up an unhedged l.t. bond portfolio of this size funded by wholesale-sized overnight commercial deposits sourced from firms designed to burn through that cash.
2) The bond market meltdown started in Q2 2022, by which point the bulk of SVB’s portfolio had been built. Stress testing rules looking at fair value losses vs reg capital would have only shut the door after the horse had bolted.
Nothing prevented the Fed from doing the same (ie questioning the portfolio losses and recommending actions- hedging, raising equity, sourcing long-dates deposits, shortening bond maturity) after Q2 2022, stress test or no. They can read a balance sheet. It seems likely the Fed did some of this, but not urgently enough. And yes, pursue an orderly liquidation or sale sooner rather than later. A complication may have been that this bank had friends in high places (typical for a big bank, less so for such a small institution)
3) Stress testing and liquidity testing focused on recession and credit crisis risks - scenarios where interest rates fall. SVB would have looked good under most scenarios. The regulators have been fighting the last war (2008 GFC scenarios) through CCAR and stress testing.
4) The rule change that could make the most sense is applying the NSFR more broadly across the banking sector and requiring banks to hold more short term assets against deposits. The Fed/FDIC has exempted banks with less than. $50b in wholesale funding from this rule, but in my view they should include institutional-sized (>250k fdic threshold) as wholesale funding.
From a distant observer, the Fed and Janet Yellen seem to have handled this with aplomb, and we rubes who aren't banking mavens should be thankful. All of the 'how I would have solved this crisis before it became a crisis' in the above post come across as arm-chair quarterbacking. 'Hedging' (like hedging what and what risk and at what cost? Dunno and do you?) 'Raising equity'? Apparently they couldn't at a cost acceptable to their shareholders. 'Sourcing long-dated deposits'? In the banking industry that's called stupid money and apparently it disappeared. 'Shortening bond maturity', sure go ahead but the last I looked maturities and rates kind of move in tandem. What's more interesting is how did a well-capitalized bank in Silicon Valley blow up one weekend. Something in the weed? There's more to this story and we're all hoping for a clever journalist to tell us what. What caused Silicon Valley to make a run on their bank?
Now that the government has shown that it will save the banks from risk doesn’t that make it more likely that the banks will just keep taking more risks?
Not quite, because the government did not save the owners of SVB, only the depositors. The shareholders were wiped out. Hopefully, regulators will claw back some of the executive bonuses too. That will help dissuade the next set of bank executives from taking on so much interest rate risk in search of higher yield.
Nice article. Whatever anyone thinks of the yuan or BRICS et cetera, it's clear that the dollar is in for tough times ahead. The Fed is caught between Scylla and Charybdis.
A strong dollar, boosted by higher interest rates, is beneficial to the middle class and the poor who retain substantial amounts of their capital in cash. A strong dollar also, of course, makes imports cheaper. Of course, the most important benefit is it helps counter inflation.
But the era of zero interest rates went on for too long( the pandemic clearly didn't help matters). It meant that regional banks all over the country were sitting on massive piles of unrealized losses. Tech, the industry which benefits most from zero interest rates, was given a full decade of it. What did they do with it? Well, the honest truth is other than Elon Musk( who I have to concede built actual useful stuff), hardly anything that matters.
Captured by venture capitalists, The tech industry has spent all of that time and resources promoting the gig economy to disrupt traditional labour, crypto in the absurd hope that institutional gambling and Ponzi schemes constituted some kind of financial alternative, and the metaverse to do precisely what exactly?
They have squandered a golden opportunity to actually build the next era of technology on American shores ( perhaps LLMs make a difference: the jury is still out there).
The decision to freeze Russia's reserves, go after Oligarchs' assets regardless of due process, and weaponize the financial system will come in time to be remembered as one of the greatest geopolitical own goals of all time. It means that the elite all over the world are more reluctant now to purchase western assets which of course drives dollar demand down, it means China can and is entrenching itself in the middle east and I can't draw out all of the massive implications there, and it means America has unwittingly made itself the midwife of an alternative financial system, albeit one with very different goals and strengths.
So, the Fed can't afford to keep raising interest rates: it will break several of American industries( banking and tech most clearly). It can't afford to not keep raising interest rates ( if inflation is not countered properly, the consequences for the median American are steep)
As it is, Biden by endorsing the right decision to make depositors whole( ironic that 'independent risk taking libertarians' panicked and clamoured for federal intervention so easily), may have damaged his own chances of reelection. The collateral damage of this decision will be vast. But, in keeping with the grand old rules of capitalism, the venture capitalists will escape much of it.
You are failing to account for all of the many deep flaws present in other economies. The dollar is the world reserve currency because all the other options are worse.
I agree that all the other options are worse. I don't think that's up for debate.
What is up for debate is the logical inference between every other option being worse and the dollar remaining the world's reserve currency.
Two things can, and in my view, will be true: there's no global currency alternative and the dollar's hegemonic status will suffer a substantial decline.
We have gotten fairly used to the 'one currency to rule them all': it was, for so long, the gold standard. Then the dollar took up the mantle. But there will be more bilateral trade agreements settled in local currencies: Brazil and Argentina are already proposing a joint regional currency ( a deeply flawed idea but a sign of the times), others will surely follow.
