Introduction: The recent collapse of Silicon Valley Bank has sent shockwaves through the banking industry, with economists warning that up to 186 other banks are at risk of failure. A new study has found that aggressive interest rate hikes by the Federal Reserve to curb inflation have eroded the value of bank assets, such as government bonds and mortgage-backed securities. As a result, a run on these banks could pose a potential risk to even insured depositors, and the FDIC’s deposit insurance fund could start incurring losses.
Highlights of the study
The report highlights the negative impact of central banks tightening monetary policy on long-term assets, including government bonds and mortgages. The report explains how banks finance long-term assets with short-term liabilities, and as interest rates rise, the value of these assets can decline, potentially leading to bank failure. The report provides examples of past bank failures and the role of uninsured depositors in such failures.
The report also analyzes the recent interest rate hike by the Federal Reserve Bank, which resulted in a decline in the market value of long-duration assets. The report uses bank call report data to assess the financial stability of U.S. banks and finds that bank assets decline on average by 10%, with the bottom 5th percentile experiencing a decline of approximately 20%. The report also highlights the impact of declines in marked-to-market asset values on the solvency and run incentives of banks, especially those with high uninsured leverage.
The report provides a detailed analysis of the potential risks associated with central banks tightening monetary policy and its impact on long-term assets and bank stability. The report highlights the importance of monitoring banks’ funding structures and uninsured leverage, especially during periods of monetary tightening. The report’s findings suggest that uninsured depositors represent a significant source of funding for commercial banks, and as such, banks need to manage their funding structures to avoid potential runs. The report concludes that assessing the financial stability of banks during periods of monetary tightening is critical for preventing bank failures and maintaining financial stability.
The Fragility of the US Banking System
Banks’ hidden losses, also known as “marking to market,” refer to the process of valuing assets based on their current market price rather than their book value. This approach is important because it provides a more accurate representation of a bank’s financial health, especially during times of economic uncertainty.
The impact of interest rate increases on banks’ asset values can be significant, and this is where marking to market becomes crucial. According to a study by Jiang et al. (2020), a substantial portion of bank portfolios, specifically loans held to maturity, are not marked to market. Therefore, the researchers rely on exchange-traded funds (ETFs) across various asset classes to conduct their analysis.
Real estate loans account for approximately 42% of the assets of the average bank, with securities linked to real estate (such as mortgage-backed securities (MBS), commercial mortgage-backed securities (CMBS), treasuries, and other asset-backed securities (ABS)) making up approximately 24% of the average bank’s assets. As these assets represent more than half of the total assets for a typical bank, the researchers concentrate on marking them to market, which may result in underestimating the effect on the remaining portion of the bank balance sheet, which they leave unchanged.
The researchers mark bank assets to market in three steps. First, they obtain the asset maturity and repricing data for all FDIC-insured banks in their regulatory filings (Call Report Form 031 and 041) in 2022:Q1. Second, they use traded indexes in real estate and treasuries to impute the market value of real estate loans held on bank balance sheets. Finally, they define the mark-to-market asset value in 2023:Q1 as total assets in 2022:Q1 minus the mark-to-market value loss defined above.
The researchers found that marking the value of real estate loans, government bonds, and other securities results in significant declines in bank assets. The median value of banks’ unrealized losses is around 9% after marking to market. The 5% of banks with worst unrealized losses experience asset declines of about 20%. These losses amount to a stunning 96% of the pre-tightening aggregate bank capitalization.
The unacknowledged losses differ slightly across the size distribution, with the smallest for GSIBs at 4.6% and the largest for large non-GSIB banks at 10%. There are also likely substantial differences in the uses of interest rate hedges across the size distribution of banks (esp. GSIBs), which the researchers were unable to account for due to data limitations.
In total, the U.S. banking system’s market value of assets is $2 trillion lower than suggested by their book value of assets as of 2023:Q1. Interestingly, the recently failed SVB does not stand out as much in the distribution of marked to market losses. About 11 percent of banks suffered worse marked to market losses on their portfolio, indicating that if SVB failed because of losses alone, more than 500 other banks should also have failed.
In conclusion, marking to market is an important approach to accurately assess a bank’s financial health, particularly during times of economic uncertainty. The study by Jiang et al. (2020) highlights the significant impact of interest rate increases on banks’ asset values and the need to account for unacknowledged losses to provide a more accurate picture of the U.S. banking system’s financial health.
The potential risk of a run on banks with uninsured depositors
The role of uninsured leverage in banking can be significant, particularly when it comes to bank runs. Bank runs occur when depositors lose confidence in a bank and withdraw their funds en masse, leading to the bank’s insolvency. Uninsured leverage refers to the proportion of a bank’s funding that comes from sources other than insured deposits or equity, such as uninsured deposits and other debt funding. When a bank relies heavily on uninsured leverage, it becomes more vulnerable to bank runs because these funding sources are more prone to withdrawal during times of stress.
