Guest Contribution: “Some Thoughts on SVB”

Today, we present a guest post written by Charles Engel, Donald D. Hester Distinguished Chair in Economics at UW Madison. 


I’d like to share some thoughts on the Silicon Valley Bank (SVB) failure. These may help clarify the extent to which the rest of the banking system is vulnerable and help understand why the Fed/Treasury/FDIC undertook the actions they did. I don’t have too much knowledge about the inner workings of SVB – just what I’ve been able to gather from the internet, mostly from the excellent coverage in the Wall Street Journal and New York Times – so I may be somewhat off base on some of the facts.

Funding of SVB:

There are a lot of moving parts to the SVB case. Start with the funding of the bank. Apparently, SVB had a disproportionate share of its liabilities in the form of deposits by Silicon Valley tech companies, and these were mainly in large amounts (for example, an average deposit size of $4 million, on up to deposits over $100 million.)

In the past couple of years, the amount of deposits at SVB grew rapidly as tech companies came into cash for several reasons. Part of it was from funds from the government to support businesses during the COVID crisis. Much of it came from financing of these firms by venture capitalists.

Funding of a bank by larger scale depositors already confronts the bank with liquidity risk. When the bank loses funding unexpectedly because a depositor withdraws its funds, it needs liquid assets on hand to pay back the depositor, unless it can find new creditors that allow the bank to roll over its funding.

In the case of SVB, there was considerable additional risk because so many of its large depositors were in related businesses in the tech sector. That increases the risk that many depositors will need to reduce their deposits at the same time.

Indeed, that seems to be what happened. When the tech sector was flush with cash, they increased deposits at SVB, but as the tech sector boom waned in the past year, they began to withdraw their deposits. None of that constitutes a bank run – this should have been normal banking activity.

What went wrong?

SVB was essentially running a “carry trade” – they were taking in demand deposits that paid a low rate of interest, and investing them primarily in longer-term Treasury bonds, government agency bonds, and bonds issued by government sponsored enterprises that had higher promised interest rates. When long-term interest rates went up, these bonds lost value.

SVB became, literally, bankrupt or insolvent, if by “bankrupt” we mean that the net worth of SVB fell below zero. This is different than saying the value of SVB’s equity fell to zero.

To be pedantic, here is a simple example:

  • A bank is funded by $95 of deposits and $5 of its own money (its initial net worth.)
  • The bank’s plan is to use the $100 to buy 1-year government bonds that pay 2% interest, and then to pay depositors a rate of 1%. (I’m using 1-year bonds in this example to keep the math simple, but SVB’s bonds were more like 10-year bonds.)
  • At the end of the year, under this plan, the bonds pay off $2. The bank pays $0.95 to depositors (the 1% return on the $95 of deposits), and earns $1.05, which is a nice 21% return on its initial net worth.
  • But suppose the market interest rate on 1-year bonds rises to 8%. Then these bonds, which pay off $102 at the end of the year are now worth $102/1.08 = $94.44. Now the bank does not have enough funds to pay back its depositors if they want their money right now, because they are owed $95. The bank is bankrupt in the sense that its net worth (the value of its assets minus its liabilities) has fallen below zero.
  • However, if we can be sure the depositors won’t want their money back until a year from now, the bank will have $102, and it can pay back its depositors and make its 21% return. The bank’s stock will be valuable, even though its net worth is negative right now.

As you can see from this example, the value of the bank stock depends on how sure the owners of the bank are that the depositors won’t want their money back before the bonds held by the bank mature.

Liquidity and Hedging Risks

In real life, banks estimate how likely it is that depositors will make withdrawals that are not replaced by other depositors and keep liquid assets on hand to satisfy those depositors. In fact, they need to keep enough liquid assets on hand to be able to satisfy depositors even if there are fairly large, unexpected withdrawals. By “liquid”, I mean assets that are short-term, so their value does not fluctuate, and which can be turned into cash at low cost and with a high degree of certainty.

SVB was not doing this. If SVB were more prudent, it would have realized that its deposits were volatile, and would have kept more liquid assets on hand, such as short-term government bonds or reserves held at the Federal Reserve that would not have lost money when interest rates went up.