In addition, a significant source of demand for the dollar all over the world is to purchase oil. That demand will fall as the Middle East pursues closer ties with China and renewable energy, which doesn't tether countries to dictatorships but also the 'petrodollar', gains widespread adoption.
You must also consider that American decoupling from the factory of the world will have its own negative effects on the dollar( China doesn't need to earn as many dollars as it needed to purchase the inputs it needs, most especially raw materials from Russia).
And that's all before we consider what a conflict over Taiwan will do to the dollar. If China's staggering reserves of American financial assets - bills, bonds, and cash - are either frozen or are dumped on the market, the consequences will be unimaginable.
A multipolar world and a singular global reserve currency are incompatible. No other currency can replace the dollar. That is true and probably always will be. But the statement is incomplete: no other currency by itself. That is the real risk. Or, depending on where you stand, the real opportunity.
“ If China's staggering reserves of American financial assets - bills, bonds, and cash - are either frozen or are dumped on the market, the consequences will be unimaginable”
How so? They’d be confiscated like German assets were seized during WWI. I think there are still two Mercks as a result. (Or they may have merged back into one company a few years ago.)
No, they have not merged afterwards. Quite interesting bit of trivia you had there. Although it's worth noting that the German owner acquired the American company almost immediately after the auction of the assets. So really,not much changed there.
China owns around a trillion dollars of Treasury bills. It possesses nearly four trillion dollars of dollar reserves. It is the largest foreign owner of those financial assets. If they were confiscated, it would constitute the decisive termination of the dollar as the global reserve currency. Other countries would readily dump their own reserves and treasuries. The faith and trust in the current financial system will be shaken beyond repair.
The immediate consequences would be an immediate shock of high inflation and a raft of bankruptcies for banks, hedge funds, and other institutions in the financial system.
Geopolitically, America would be de facto surrendering its pre-eminent position permanently. Politically, extreme populists would capitalize on the chaos to propose and inspire dangerous policies. Those are merely the immediate and obvious consequences.
There will, of course, be a bunch of others which will only reveal themselves upon I agree that all the other options are worse. I don't think that's up for debate.
What is up for debate is the logical inference between every other option being worse and the dollar remaining the world's reserve currency.
Two things can, and in my view, will be true: there's no global currency alternative and the dollar's hegemonic status will suffer a substantial decline.
We have gotten fairly used to the 'one currency to rule them all': it was, for so long, the gold standard. Then the dollar took up the mantle. But there will be more bilateral trade agreements settled in local currencies: Brazil and Argentina are already proposing a joint regional currency ( a deeply flawed idea but a sign of the times), others will surely follow.
In addition, a significant source of demand for the dollar all over the world is to purchase oil. That demand will fall as the Middle East pursues closer ties with China and renewable energy, which doesn't tether countries to dictatorships but also the 'petrodollar', gains widespread adoption.
You must also consider that American decoupling from the factory of the world will have its own negative effects on the dollar( China doesn't need to earn as many dollars as it needed to purchase the inputs it needs, most especially raw materials from Russia).
And that's all before we consider what a conflict over Taiwan will do to the dollar. If China's staggering reserves of American financial assets - bills, bonds, and cash - are either frozen or are dumped on the market, the consequences will be unimaginable.
A multipolar world and a singular global reserve currency are incompatible. No other currency can replace the dollar. That is true and probably always will be. But the statement is incomplete: no other currency by itself. That is the real risk. Or, depending on where you stand, the real opportunity. a crisis. The outcome, to put it mildly, would be grave
“Other countries would readily dump their own reserves and treasuries. The faith and trust in the current financial system will be shaken beyond repair.”
Sure, Japan, Germany, Australia would bail. It’s simply not going to happen. It didn’t happen when we seized Russia’s assets - what was that $400 billion? Not so much as a ripple.
Russia became the equivalent of Nazi Germany and Japan in WW II when they invaded Ukraine. If China invades Taiwan, they could become equivalent. Otherwise, it is just a Cold War type of scenario where you have to pretend to like each other.
“But I think Summers’ argument just doesn’t make sense here.”
Small wonder, because Summers has been wrong on important policy for decades. He’s arguably the worst applied economist of his generation. It baffles me as to why people can’t find a better economist to quote. Summers’ thinking should not be the Good Housekeeping Seal of approval for economic columns.
He was also at the heart of repealing Glass-Steagal in 1999 which allowed us all to learn far more about our financial system than we wanted to in 2008. so very in keeping with his academic background.
I think maybe some people are being unfair. You're welcome to think Larry Summers is a jerk. But he applied a pretty standard analysis comparing the size of the fiscal expansion to the size of the output gap and arrived at a pretty orthadox conclusion. That being that the expansion was inflationary. This was the same analysis that said that the 08 stimulus was inadequate. I think history shows both to be accurate and pretty conventional assessments. Im not sure history will look so kindly on Team Transitory.
"I think there is about a one-third chance that inflation will accelerate significantly over the next several years, and we’ll be in a stagflationary situation, like the one that materialized between 1966 and 1969, where inflation went from the range of ones to the range of sixes. I think there’s a one-third chance that we won’t see inflation, but the reason we won’t see it is that the Fed hits the brakes hard, markets get very unstable, the economy skids downward close to recession. And I think there’s about a one-third chance that the Fed and the Treasury will get what they’re hoping for, and we’ll get rapid growth that will moderate in a non-inflationary way."
