4 questions to Teodor Dyakov on the failure of the Silicon Valley Bank

A first - almost naïve - question: why do banks fail?

There are many reasons why banks fail, but I will focus on the essential role of banks in providing liquidity and how that is related to the failure of the Silicon Valley Bank (SVB). A common saying in the business world is that banks “borrow short and lend long”. In the case of SVB, the bank borrowed from depositors who can redeem their deposits at any time, and hence, the bank borrowed “short”. The bank used deposits to fund a portfolio of loans and bonds whose value can be redeemed at a much longer term, and hence, it lent “long”.

This exposes banks to a liquidity risk – when depositors want their money back, banks have to sell their assets at a discount, even if the long-term performance would be beneficial to the bank and its depositors. If depositors worry about the solvency of the bank, they may “run” to the bank to get their money back, fearing that the withdrawals of other investors may cause the bank to sell assets at a discount, eventually generating losses and triggering an insolvency. Thus, the essential role of banks in transforming illiquid assets into liquid liabilities may rationally lead to bank runs. This insight earned Douglas Diamond and Philip Dybvig the 2022 Nobel Prize in Economic Sciences and is the main reason behind deposit insurance schemes around the world.

Could you elaborate on why raising interest rates and risk-management are the two main reasons behind this failure?

SVB promised investors to pay a slightly higher interest rate on deposits, relative to other banks. It managed to attract deposits from many venture capital backed firms in the Silicon Valley, historically its main area of focus as the name of the bank suggests. And how would the bank make profit? It would (mainly) invest the deposits in long-term maturity government bonds and thus capture the spread between the bond yield and the rate it pays on deposits. This business model has very little exposure to credit risk – after all, the bank invested in some of the safest assets in the world.

However, the investment in long-term bonds exposed the bank to a significant interest rate risk. An increase in interest rates by the FED lowers the value of bonds. Suppose a bond promises to pay 1000$ in two years and currently trades at 990$. When interest rates go up, investors would now be willing to pay less for the same bond (say, 980$), because they have a new, more profitable alternative: banks offer higher yield to investors. This inverse relationship between rates and prices becomes more pronounced the longer the maturity of the bond. In finance, we call this “duration” risk – long term bonds have a higher duration, and hence a stronger exposure to interest rate risk. Thus, when interest rates rose, the value of long-term bonds was particularly affected and the assets of SVB experienced a very strong decline in value.

In addition to that, banks are allowed to report the “hold to maturity” value of their long-term assets rather than the market value. If the bond promises to pay 1000$ in 10 years and the bank intends on holding the bond for 10 years, it can use the 1000$ value in its books. But when interest rates rise and the bank is forced to sell the bond on the market, it might only be able to sell it for, say, 800$ if it needs to stop a run. Thus, once depositors realized that raising interest rates had such a large negative impact on the value of the assets of SVB, the bank was already subject to the classical “bank run” described in the model of Diamond and Dybvig.

In a way, “the worm was in a fruit” for a long time, right?

Of course, the question we are all asking is: why didn’t SVB manage its interest rate risk? For example, the bank could have entered into SWAP agreements (an exchange of variable for fixed payments), essentially eliminating interest rate risk. And why did it have such high reliance on large, uninsured deposits from venture-capitalist backed firms? Deposits up to 250,000$ in the US are insured, but investors holding more than that amount are very likely to initiate a bank-run, should they fear for the solvency of the bank.

But even more importantly: how did the regulators miss this? As many observers noted, there were several red flags in the build-up to SVB’s failure. Nearly 90% of SVB’s deposits were uninsured, making the bank particularly prone to a classical bank run. In addition, the mismatch in duration between the assets and liabilities of SVB should have been apparent.

What is the role of education? Can we train the next generation of professionals who can avoid such failures in the future?

We know that banks borrow short and lend long and we know that bank runs can arise rationally, even if every market participant is competent and acting in the best interest of society. We cannot avoid the risk of bank runs outlined by Diamond and Dybvig: it is inherent in the nature of banking. However, we can use the case of SVB’s failure to emphasize the importance of managing interest rate risk. Perhaps we as a society got too comfortable with low interest rates. We must, however, stress that banks should limit the duration mismatch between their assets and liabilities, so that depositors and society are not adversely affected by rising interest rates. In addition, we need to have an important conversation about the role of regulation in modern financial markets – behind thousands of pages of rules, regulators somehow managed to miss a simple duration mismatch and a run-prone investor base. EDHEC is very well equipped to train the next generation of business professionals, and we need to show our students the root cause of debacles, such as the one we are observing right now with SVB and that finance has the tools to prevent them.