In addition, the SVB could have bought “insurance” against interest rate increases – that is, it could have reduced some of its vulnerability to interest rate increases by buying options. Although I have not seen any reliable reporting on this, I have seen unreliable reports that SVB did not hedge interest rate risk very much.

Also, perhaps SVB allowed itself to grow too quickly, making their balance-sheet risks harder to manage.

In short, SVB was gambling that interest rates would not go up much. They were not hedging interest-rate risk prudently, and they did not have enough short-term assets. If interest rates had not risen, SVB could easily have handled the decrease in their deposits from the tech sector, because they were holding bonds that are traded in very deep markets (so, at least their assets were liquid in that way.) If interest rates had not gone up, those bonds would not have lost value.

The bottom line was that SVB was running a risky business, betting that interest rates would not go up much. I say this with 20-20 hindsight, of course. Sometimes even very safe bets go wrong, but it does seem to me that the risks on SVB’s balance sheet were pretty apparent.

Is this a “classic bank run”?

I would say not, though I don’t want to get in the business of quibbling over definitions.

Here is the distinction I would make:

In a classic bank run, the bank may be forced to sell assets at a “fire-sale” price. By a fire-sale price, I mean the assets need to be sold at a price below what they would bring in a normally functioning financial market where potential investors have deep pockets and there is not a lot of hidden information about the value of the asset. For example, suppose a bank’s portfolio consists primarily of loans to local businesses and households. There are two circumstances where a fire sale could occur. One is if the bank is forced to sell off its assets (to pay off depositors) but potential buyers have difficulty valuing the assets. The bank maybe has made good loans, but potential buyers of those loans might not be able to assess the value of the bank’s portfolio on short notice. If the bank needs cash quickly, it might agree to sell off the assets for less than they are worth. A second case in which fire sales occur is if, even if the value of the assets is understood by potential buyers, there are not enough deep-pocketed investors to buy off the assets. This might happen in a time of widespread financial distress.

Note that neither of these conditions hold for SVB’s portfolio. It was holding Treasury bonds or other bonds that are traded in deep markets. The problem was not that SVB needed to sell off its assets at a fire-sale price. Its problem was that the value of those assets really had fallen.

In a classic bank run, the bank is solvent in the sense that its net worth is positive. The bank run is precipitated when depositors start believing – maybe based on rumors – that something is wrong with the bank and begin to withdraw their money. The bank is forced to sell their assets at fire sale prices. If enough depositors withdraw their money, the bank cannot repay all the depositors when they sell their assets at fire-sale prices (though they could repay if they were able to sell the assets at their true value.) Once depositors realize this, each depositor wants to run to the bank and get their money out when they can, and not be one of the depositors that doesn’t get their money back. When everyone does this, the bank fails. The depositors fear of bank failure leads to a self-fulfilling prophesy.

That was not what happened to SVB. It was not solvent – apparently, according to the press, its net worth, if assets were all marked to market, was negative. It could have held on had depositors kept their money in the bank, and that’s why SVB stock was still valued before last week. But the story of SVB is that they gambled and lost.

What is SVB worth now?

One indication that SVB was not the victim of a classic bank run is that no other bank wanted to buy SVB. The only real value from buying SVB might be if the bank had a good reputation which a buyer could still tap into.

The carry trade that SVB was running seemed profitable. However, if the carry trade really is valuable, there really is no need for another bank to buy SVB to profit from it. They could just do the carry trade themselves. The only value from buying SVB is if another bank thought it could keep SVB’s depositors because of the reputation SVB built up for being good to its customers. Apparently, at least for now, that good will was not valuable enough to induce another bank to buy SVB.

Did the Fed/Treasury/FDIC do the right thing?

There are two major policy changes undertaken by the regulators. The one with the most publicity is that depositors were repaid, even when their deposits were greater than the $250,000 limit that the FDIC had guaranteed. The second, which I describe below, is to turn long-term bonds held by banks into liquid assets.

Note that the owners of SVB were not bailed out. It appears the owners will lose their entire investment. Vanguard was a big owner of SVB. So, it isn’t just the rich guys that lose out. A lot of small investors have their wealth and pensions managed by Vanguard. On the other hand, the executives at SVB have made a lot of money the past few years, so they will lose their jobs, but they aren’t exactly hurting.