That's quite a few scenarios. But bottom line he says there's a two-thirds chance we'll be in a stagflationary situation (high inflation, high unemployment) or we get a recession.
Two years later we haven't had a recession, employment is good, and inflation is higher than we like.
We'll see what happens in the future, but I'd say his prognostication is looking less prescient over time.
Not to beat a dead horse (although it's probably too late), but I think if you read that quote carefully, he's saying the American Rescue Act (it's release was the context in which it was made) had a 2 in 3 chance of being inflationary, and the course of that inflation depended on how the fed reacted, which I don't think anyone could predict (are they going to raise rates this afternoon? by how much?)
I think if you compare it too a lot of the left commentary at the time, I think he got the big question right. That inflation was coming, and that it was not a temporary supply side issue that would take care of itself, but rather due to a hot economy that required intervention. As for the feds response, I think they ran with scenario one for a good year or so, then shifted to scenario 2. Perhaps they will pull off their soft landing, but we are clearly not out of the woods yet.
Just to be clear, I'm not some Larry Summer super-fan that thinks he has super human powers, but I also don't think he's the comic book villain that some people like to portray him as. I do think he's a serious and accomplished academic that you ignore at your peril. We see on the right what happens when punishing heretics is valued over learning from your mistakes. I don't want those on the left to make the same mistake.
Summers was clearly more correct about inflation than, e.g., Paul Krugman and Brad DeLong. Fine if he had stopped there. But then he went too far in laying out scary scenarios. You don't have to do that! For example, Jason Furman was even more accurate than Summers (let alone PK and BDL) and he didn't indulge in frightening scenarios.
I too think Summers is very smart, though perhaps not as smart as he thinks he is. (E.g., lecturing the Obama administration in 2009 on what size of a stimulus package would be politically feasible. Stick to economics, Larry!)
To say Treasuries are subject to interest rate risk from the fed ignores their inherent risk. They are subject, like all dollar denominated bonds, to INFLATION risk. And as the largest institution, the government is also ironically the largest DRIVER of said inflation, whenever they make a poor investment decision with the funds they are loaned - investments that don't return good economic growth.
When that government spending investment returns poorly, resources have been wasted and supply shortages occur. This causes the price inflation that devalues bonds before the fed even tries to raise rates to stop it.
When you invest in Treasuries, you are investing in the government spending that investment wisely.
Ahhh, another economic event, another cycle, and there will be more acts to the American financial play until democratic economics eventually replaces autocratic economics. The only clarification is that there are multiple autocrats, not just one on top. It’s a whole class on top who work to perfect the sly skills of wealth while appearing to play by the rules. The drastic income gap in this country is not an accident — the SVB boys collected theirs while walking out the door. This sly winning is helped by the inscrutability of economics; Larry Summers just isn’t very good at it. But the human supply chain is wearing down, so maybe more fundamental change is in store? Hopefully soon . . . .
Maybe this point in American history is where we get our dose of monarchy? Financial monarchy, capitalistic as we are. Going off the tracks or not will depend on the quality of education that is enacted going forward — we stand a chance if it is based on human literacy. Europe discovered Bildung and rescued themselves. We should be so lucky — today it is called Human Ecology, the proper term for an honorable proper human ecosystem.
The Fed is trying to break the economy to reduce inflation. Mind you, it won't break these formerly
Libertarian turned Socialist VCs, no it'll save them--its waiting to break the middle and poor classes to quell inflation. A good portion of the crap in the system was just bailed out like 2008. We are a top heavy country now--is it any coincidence that our deficit has skyrocketed and productivity plummeted along with the pigs learning to walk on two legs?
"And if you’re not going to borrow short and lend long, why are you a bank in the first place?"
Banks do provide another important service (the service that SVB's and other banks depositors are likely most interested in): they allow individuals and businesses to store and transmit money electronically without having to manage piles of cash. The banks could even charge a fee for this service with no lending required.
That is what banks started out doing. Goldsmiths were the original bankers because they already had safes and security guards, so people paid them to safely store their money. Then they realised they had all these deposits (gold in those days) sitting doing nothing. Why not lend some out and earn interest? Who would know? So greed took over.
You can get away with it because no-one really knows if the loan is from the goldsmith's personal gold (=bank capital) or the depositors' gold. But if you get too greedy and loan money to people who don't pay you back then one rumour can start a bank run and your business is finished.
Over time the lending out practice became more commonplace and accepted....but greed always wins out in good times so bank runs have been going on for a very long time. I think the reason that the practice has thrived and become normal is that lending out stored wealth means it can be put to good use and build stronger communities. Unfortunately lending for useless or harmful things can also happen and the temptations of greed and corruption are always there.
Thanks Noah for this. Reading all the hot takes on this over the last several days was driving me nuts. Seemed like everybody wanted to use this as an excuse to ride their favorite hobby horse (woke capitalism, annoying tech bros., etc). Even the usually reliable Derek Thompson seemed to be getting in on the action. I look forward to your follow-up articles
I'm more in the Larry Summers camp here. Managing the risk of borrowing short to lend long is the primary value add of a bank. Lots of other things intermixed along the way but that is the foundation. That we have a generation of bank managers who haven't experienced a rising interest rate environment does not excuse them from failing. In the SVB case this is exaggerated by their specific, intentionally sought (and created), market niche of focusing on start-ups. Startups have an unusual relationship with cash in the best of times and when times are stressful that gets even worse. SVB's role in risk management was to anticipate and manage that risk or alterntaively, participate in the broader market instead. So we have naive bank managers and naive startup founders (and startup CFOs) mishandling the risk surrounding tens, hundreds of millions of dollars and the bank fails when the very favorable risk environment rapidly changes. Not too much of a surprise.
