Silicon Valley Bank - A Quick Demise

Silicon Valley Bank went into receivership on Friday morning. As it was the 16th largest bank in the US, what actually happened and what does it mean for the banking sector?

This post was originally going to cover the Senate Committee on Banking, Housing and Urban Affairs with Federal Reserve Chairman Jerome Powell regarding some of the economic misrepresentations that occurred during the meeting. However, on Friday morning, Silicon Valley Bank (SVB) failed, potentially sending shock-waves through the US economy. We will cover this story today, while postponing the hearing till next time.

SVB Receivership

What happened

On March 10, 2023, SVB entered Federal Deposit Insurance Corporation (“FDIC”) receivership, as it became financially insolvent. FDIC receivership is a process whereby the FDIC manages a bank’s assets and deposits with the aim to return money to depositors. This can be done via transferring the bank to another institution (most often another bank) or slowly unwinding the bank.

How did this happen

SVB was a commercial bank. Commercial banks are in the business of taking money from individuals and corporations (i.e. deposits) and using these funds to provide lending to other individuals and corporations. Commercial banks make money on the interest rate spread between the lending rate and the deposit rate (most deposit accounts at banks, until recently, paid no or negligible interest). 

SVB was a bank that specialized in serving early ‘tech’ startups. During the Covid-19 pandemic, SVB received a large amount of deposits (approx. $180bln) from the startup companies. As lending opportunities for this capital dried up, SVB had nowhere to place the money. The bank decided to invest the money in certain safe assets – in their particular case, it was Mortgage Backed Securities (MBS). They intended to hold these assets till maturity (Held-to-Maturity, HTM). This means that they were never planning to sell these assets. They would receive the stated coupon amount throughout the duration of this asset, and the principal when it matured. These were long-dated securities, many years into the future. The yield (i.e. the rate of return) on these securities was very low, around 1%-2% because they were purchased when interest rates were low. 

Over the last few years, interest rates have significantly increased and it is now easy to find savings accounts in banks paying well above 3%-4%. This means that a security that yields 1% is very unattractive for investors, as it pays much less than a simple savings account. This results in the value of that security to fall. Therefore, on paper, the value of SVB’s investments was lower than what they paid for. However, by itself this is not a problem. SVB was still profiting from these assets because it was receiving a rate of return of 1%-2% on this investment, while paying depositors (i.e. whose money they invested with) less than that rate. Nevertheless, certain investors in SVB noticed that since the value of assets held by SVB have fallen, SVB might not be able to cover the deposits it has. Typically, this would also not be a problem because of FDIC insurance (an insurance policy into which all banks pay in), which insures every bank account up to $250,000 (if your money is in any FDIC insured bank, and it is below $250,000, you will not lose a dollar regardless of what happens to the bank). Unfortunately, SVB had a lot of corporate accounts that carried a lot more than $250,000. Over 90% of the total deposited amount at SVB did not officially fall under FDIC insurance. After worries emerged that SVB currently does not have enough funds to potentially pay back depositors, a bank run occurred (i.e. everyone went to the bank to ask for their money back). Overnight, around $40bln of deposits was returned to depositors. Since no bank carries such a large amount of cash on their account in proportion to their deposits (the $40bln equated to 25% of all SVB deposits, while the current reserve requirement in the US is 0%), SVB had to sell some of those MBS assets at a loss in order to give depositors their money back. This loss, which would not have mattered if a bank run did not occur, now spread panic to all other customers of SVB who demanded their money. SVB did not have this money, even if it sold all its assets. This all ended with FDIC receivership.

What’s next

Although the FDIC only guarantees deposits up to $250,000, it implicitly has started to guarantee all depositor amounts. The FDIC will attempt to resolve this issue in a way to get depositors all their money back. This could be done via transfer of SVB to a bigger bank that will have enough capital to pay depositors. With such a transfer to a bigger bank, these depositors will also not want to withdraw money, reversing the bank run. However, as the FDIC receivership process may take some time, there may be payroll and liquidity issues for corporate clients of the bank. The FDIC intends to limit the disruptions by offering access to corporate clients to all of their money on Monday, March 13. 

Should we be worried

Based on our understanding of the situation, the answer is no from a macroeconomic perspective. If you are an individual or corporation that has deposits above $250,000, the risk of a bank run impacting you is always present and should be managed going forward. There appears to be no danger to the rest of the financial system, as SVB was not a big bank and it did not hold any securities that could impact other banks. The main issue for SVB was the bank run. Bank runs are an inherent problem for the financial system because of what is called maturity mismatch between deposits and lending. Depositors can access their money at any time, while the money that banks lend out is typically locked for a period of time. Thus, if all depositors would want their money back instantly, no bank on its own would be able to pay them back. With that in mind, the Federal Reserve has mechanisms in place for the largest banks (so called “too big to fail”) that allow them to borrow money from the Federal Reserve against their assets in case there are big deposit withdrawals. Therefore, there shouldn’t be any financial contagion to the banking system due to these guarantees. The only serious risks will be borne by the corporate clients of SVB as they may have operational issues due to them not being able to access their money in the near term to pay vendors, staff, and other liabilities. [Update: on Sunday evening, the FDIC announced that all deposits will be backstopped, allowing businesses to access their money on Monday.]