Will depositors ignore risks?

There is an argument that reimbursing all depositors creates “moral hazard”. That is, depositors should be monitoring the health of the bank in which they put their money. But if they can be sure that they will get their money back, even if the bank is careless, then depositors have little incentive to monitor.

Here, I think the small depositors should not shoulder blame. It is just too much to ask of a plumber or grade-school teacher to monitor the bank and assess the risk on the bank’s balance sheet.

Larger scale investors that had multi-million-dollar deposits bear more responsibility. I’m skeptical even then that there is enough sophistication in the finance departments of tech companies to recognize these risks, except in the largest ones.

In any case, the Fed/Treasury/FDIC at this point must trade off the problem of creating moral hazard by depositors with the risk of panic leading to “classic bank runs”, as I defined them.

Creating liquidity

The second major policy move was that the Fed agreed to supply short term loans to banks and take long-term bonds as collateral. The key provision is that the collateral value of these bonds is taken to be full face value. In other words, a bond that pays off $10,000 in 10 years is worth $10,000 of collateral, even if its current price on the market is much less. A bank can borrow $10,000 from the Fed, for up to one year, and post this bond as collateral.

That means that the long-term bond now provides $10,000 of cash to the bank with which it can pay off depositors.

Any bank that finds itself in the position of SVB or Signature bank will now be in a much stronger position.

I don’t think there are very many banks, though, that have balance sheets as risky as SVB’s and Signature’s. (Signature’s case is somewhat different than SVB’s but related.)

However, while most banks might stay solvent even if a lot of their funding disappears, the banks would still shrink in size. That would mean a reduction in the services these banks provide, especially to firms and households that need loans. While this would not constitute a financial panic, it would still be undesirable from a macroeconomic standpoint because productive investments would not get funded. It makes sense for the Fed to try to avoid such a situation.

It seems fair to let the banks that took on too much risk, like SVB and Signature, fail, but to provide liquidity to other banks. No policy can achieve a perfect outcome in the current situation, but I believe the policies that have been undertaken are good ones.

Parting thoughts

I realize this is not a definitive or detailed account of what happened. I’m just trying to fill in some of the blanks. You’ll see a lot of commentary of the form “we just need to reform the financial system by doing …”, or “regulators just need to …” There are many reasons why the market cannot, with complete efficiency, channel funds from investors to borrowers. There is probably an optimal way to organize and regulate the financial system, and I’m sure we haven’t hit on it yet. Even the optimal system, however, won’t be perfect. We will all learn from the current mess, and, I hope, make changes to the system that make it more nearly perfect.


This post written by Charles Engel.

24 thoughts on “Guest Contribution: “Some Thoughts on SVB”

  1. w

    I’m curious whether you think that any bank in the country can handle a run on over twenty percent of its deposits in less than 48 hours. We know that all banks – of any size – now hold some percentage of reserves that would make losses if sold at market.

    1. baffling

      this is a problem in todays financial world. banks and regulations were developed when you had to go to the bank to issue the withdrawal, and news spread by paper. now everybody has the “right” to transfer money in real time while getting spun by misinformation on twitter 24/7. not sure how the banks need to evolve to handle this situation, but the current rules do not allow for it to be done efficiently. you saw this in the reddit fueled GameStop and AMC episodes of a couple years ago. I would be curious, those VC folks who claim to only withdraw their money to protect it. did any of them short the stock in the process of hollering on twitter how the bank was insolvent? especially when they were the ones who created the illiquid condition with large withdrawals to begin with? I would like to have that investigated.

    2. King James return

      The issue with SVB was that 25% of its depositors accounted for 88% of the deposits. In other words, a small number of customers accounted for a HUUUGE amount of the deposits. So a few large customers accounted for the run. I find it hard to believe that any other banks are this exposed.

    3. Ivan

      The real question is not “handle” but avoid. A bank that is 90% insured consumer deposits will never get a run of 20% of its deposits because there are not that many depositors who will fear the loss of money – no matter what the internet says about that bank.