There is a slope change in the depositories curve around 7/2018. Why?
I'm still thinking that a core tenet of macro theory is based on elastic response instead of many different time based viscous responses. I think this makes the Inflation fighting = k(interest rate), inaccurate and not as effective.
Our national and global economy is 10x more complicated than 1970s - 80s
Trump taxes are probably a good reason. Why does everyone forget that trump deliberately goosed the economy by bringing down taxes and then asking the fed to hold off on interest rate increases. The big levers: interest rates, taxes, government spending. He cranked them all to 11 in 2018.
I also think a good portion of that change is due to the Trump tax cuts. The graph is just Treasury holdings by depository institutions and the Trump tax cuts made holding municipal bonds relatively less attractive, since the lower tax rate makes tax-exempt interest less valuable, so I think at least some of the increase is banks substituting out of municipal bonds into Treasuries.
Noah, how can you not mention duration matching even once in this article? I believe this is the norm in banking, to reduce interest rate risk and essentially lock in a low risk spread on deposits. SVB deliberately chose not to do this. They were gambling that rates wouldn't go up as fast as they did in order to earn a higher spread. I think future regulations should force this type of gamble to be disclosed to depositors, more like surgeon's general warning on a cigarette carton, rather than a couple bullet points and a small table buried deep in a 10k.
The problem for these smaller banks isn't just that interest rates went up. It's that they had a flood of deposits when interest rates were extremely low, mainly due to the various stimulus programs. The smaller banks couldn't grow their loan portfolio at a comparable rate to the increase in deposits so they had to buy more bonds. Then with swollen bond portfolios, rates went up causing large unrealized losses and at the same time, deposits were shrinking as the excess cash was disappearing. SVB is an extreme version of this story but I've been hearing about community banks feeling this funding stress since last fall.
SVB bought lots of long duration bonds at close to the lowest interest rates in history. That was a conscious choice. They made a decision to chase yield ignoring the very high likelihood that the short and long-term interest rates would rise significantly in the next couple of years. At the time they were buying these bonds, we were scrambling to get through the bank approval process to get our mortgage refinance locked in at historically low fixed rates (sub-3% for 15 years) because I assumed those rates would not be available again for many years and I expected them to snap back to "normal" levels by late 2021. Apparently, we were on the opposite side of the "smart money."
At low interest rates like 2020-21, the only reason to buy long bonds is to hold them for a specific purpose for a time that matches their duration. That is why annuities cost a lot for a given amount of income in 2021 - the life insurance companies would need to fund them with low interest rate bonds matching the durations of the mortality period. Annuities are much better priced today.
SVB wanted to report decent earnings in 2021 and did nothing to manage the duration risk of what they had bought which is astonishing. Once the wealthy "sophisticated" investors with lots of uninsured money in the bank figured this out a couple of weeks ago, they ran for the door crushing the bank and regulators on the way. The bank, depositors, and regulators showed high levels of incompetence and obliviousness. FDIC appeared to perform well, but they are the ones that come out with the shovels and wheelbarrows after the elephants finish their show.
I think a key lesson for the regulators out of this is that banks with a high percentage of assets above FDIC limits will need to be regulated to have much higher levels of low duration assets that they can sell quickly without taking a significant loss. Or, just insure all bank deposits and charge higher FDIC premiums to the banks - that will not make the well-run banks happy as they will have to reduce their profits to cover the costs of the idiots.
Of all the risks in banking I've seen cataloged, I can't remember ever seeing 'too many deposits' included . I think typically banks with excess reserves just lend them on the overnight federal funds market.
I can't see how the FDIC can make a credible commitment not to bail out depositors in future (they didn't bail out Indymac, but that brief period of bravery ended with the GFC). So, it would be better to make the guarantee explicit, and regulate accordingly (roughly, turn them into public utilities).
This will be the subject of one of my 2 remaining SVB posts...
The private credit markets have boomed since 2010. HNW money getting 8% doing bank like things which banks were regulated out of.
SVB of course was involved in lobbying to de-regulate mid-size banks like itself, a few years back. This sort of failure was a predictable outcome, at the time, if people had spent even a few moments thinking about how, as Noah describes, their business model made them extra risky if interest rates ever went up. Hopefully we will hold the line on making their management and stockholders suffer losses. (There's even an argument for making non-depositor lenders, if they have any, take a haircut. If they had bonds, the folks under-writing those bonds ought to have assessed the riskiness of the business more clearly.)
There would still have been a fundamental blind spot for the regulators. since they were holding TBonds and government backed MBS, those bonds could be considered by the regulators as hold to maturity and not marked to market, so regarded as assets at par value. In reality, they ran into a cash flow crunch when the depositors started asking for money and then exposed themselves as selling assets at a loss impairing their capital. That is when the "sophisticated" VC investors bolted for the door and sucked out cash at an astonishing rate. There are probably very few banks that could withstand having 20% of their deposits transferred out in one business day without emergency lending from the Fed. If only Mary was there waving a handful of honeymoon cash to save the day....