The reasons behind Silicon Valley Bank’s collapse
In the case of Silicon Valley Bank (SVB), the bank was not particularly thinly capitalized compared to other banks, but it stood out in its reliance on uninsured leverage. Prior to monetary tightening, the average bank funded 10% of their assets with equity, 63% with insured deposits, and 23% with uninsured debt. SVB, however, funded over 78% of its assets with uninsured deposits, putting it in the 1st percentile of distribution in insured leverage. This made it much more run-prone than other banks.
The impact of this reliance on uninsured leverage became clear when SVB failed during the financial crisis. Even if only half of uninsured depositors panicked and withdrew their funds, this led to a withdrawal of one quarter of the total marked to market value of the bank. Any fire sale discounts resulting from these withdrawals could impose substantial losses on the remaining creditors, increasing their incentives to run. This suggests that uninsured deposits played a critical role in the failure of SVB.
Overall, the case of SVB highlights the importance of considering a bank’s funding structure, particularly its reliance on uninsured leverage, when assessing its vulnerability to bank runs. Banks with high levels of uninsured leverage may be more prone to runs and therefore more vulnerable to failure during times of stress.
The impact of the bank’s disproportionate share of uninsured funding
In the case of a bank with uninsured deposits, the possibility of a bank run becomes more likely compared to a bank with insured deposits. In this simple numerical example, we assume a bank with $100BN in assets, of which $90BN is in risk-free perpetuities, and $10BN is in cash. The bank has $80BN in uninsured deposits and $10BN in long-term debt at a fixed rate of 3%. The bank’s equity is $10BN, and its deposits cost 3% per year.
Suppose the Federal Reserve unexpectedly increases the risk-free rate by 100 basis points due to inflation. The value of the bank’s long-term assets will decline by 25%, to $67.5BN, while the value of its total assets will be $77.5BN. Despite the significant swing in the value of the assets, it is not immediately clear whether the bank is insolvent or whether depositors should run. This depends critically on the composition of depositors, i.e., how many are insured versus uninsured, and on the interest rates paid on deposits.
Run Incentives and Uninsured Leverage: Simple Example
We can generate a range of scenarios by considering different behavior by uninsured depositors. For instance, suppose that all depositors are insured, and the bank’s deposit costs remain at 3%. In this case, the value of equity will fall to $7.5BN, and the value of insured deposits will fall by 6.25%, or $5BN. Therefore, the bank remains solvent, and insured depositors will not experience a loss.
However, suppose that the bank has $40BN in uninsured deposits, and the uninsured depositors have a run probability of 50%. This means that if half of the uninsured depositors decide to withdraw their funds, the bank’s uninsured deposits will decline by $20BN. Since the bank has only $10BN in equity, its insolvency is triggered, and all depositors, including insured depositors, will experience a loss. In this scenario, uninsured depositors have an incentive to run due to the possibility of a loss of their funds.
Therefore, this example illustrates how the presence of uninsured deposits can make a bank more vulnerable to runs. Uninsured depositors have a greater incentive to withdraw their funds, which can trigger a bank’s insolvency and cause losses for all depositors. In contrast, insured depositors are protected by the FDIC and are less likely to withdraw their funds in a bank run scenario.
The baseline scenario described in this note considers the case where all depositors in a bank are “sleepy,” meaning they do not require a change in deposit rates offered to them, even if higher rates are being offered elsewhere. This scenario could be applicable to a bank that is primarily financed with insured deposits, as insured depositors may feel secure in the protection offered by deposit insurance and not be motivated to seek higher returns elsewhere.
In this scenario, the cost of deposits for the bank includes the deposit rate and acquisition costs, and the deposits are cheaper than the risk-free rate due to some special value that depositors attach to the bank’s debt. This is consistent with previous research (Jiang et al., 2020) that suggests deposit franchise allows banks to pay low deposit rates that are insensitive to market interest rates (Dreschler et al., 2017; Egan et al., 2017).
The note then considers several scenarios in which depositors may be motivated to withdraw their funds from the bank. In scenario 1, 50% of depositors are uninsured and half of them run. The bank would need to pay uninsured depositors $20BN, but the remaining marked-to-market value of bank assets would still be $57.5BN, which is more than the remaining face value of insured deposits ($40BN). Therefore, the bank could theoretically continue its operations on a scaled-down basis.
In scenario 2, all uninsured depositors run, and the bank needs to pay them $40BN. The remaining marked-to-market value of bank assets is $37.5BN, which is less than the face value of remaining insured deposits ($40BN). In this case, the FDIC would have to close the bank, and even insured depositors would be impaired.
Scenario 3 considers a situation where all depositors run or leave the bank for higher-yielding alternatives. In this case, the bank would be insolvent, as the marked-to-market value of its assets is less than the face value of deposits. The bank could sell its assets, but even with the proceeds from liquidation ($67.5BN) and its cash buffer ($10BN), it would not be able to cover the face value of deposits ($80BN). If all depositors are uninsured, they would have an incentive to run on the bank since the marked-to-market value of bank assets is less than the face value of uninsured deposits.