Overall, SVB made three critical mistakes throughout this process:It invested a very large amount of the deposits at one time into one type of security. This exposed them to a large amount of timing and interest rate risk (i.e. interest rates changing). They were also not diversified in their investment. Given the above decision, it did not purchase interest rate swaps (i.e. insurance against interest rates rising) even though interest rates were at all time lows at the time of the acquisition. Given both of the above decisions, once interest rates started rising, SVB did not cut its losses by selling their assets. 

All three decisions, combined with poor communication from SVB’s management to the public, resulted in a bank run. 
Why was this missed by the regulators/market: This is a question that does not yet have a clear answer. The reason this wasn’t caught by regulators basically boils down to the fact that SVB most likely did not violate any regulatory rules. As explained above, the main issue was the concern by depositors that SVB would not be able to cover their depository obligations with their assets. This is then a question on how we mark the value of assets of a bank. Let’s simplify the SVB example to illustrate this issue.

Suppose a depositor puts $98 in the bank account at Bank A. Bank A can now use this money to either lend or invest. Suppose Bank A is very ‘safe’ and chooses to purchase a risk free asset that promises to give you $100 in a year. How much would you pay for this asset? That depends on the risk free interest rate. Let’s assume that it is 2%.  This implies that the value of this asset that gives you $100 in one year, is approximately $98, because the $2 difference is equivalent to the 2% rate. Bank A purchases this asset for $98 and records on their balance sheet that they have a $98 asset, while their liabilities are also $98 (the deposit they must give back at some point). Two weeks later, unexpectedly, the Federal Reserve raises rates to 4%. Using the logic above, this asset would now be valued at around $96. If Bank A were to update their balance sheet (this is called marking-to-market or MTM), it would now say that they have a $96 asset and a $98 liability (the deposit). This looks like they’re not able to cover the liability and initiates the bank run by the depositors. However, they will still get $100 at the end of the year, and will comfortably pay back the depositor then, even though they are unable to do so now. 

So how does this matter? Held to maturity (HTM) assets, which are the assets SVB was mainly holding, do not have to be marked-to-market – the value on the balance sheet does not need to be changed and can be kept at original cost. This ‘hid’ the insolvency issue (i.e. that assets were less than liabilities at this point in time). Without the bank run, this wouldn’t be a problem. 

The reason why not all assets are marked-to-market is because it would create a lot of volatility on banks’ balance sheets, even though the bank actually intends to hold these assets to maturity, meaning this volatility is irrelevant in the long-run. Furthermore, for large banks, the Federal Reserve ‘stress-tests’ these HTM assets also. For small banks, these tests are not required. This does create a possibility for small banks to manipulate their finances. Marking-to-market can show losses, as was the case with SVB. If a bank has losses, they must set aside capital (typically equity or retained earnings) in proportion to the losses, in order to give it an extra cushion to pay back depositors. This additional capital costs the bank, as it cannot be invested and, if additional capital is needed, raising it can be very costly. Therefore, by not having to adjust for mark-to-market, you can have more capital invested, even if the return on the investment was itself low. 

The issue, however, is that if this problem was caught earlier, the same situation would transpire – people would be worried and a bank run could occur. Thus, some suggested that the best approach in hindsight would have been to pass this bank to another, bigger bank several weeks ago. 

Conclusion

Tomorrow (Monday, March 13) could be a big day for the banking sector. Economists have studied the issue of bank runs quite extensively. Coincidentally, the 2022 Nobel Prize in Economics was given for the Diamond-Dybvig model of bank runs. This was a foundational model that explained the issue of maturity mismatch of deposits and lending, why banks charge a spread between lending and deposits to be compensated for this risk, and how individual assumptions and beliefs can precipitate a bank run. This bank run fear may have been further exacerbated, as explained in this Twitter thread, by social media where information spreads much faster (this could be an interesting avenue for further research). We will quickly see whether the SVB bank panic will spread to other banks tomorrow. Although the financial system is safe due to the actions of the Federal Reserve, it might not stop the public from being worried.

Post Scriptum

At the time of writing this article, there is information that the FDIC is accepting bids for SVB, with the intent to find a buyer by Sunday. Furthermore, the FDIC announced that all depositors will be guaranteed their money.