So the relationship between HTM and MTM assets needs to be looked at very hard now that there is a real world example of a "stress test" of what a modern bank run looks like. My understanding is that this scenario is not part of the vaunted stress testing protocols in Dodd-Frank.
Yes- Holding company bond holders seem very likely to lose money (rightly so), possibly bank-level lenders and bond holders also- a reason why every publicly traded liability instrument in the banking sector below depositors in the retroactively changed waterfall has had its market price adjusted downward
Nothing more brilliant than the idea of our politicians running banks!
“Without borrowing short and lending long, there really isn’t much of a reason for a bank to exist!”
The job of a bank is to manage the RISK of borrowing short to lend long. SVB appear to have done the thing that is the obvious banking option, but with:
- much longer dated long lending than is in any way sensible for an institution with 100% demand deposits - 1 year treasuries are still “lending longer” than overnight deposits but are much less exposed to rate moves; and
- zero risk management of the risk by swapping those longer rates to floating.
By lending long and swapping to floating could still have achieved additional margin due to the longer-term commitment, but would have had a gain on the derivative to offset the losses on the bond portfolio.
It really is banking risk management 101 and to have done none of it seems utter madness.
There are two different issues here, I think.
Summers' comment, as written, is indeed stupid. "Borrowing short and lending long" is in fact what banks DO. Of course they are supposed to manage their risks in doing so (something that SVB seems to have inadequately, to be charitable), but that is a different sort of issue.
Fair; I would give him the benefit of the doubt that (even extended) tweet enforced brevity meant the “(managing the risks of...) borrowing short and lending long” was elided, but on a strict read of his tweet I agree.
I think your charitable assessment of their risk management processes is, yes, charitable!!
It does also make an interesting discussion case for impact of regulation: Following Trump reforms banks of SVB’s size didn’t need to take the impact of unrealised losses to their regulatory capital, which is essentially their buffer for risk. By removing that risk from the regulatory capital, it definitely feels like SVB decided it was a risk they didn’t need to worry about at all, which seems foolish. Tricky game that regulators have to play!!
The Fed knew what SVB was doing and it wasn’t ordinary course of business. Was on an unprecedented scale for SVB or for any bank (increasing bond portfolio from 10 billion to 100 billion in two years) and I can assure you the Fed doesn’t need a regulation to throw the gauntlet down to a bank over unusual actions and ask them for the policies and oversight procedures that govern and control these actions. Macro prudential regulation is as much about consultation and questioning as it is about a rule book (and I don’t see that the 2018 stress testing rule changes would have made a huge difference). It doesn’t take a genius to note that fair value MTM losses (which occurred AFTER the bond portfolio had already been built up) exceeded tangible common equity. The Fed and FDIC allowed them to keep taking deposits regardless.
Surely preventing SVB taking deposits in virtually indistinguishable from just causing bank run and/or putting SVB into receivership directly? Or is that your point as to what the Fed should have done when SVB showed MTM insolvency?
1) The Fed should have stopped them from building up an unhedged l.t. bond portfolio of this size funded by wholesale-sized overnight commercial deposits sourced from firms designed to burn through that cash.
2) The bond market meltdown started in Q2 2022, by which point the bulk of SVB’s portfolio had been built. Stress testing rules looking at fair value losses vs reg capital would have only shut the door after the horse had bolted.
Nothing prevented the Fed from doing the same (ie questioning the portfolio losses and recommending actions- hedging, raising equity, sourcing long-dates deposits, shortening bond maturity) after Q2 2022, stress test or no. They can read a balance sheet. It seems likely the Fed did some of this, but not urgently enough. And yes, pursue an orderly liquidation or sale sooner rather than later. A complication may have been that this bank had friends in high places (typical for a big bank, less so for such a small institution)
3) Stress testing and liquidity testing focused on recession and credit crisis risks - scenarios where interest rates fall. SVB would have looked good under most scenarios. The regulators have been fighting the last war (2008 GFC scenarios) through CCAR and stress testing.
4) The rule change that could make the most sense is applying the NSFR more broadly across the banking sector and requiring banks to hold more short term assets against deposits. The Fed/FDIC has exempted banks with less than. $50b in wholesale funding from this rule, but in my view they should include institutional-sized (>250k fdic threshold) as wholesale funding.
https://www.bis.org/bcbs/publ/d295.pdf
From a distant observer, the Fed and Janet Yellen seem to have handled this with aplomb, and we rubes who aren't banking mavens should be thankful. All of the 'how I would have solved this crisis before it became a crisis' in the above post come across as arm-chair quarterbacking. 'Hedging' (like hedging what and what risk and at what cost? Dunno and do you?) 'Raising equity'? Apparently they couldn't at a cost acceptable to their shareholders. 'Sourcing long-dated deposits'? In the banking industry that's called stupid money and apparently it disappeared. 'Shortening bond maturity', sure go ahead but the last I looked maturities and rates kind of move in tandem. What's more interesting is how did a well-capitalized bank in Silicon Valley blow up one weekend. Something in the weed? There's more to this story and we're all hoping for a clever journalist to tell us what. What caused Silicon Valley to make a run on their bank?
Exactly
Now that the government has shown that it will save the banks from risk doesn’t that make it more likely that the banks will just keep taking more risks?