  2. Macroduck

    Pardon my off-topic. I’m about to read Professor Engel’s post,assoon as I get this off my chest:

    Two factors have interacted to foster the collapse of three banks in a single week – weak regulation and a sharp rise in interest rates. There are probably other factors on the list – drag queens and good environmental stewardship, for instance – but let’s not take on too much in a single comment.

    Let’s limit ourselves to just on factor – rates. For a number of years, interest rates were very low. Rates were low for long enough to induce big distortions. Housing prices reached the highest multiple of household income on record:

    https://fred.stlouisfed.org/graph/?g=11946

    Stock market capitalization reached a new high relative to GDP:

    https://fred.stlouisfed.org/series/DDDM01USA156NWDB

    We know from the recessions of the 21st century that these are dangerous conditions. We know from the 2014 taper tantrum and its little cousin in 2019 that reducing central bank liquidity after a period of zero rates and large central bank asset purchases is dangerous. We know from the extended period of below-target inflation that central banks have limited power over inflation in a low-rate environment.

    Raising rates is dangerous, and prolonged periods of low rates seem to create conditions which magnify the danger. Let’s remember, the Fed has downplayed capital investment as a channel for monetary policy and emphasized the wealth effect – and there is little question that is an accurate characterization of effects. Problem is, physical capital remains after a recession, ready to be put to use. Wealth evaporates as rates rise, and the evaporation in 2022 was the worst in over a century. Sudden swings in wealth add to financial and economic risk. Can we afford a policy which is heavy on financial effects, while doing little to encourage productive investment? I am not sure that’s fair, but if the Fed has been unable to kill job growth with draconian rate hike, how much credit do low rates deserve in the long expansion and the Covid recovery?

    Let’s also remember that banks necessarily hold Treasury debt as Tier-One capital. Regulators require banks to hold assets, for prudential reasons, which collapsed in value because one of those regulators went from a policy of zero rates and asset accumulation to rapid rate rise and shedding assets. Left hand, meet right hand.

    All of which is to say, hindsight is likely to encourage the Fed to maintain higher rates, relative to economic and financial conditions, in the future than in recent cycles. The risks inherent in maintaining rates near zero may not be all that well balanced by benefits.

    That is not to say monetary policy is the villain in this story. Regulator policy allowed the mortgage crisis to happen and the Fed used the tools it had in response. Larry Summers told Obama that a fiscal stimulus adequate to the task would look bad, and Obama believed him. Bank bailouts were OK, but household bailouts were anathema. The Fed was left to fill the gap.

    This time, a pandemic created a supply shock and Trump and Biden and both Congresses realized that household support was wise policy. Russia’s war caused another supply shock and profiteering by corporations piled on. Dodd-Frank was gutted. In that environment, the Fed overreacted to inflation.

    Even though monetary policy isn’t the only cause of inflation or bank failures, I don’t imagine zero rates for as long as it takes is going to be a popular prescription in the next couple of decades. Distortion and risk will be remembered too well.

    I wonder what tools can replace super-easy monetary policy. The lesson of the Great Recession was that intervention should be adequate to the shock. That lesson was put to use during Covid, with good results. Those who are treating inflation like the end of the world have short memories – the Great Recession had much worse economic consequences than Covid in large measure because we learned in the Great Recession that counter-cyclical policies need to be big and persistent.

    So, a future of higher rates with more reliance on financial regulation and fiscal policy? Does that seem likely to anyone?

    1. w

      Macroduck,
      “Those who are treating inflation like the end of the world have short memories”
      Yes, that much for certain. Inflation is not a smooth flowing stream. We’ve had a long period of unusual stability during which wages relative to productivity got stepped on. Now we have a catch-up time. It’s not a crisis and it will come to a natural end. There is no reason to break housing, to break commercial real estate, to put people out of work, drive stock markets down and to break banks to end it. That’s some kind of Fed hubris – it’s a hero delusion: “Look at us, we’re the Paul Volkers of today”.
      Of course all the finger pointing by government will be at the banker; they will protect their own. Yes, the bank made mistakes… but so has the Fed and the Fed’s are bigger.