Not quite, because the government did not save the owners of SVB, only the depositors. The shareholders were wiped out. Hopefully, regulators will claw back some of the executive bonuses too. That will help dissuade the next set of bank executives from taking on so much interest rate risk in search of higher yield.
Nice article. Whatever anyone thinks of the yuan or BRICS et cetera, it's clear that the dollar is in for tough times ahead. The Fed is caught between Scylla and Charybdis.
A strong dollar, boosted by higher interest rates, is beneficial to the middle class and the poor who retain substantial amounts of their capital in cash. A strong dollar also, of course, makes imports cheaper. Of course, the most important benefit is it helps counter inflation.
But the era of zero interest rates went on for too long( the pandemic clearly didn't help matters). It meant that regional banks all over the country were sitting on massive piles of unrealized losses. Tech, the industry which benefits most from zero interest rates, was given a full decade of it. What did they do with it? Well, the honest truth is other than Elon Musk( who I have to concede built actual useful stuff), hardly anything that matters.
Captured by venture capitalists, The tech industry has spent all of that time and resources promoting the gig economy to disrupt traditional labour, crypto in the absurd hope that institutional gambling and Ponzi schemes constituted some kind of financial alternative, and the metaverse to do precisely what exactly?
They have squandered a golden opportunity to actually build the next era of technology on American shores ( perhaps LLMs make a difference: the jury is still out there).
The decision to freeze Russia's reserves, go after Oligarchs' assets regardless of due process, and weaponize the financial system will come in time to be remembered as one of the greatest geopolitical own goals of all time. It means that the elite all over the world are more reluctant now to purchase western assets which of course drives dollar demand down, it means China can and is entrenching itself in the middle east and I can't draw out all of the massive implications there, and it means America has unwittingly made itself the midwife of an alternative financial system, albeit one with very different goals and strengths.
So, the Fed can't afford to keep raising interest rates: it will break several of American industries( banking and tech most clearly). It can't afford to not keep raising interest rates ( if inflation is not countered properly, the consequences for the median American are steep)
As it is, Biden by endorsing the right decision to make depositors whole( ironic that 'independent risk taking libertarians' panicked and clamoured for federal intervention so easily), may have damaged his own chances of reelection. The collateral damage of this decision will be vast. But, in keeping with the grand old rules of capitalism, the venture capitalists will escape much of it.
You are failing to account for all of the many deep flaws present in other economies. The dollar is the world reserve currency because all the other options are worse.
I agree that all the other options are worse. I don't think that's up for debate.
What is up for debate is the logical inference between every other option being worse and the dollar remaining the world's reserve currency.
Two things can, and in my view, will be true: there's no global currency alternative and the dollar's hegemonic status will suffer a substantial decline.
We have gotten fairly used to the 'one currency to rule them all': it was, for so long, the gold standard. Then the dollar took up the mantle. But there will be more bilateral trade agreements settled in local currencies: Brazil and Argentina are already proposing a joint regional currency ( a deeply flawed idea but a sign of the times), others will surely follow.
In addition, a significant source of demand for the dollar all over the world is to purchase oil. That demand will fall as the Middle East pursues closer ties with China and renewable energy, which doesn't tether countries to dictatorships but also the 'petrodollar', gains widespread adoption.
You must also consider that American decoupling from the factory of the world will have its own negative effects on the dollar( China doesn't need to earn as many dollars as it needed to purchase the inputs it needs, most especially raw materials from Russia).
And that's all before we consider what a conflict over Taiwan will do to the dollar. If China's staggering reserves of American financial assets - bills, bonds, and cash - are either frozen or are dumped on the market, the consequences will be unimaginable.
A multipolar world and a singular global reserve currency are incompatible. No other currency can replace the dollar. That is true and probably always will be. But the statement is incomplete: no other currency by itself. That is the real risk. Or, depending on where you stand, the real opportunity.
“ If China's staggering reserves of American financial assets - bills, bonds, and cash - are either frozen or are dumped on the market, the consequences will be unimaginable”
How so? They’d be confiscated like German assets were seized during WWI. I think there are still two Mercks as a result. (Or they may have merged back into one company a few years ago.)
No, they have not merged afterwards. Quite interesting bit of trivia you had there. Although it's worth noting that the German owner acquired the American company almost immediately after the auction of the assets. So really,not much changed there.
China owns around a trillion dollars of Treasury bills. It possesses nearly four trillion dollars of dollar reserves. It is the largest foreign owner of those financial assets. If they were confiscated, it would constitute the decisive termination of the dollar as the global reserve currency. Other countries would readily dump their own reserves and treasuries. The faith and trust in the current financial system will be shaken beyond repair.
The immediate consequences would be an immediate shock of high inflation and a raft of bankruptcies for banks, hedge funds, and other institutions in the financial system.
Geopolitically, America would be de facto surrendering its pre-eminent position permanently. Politically, extreme populists would capitalize on the chaos to propose and inspire dangerous policies. Those are merely the immediate and obvious consequences.
There will, of course, be a bunch of others which will only reveal themselves upon I agree that all the other options are worse. I don't think that's up for debate.
What is up for debate is the logical inference between every other option being worse and the dollar remaining the world's reserve currency.
Two things can, and in my view, will be true: there's no global currency alternative and the dollar's hegemonic status will suffer a substantial decline.
We have gotten fairly used to the 'one currency to rule them all': it was, for so long, the gold standard. Then the dollar took up the mantle. But there will be more bilateral trade agreements settled in local currencies: Brazil and Argentina are already proposing a joint regional currency ( a deeply flawed idea but a sign of the times), others will surely follow.
In addition, a significant source of demand for the dollar all over the world is to purchase oil. That demand will fall as the Middle East pursues closer ties with China and renewable energy, which doesn't tether countries to dictatorships but also the 'petrodollar', gains widespread adoption.
You must also consider that American decoupling from the factory of the world will have its own negative effects on the dollar( China doesn't need to earn as many dollars as it needed to purchase the inputs it needs, most especially raw materials from Russia).
And that's all before we consider what a conflict over Taiwan will do to the dollar. If China's staggering reserves of American financial assets - bills, bonds, and cash - are either frozen or are dumped on the market, the consequences will be unimaginable.
A multipolar world and a singular global reserve currency are incompatible. No other currency can replace the dollar. That is true and probably always will be. But the statement is incomplete: no other currency by itself. That is the real risk. Or, depending on where you stand, the real opportunity. a crisis. The outcome, to put it mildly, would be grave
“Other countries would readily dump their own reserves and treasuries. The faith and trust in the current financial system will be shaken beyond repair.”
Sure, Japan, Germany, Australia would bail. It’s simply not going to happen. It didn’t happen when we seized Russia’s assets - what was that $400 billion? Not so much as a ripple.
Russia became the equivalent of Nazi Germany and Japan in WW II when they invaded Ukraine. If China invades Taiwan, they could become equivalent. Otherwise, it is just a Cold War type of scenario where you have to pretend to like each other.
“But I think Summers’ argument just doesn’t make sense here.”
Small wonder, because Summers has been wrong on important policy for decades. He’s arguably the worst applied economist of his generation. It baffles me as to why people can’t find a better economist to quote. Summers’ thinking should not be the Good Housekeeping Seal of approval for economic columns.
He did accurately predict our current inflation problem
He was also at the heart of repealing Glass-Steagal in 1999 which allowed us all to learn far more about our financial system than we wanted to in 2008. so very in keeping with his academic background.
It seems to me he covered his bets by making a variety of predictions, some of which came to fruition.
I think maybe some people are being unfair. You're welcome to think Larry Summers is a jerk. But he applied a pretty standard analysis comparing the size of the fiscal expansion to the size of the output gap and arrived at a pretty orthadox conclusion. That being that the expansion was inflationary. This was the same analysis that said that the 08 stimulus was inadequate. I think history shows both to be accurate and pretty conventional assessments. Im not sure history will look so kindly on Team Transitory.
This is what Summers said in March 2021:
"I think there is about a one-third chance that inflation will accelerate significantly over the next several years, and we’ll be in a stagflationary situation, like the one that materialized between 1966 and 1969, where inflation went from the range of ones to the range of sixes. I think there’s a one-third chance that we won’t see inflation, but the reason we won’t see it is that the Fed hits the brakes hard, markets get very unstable, the economy skids downward close to recession. And I think there’s about a one-third chance that the Fed and the Treasury will get what they’re hoping for, and we’ll get rapid growth that will moderate in a non-inflationary way."
That's quite a few scenarios. But bottom line he says there's a two-thirds chance we'll be in a stagflationary situation (high inflation, high unemployment) or we get a recession.
Two years later we haven't had a recession, employment is good, and inflation is higher than we like.
We'll see what happens in the future, but I'd say his prognostication is looking less prescient over time.
Not to beat a dead horse (although it's probably too late), but I think if you read that quote carefully, he's saying the American Rescue Act (it's release was the context in which it was made) had a 2 in 3 chance of being inflationary, and the course of that inflation depended on how the fed reacted, which I don't think anyone could predict (are they going to raise rates this afternoon? by how much?)
I think if you compare it too a lot of the left commentary at the time, I think he got the big question right. That inflation was coming, and that it was not a temporary supply side issue that would take care of itself, but rather due to a hot economy that required intervention. As for the feds response, I think they ran with scenario one for a good year or so, then shifted to scenario 2. Perhaps they will pull off their soft landing, but we are clearly not out of the woods yet.
Just to be clear, I'm not some Larry Summer super-fan that thinks he has super human powers, but I also don't think he's the comic book villain that some people like to portray him as. I do think he's a serious and accomplished academic that you ignore at your peril. We see on the right what happens when punishing heretics is valued over learning from your mistakes. I don't want those on the left to make the same mistake.
Summers was clearly more correct about inflation than, e.g., Paul Krugman and Brad DeLong. Fine if he had stopped there. But then he went too far in laying out scary scenarios. You don't have to do that! For example, Jason Furman was even more accurate than Summers (let alone PK and BDL) and he didn't indulge in frightening scenarios.
I too think Summers is very smart, though perhaps not as smart as he thinks he is. (E.g., lecturing the Obama administration in 2009 on what size of a stimulus package would be politically feasible. Stick to economics, Larry!)
Seems like the obvious answer is some help fighting inflation from the fiscal side.