  3. JohnH

    Michael Hudson puts the SVB meltdown into the context of the broader problem of low asset valuations in a time of high interest rates: https://michael-hudson.com/2023/03/why-the-banking-system-is-breaking-up/

    I have yet to see any reports on what banks’ net worth would look like if they had to mark to market all of their “hold to maturity” assets, not just bonds. Brokered CD rates are now over 5% and some money market accounts are offering 4.7%. Meanwhile most treasury bonds exceeding one year maturity yield less than 4%; mortgages written 2-3 years ago return less than 3% on face value. Sure, banks could have funded their mortgages with 10-year treasuries, but they could also have funded them with the more profitable “borrow short, lend long” approach.

    It will be interesting to see what the Fed’s balance sheet looks like in six months.

    1. pgl

      “The collapses of Silvergate and Silicon Valley Bank are like icebergs calving off from the Antarctic glacier. The financial analogy to the global warming causing this collapse is the rising temperature of interest rates, which spiked last Thursday and Friday to close at 4.60 percent for the U.S. Treasury’s two-year bonds. Bank depositors meanwhile were still being paid only 0.2 percent on their deposits.”

      Oh my – tight money and global warming?! No wonder little Jonny boy got excited. But wait Jonny boy the other day Jonny boy was praising high interest rates as they allegedly gave depositors 4% returns. Well may not but that Jonny boy has malleable opinions is nothing new.

  4. Baffling

    Nice summary, thank you. I have heard the comment that the management at svb were not nad hiys, but made foolish mistakes. Of that is indeed the case, then the government action has been appropriate in not penalizing the depositors or banking customers.

  5. Macroduck

    Other banks can do SVB’s carry trade without buying SVB. Yes. And in the medium term, that’s what matters. In the short term, what matters to clients is getting their banking done. The FDIC is going to cover their deposits, but at what bank?

    So a bunch of tech execs, many of whom were told where to bank, now need to establish banking relationships. Not just bank accounts, banking relationships. These guys were told to bank at SVB by the venture capital firms which funded them. The banking relationship we’re looking at here involves surveillance of the tech firm by the VC firm through SVB. That’s what SVB was good at, and SVB was the biggest. Now, it’s gone.

    Not that I feel sorry for the tech execs or the VC execs, but they’ll have their hands full for a while. Surveillance is a big part of this game. Being a bank isn’t enough. So I guess most of the account execs from SVB will find work pretty quickly and bring their client books with them.

    1. Moses Herzog

      Looks pretty great and wise about 3 and 1/2 year later, aye?? I just think she has differentiated (in the most complimentary and positive way) herself from the other governors. But then I liked Alan Blinder a lot also when he was there, and a strong case can be made Blinder was too “dovish” under Greenspan. I don’t think he was, but I’m willing to concede the case could be made.

      The words “sharp cookie” keep coming to mind when I think of Brainard. I guess the “woke” folks will have to sit around at a conference table 16 hours a day for a full week and see if they can contort the phrase “sharp cookie” into some deep personal violation or the metaphorical version of nonconsensual penetration. Until then I guess we’ve still got 7 days we can call Brainard a sharp cookie.

    2. pgl

      Brainard was ahead of the times there! BTW CoRev is BAAAACK blaming the regulators for spending part of a minutes on financing clean energy. YEA CoRev reads a dumb headline from Faux Business. Never mind the regulators talked about other things during this meeting. Never mind no one said holding stocks in clean energy tech stocks led to the collapse of SVB. But wait – here is a story that should make CoRev happy as the collapse of SVB might reduce funding for these stocks:

      https://www.msn.com/en-us/money/markets/silicon-valley-bank-collapse-could-slow-the-transition-to-clean-energy/ar-AA18AF5z

      CoRev is elated that transitioning away from his precious fossil fuels might be slowed a bit. MAGA!

    3. baffling

      at about this point in time, svb started to grow its assets quite a bit. interesting timeline. my understanding of svb management was they were smart and morally upright, but rather naive in the banking and finance world overall. they catered to a niche clientele, and did that well, but would have been overmatched with greater competition. their reach for yield seems to be an example of that in practice. my guess is the reach for yield was brought on by some “outside” expert opinions…which is why the bank could not successfully complete the maneuver.

  6. AS

    Thank you.
    Excellent summary.
    I wonder if the SVB executives wanted to believe that the rise in inflation was “transitory” and thus took too much risk by making ten-year treasury investments. I also wonder if any of the board members lived through the inflation of the 1970s to 1980s. Many of us who lived through that period have inflation and high interest rates singed into our memories and have been waiting for a return. We may not repeat the prior decade’s high inflation and interest rates, but we seem to have a bit of “rhyming” happening.

  7. pgl

    BLS speaks, you listen:

    https://www.bls.gov/news.release/cpi.nr0.htm

    ‘CONSUMER PRICE INDEX – FEBRUARY 2023

    The Consumer Price Index for All Urban Consumers (CPI-U) rose 0.4 percent in February on a seasonally
    adjusted basis, after increasing 0.5 percent in January, the U.S. Bureau of Labor Statistics reported
    today. Over the last 12 months, the all items index increased 6.0 percent before seasonal adjustment.

    The index for shelter was the largest contributor to the monthly all items increase, accounting for over
    70 percent of the increase, with the indexes for food, recreation, and household furnishings and
    operations also contributing. The food index increased 0.4 percent over the month with the food at home
    index rising 0.3 percent. The energy index decreased 0.6 percent over the month as the natural gas and
    fuel oil indexes both declined.

    The index for all items less food and energy rose 0.5 percent in February, after rising 0.4 percent in
    January. Categories which increased in February include shelter, recreation, household furnishings and
    operations, and airline fares. The index for used cars and trucks and the index for medical care were
    among those that decreased over the month.’

    OK. Now what would be the change in the index for all items less food, energy, and shelter?

  8. Moses Herzog

    I don’t even know Mr Engel. But my “spidey sense” tells me he won’t at all if I add this off-topic comment to his terrific and edifying post.

    I suppose if Professor Chinn put this link up in a post, folks like Kopits and Bruce Hall would label Menzie a Chinese spy. That’s ok. Some of us other white dorks have got Menzie’s back, and other Chinese Americans who contribute greatly to the progression of America, we got their back also. And I’ll tell you this, when I’m “out and about” in public, it’s not happening with no response on my watch. Yup, that means if I have to get physical to defend an innocent party, then that’s where it’s going. I owe too much great memories and feeling to my time in China (amongst a very few Asian friends I have had here in the states):
    https://www.reuters.com/world/us/hate-crimes-us-surged-116-2021-2023-03-13/

    1. Moses Herzog

      *he won’t mind at all.

      Brain not working this morning cuz I got a sweater on and it’s hot in this building. It’s like my brain malfunctions when I’m hot. I know that sounds lame, but it’s true.

  9. Steven Kopits

    This analysis is helpful, but not entirely correct.

    First, to suggest that SVB was investing in risky assets — US treasuries and MBSs — is not correct. After all, when was a US treasury considered a risky asset?
    There is no apparent repayment risk at present, but only interest rate risk.

    Now, the author above suggests that there was no classic run on the bank. This is structurally incorrect. Why? If the bank is locked into low interest rate assets, then it cannot adjust its deposit rates. So if the bank lent long at 3% and when deposit rates were 1%, then everything is fine. However, if 3 mo Treasuries go from 0% to 5% (4.72% today) in fourteen months (https://fred.stlouisfed.org/series/DGS3MO), then the bank will struggle to adjust. That’s what happened.

    Importantly, note that the treasurers of the depositing companies will find the 1% deposit rate uncompetitive with 3 mo treasuries at 4.72%, and so will pull their money out of the bank. This is not a bank run in the sense of a panic (at first), but in the sense of the bank being unable to offer competitive deposit rates, thereby in effect causing depositors to withdraw money in a ‘shadow’ run, which turns into a real run when the bank has to liquidate assets below historical cost.

    Is SVB unique in this regard? I wouldn’t think so.

    1. pgl

      You start by denying investing in Treasuries and MBS is not investing in “risky” assets.

      ‘First, to suggest that SVB was investing in risky assets — US treasuries and MBSs — is not correct. After all, when was a US treasury considered a risky asset? There is no apparent repayment risk at present, but only interest rate risk.’

      But the rest of your latest screed admits interest rate risk can be serious. Listen dude – there is a reason why no one want to read your arrogant babbling. Such contradictions is your only forte,

      Charles Engel is a very good economist. You are a moron who does not deserve to shine his shoes.

Comments are closed.