To say Treasuries are subject to interest rate risk from the fed ignores their inherent risk. They are subject, like all dollar denominated bonds, to INFLATION risk. And as the largest institution, the government is also ironically the largest DRIVER of said inflation, whenever they make a poor investment decision with the funds they are loaned - investments that don't return good economic growth.
When that government spending investment returns poorly, resources have been wasted and supply shortages occur. This causes the price inflation that devalues bonds before the fed even tries to raise rates to stop it.
When you invest in Treasuries, you are investing in the government spending that investment wisely.
Ahhh, another economic event, another cycle, and there will be more acts to the American financial play until democratic economics eventually replaces autocratic economics. The only clarification is that there are multiple autocrats, not just one on top. It’s a whole class on top who work to perfect the sly skills of wealth while appearing to play by the rules. The drastic income gap in this country is not an accident — the SVB boys collected theirs while walking out the door. This sly winning is helped by the inscrutability of economics; Larry Summers just isn’t very good at it. But the human supply chain is wearing down, so maybe more fundamental change is in store? Hopefully soon . . . .
Nice to know that there is at least someone who sees through the veil of deception.
Maybe this point in American history is where we get our dose of monarchy? Financial monarchy, capitalistic as we are. Going off the tracks or not will depend on the quality of education that is enacted going forward — we stand a chance if it is based on human literacy. Europe discovered Bildung and rescued themselves. We should be so lucky — today it is called Human Ecology, the proper term for an honorable proper human ecosystem.
The Fed is trying to break the economy to reduce inflation. Mind you, it won't break these formerly
Libertarian turned Socialist VCs, no it'll save them--its waiting to break the middle and poor classes to quell inflation. A good portion of the crap in the system was just bailed out like 2008. We are a top heavy country now--is it any coincidence that our deficit has skyrocketed and productivity plummeted along with the pigs learning to walk on two legs?
"And if you’re not going to borrow short and lend long, why are you a bank in the first place?"
Banks do provide another important service (the service that SVB's and other banks depositors are likely most interested in): they allow individuals and businesses to store and transmit money electronically without having to manage piles of cash. The banks could even charge a fee for this service with no lending required.
Wy not just have the Fed handle all the basic transactions, as proposed by Dean Baker?
That is what banks started out doing. Goldsmiths were the original bankers because they already had safes and security guards, so people paid them to safely store their money. Then they realised they had all these deposits (gold in those days) sitting doing nothing. Why not lend some out and earn interest? Who would know? So greed took over.
You can get away with it because no-one really knows if the loan is from the goldsmith's personal gold (=bank capital) or the depositors' gold. But if you get too greedy and loan money to people who don't pay you back then one rumour can start a bank run and your business is finished.
Over time the lending out practice became more commonplace and accepted....but greed always wins out in good times so bank runs have been going on for a very long time. I think the reason that the practice has thrived and become normal is that lending out stored wealth means it can be put to good use and build stronger communities. Unfortunately lending for useless or harmful things can also happen and the temptations of greed and corruption are always there.
Thanks Noah for this. Reading all the hot takes on this over the last several days was driving me nuts. Seemed like everybody wanted to use this as an excuse to ride their favorite hobby horse (woke capitalism, annoying tech bros., etc). Even the usually reliable Derek Thompson seemed to be getting in on the action. I look forward to your follow-up articles
I was similarly baffled when I read that tweet from Larry Summers.
I'm more in the Larry Summers camp here. Managing the risk of borrowing short to lend long is the primary value add of a bank. Lots of other things intermixed along the way but that is the foundation. That we have a generation of bank managers who haven't experienced a rising interest rate environment does not excuse them from failing. In the SVB case this is exaggerated by their specific, intentionally sought (and created), market niche of focusing on start-ups. Startups have an unusual relationship with cash in the best of times and when times are stressful that gets even worse. SVB's role in risk management was to anticipate and manage that risk or alterntaively, participate in the broader market instead. So we have naive bank managers and naive startup founders (and startup CFOs) mishandling the risk surrounding tens, hundreds of millions of dollars and the bank fails when the very favorable risk environment rapidly changes. Not too much of a surprise.
Fabulous.
There is a slope change in the depositories curve around 7/2018. Why?
I'm still thinking that a core tenet of macro theory is based on elastic response instead of many different time based viscous responses. I think this makes the Inflation fighting = k(interest rate), inaccurate and not as effective.
Our national and global economy is 10x more complicated than 1970s - 80s
Trump taxes are probably a good reason. Why does everyone forget that trump deliberately goosed the economy by bringing down taxes and then asking the fed to hold off on interest rate increases. The big levers: interest rates, taxes, government spending. He cranked them all to 11 in 2018.
I also think a good portion of that change is due to the Trump tax cuts. The graph is just Treasury holdings by depository institutions and the Trump tax cuts made holding municipal bonds relatively less attractive, since the lower tax rate makes tax-exempt interest less valuable, so I think at least some of the increase is banks substituting out of municipal bonds into Treasuries.
Thanks!
Thanks!
Noah, how can you not mention duration matching even once in this article? I believe this is the norm in banking, to reduce interest rate risk and essentially lock in a low risk spread on deposits. SVB deliberately chose not to do this. They were gambling that rates wouldn't go up as fast as they did in order to earn a higher spread. I think future regulations should force this type of gamble to be disclosed to depositors, more like surgeon's general warning on a cigarette carton, rather than a couple bullet points and a small table buried deep in a 10k.