In summary, the baseline scenario assumes that depositors are “sleepy” and do not require higher deposit rates or consider withdrawing their funds from the bank. However, the note highlights that there are several scenarios in which depositors may become motivated to withdraw their funds, and the bank’s ability to handle these scenarios depends on the proportion of uninsured depositors and the extent of their withdrawals.
Banks’ Hidden Losses: “Marking to Market
Marked to market losses refer to the practice of revaluing assets based on their current market value. This means that the value of the assets on a bank’s balance sheet can fluctuate with changes in market conditions. If the value of a bank’s assets declines, it can result in marked to market losses, which can impact the bank’s solvency and expose it to run risk.
Solvency refers to a bank’s ability to meet its financial obligations as they come due. If a bank is insolvent, it means that its liabilities exceed its assets, and it may not be able to meet its financial obligations. Marked to market losses can impact a bank’s solvency if the decline in asset values is large enough to render the bank unable to cover its obligations.
Run risk refers to the risk of a bank experiencing a sudden and significant outflow of deposits, which can result in liquidity problems and potential insolvency. Marked to market losses can expose a bank to run risk if depositors lose confidence in the bank’s ability to meet its obligations and decide to withdraw their deposits.
In the context of U.S. banks, it is difficult to evaluate the market value of deposit liabilities. Deposit liabilities are on demand and can be evaluated at their face value at prevailing rates, but there may be a spread between deposit rates and fed funds rates due to banks’ market power, which can allow banks to earn rents. This can make it challenging to determine the true value of deposit liabilities and the impact of marked to market losses on a bank’s solvency and run risk.
However, a benchmarking exercise can be used to evaluate whether the assets of U.S. banks are sufficient to cover uninsured deposits. The exercise considers the run incentives of uninsured depositors from the perspective of assets after marking assets to market. The results of the exercise show that virtually all banks have enough assets to cover their uninsured deposit obligations, except for a few, such as SVB, whose marked-to-market assets are barely enough to cover its uninsured deposits. This means that even a small fire sale discount could result in uninsured depositors losing money in a run, making a run rational.
This passage discusses the fragility of banks to uninsured depositor runs and studies scenarios related to such events. The authors consider two cases: in the first case, all uninsured depositors run, while in the second case, half of them run. The study compares the pre and post-FED monetary tightening scenarios. Before the FED interest rate increases, U.S. banks were solvent under both scenarios, and uninsured depositors had no incentives to run. However, after the tightening, banks are much more fragile to uninsured depositors runs, and some banks may become insolvent, which would impair insured depositors. The authors also plot the distribution of insured deposit coverage ratio and the 10 largest banks at risk of a run. They conclude that some banks are certainly at a potential risk of a run, and there may be no funds left for the remaining uninsured depositors if the FDIC shut these banks following a run.
Extreme Insolvency: No Deposit Franchise
Extreme Insolvency: No Deposit Franchise is a scenario where the solvency of banks is evaluated based on whether the market value of their assets is sufficient to cover all non-equity liabilities, assuming that there is no value to banks’ deposit franchise. This means that if all depositors and debtholders withdrew their funding today, banks would need to repay their debts without any support from their deposit franchise. In this scenario, the assumption is that assets can be sold at their current market value without any additional discount due to liquidation.
This extreme scenario is useful in understanding the de facto capitalization of the U.S. banking sector. However, it is important to note that insured depositors have no incentives to withdraw funds as a function of default risk, so this scenario is not likely to occur in practice.
However, after the recent decrease in the value of bank assets, 2,315 banks accounting for $11 trillion of aggregate assets have negative capitalization. This implies that they would be unable to repay their debts if all depositors and debtholders withdrew their funding today.
The study also highlights the potential risk of impairment to even insured depositors if a large share of uninsured depositors decides to withdraw their funds. The study estimates that almost 190 banks are at potential risk of impairment to even insured depositors, with potentially $300 billion of insured deposits at risk. If uninsured deposit withdrawals cause even small fire sales, substantially more banks are at risk.
Overall, the calculations suggest that recent declines in bank asset values significantly increased the fragility of the US banking system to uninsured depositors runs. The study recommends several medium-run regulatory responses, including expanding more complex banking regulation on how banks account for mark-to-market losses and imposing stricter capital requirements on banks.
In conclusion, a new study has found that close to 190 banks could face the same fate as Silicon Valley Bank’s collapse, due to the Federal Reserve’s aggressive interest rate hikes that have eroded the value of bank assets. The U.S. banking system’s market value of assets is $2 trillion lower than suggested by their book value of assets. If only half of uninsured depositors decide to withdraw their funds, almost 190 banks are at potential risk of impairment to even insured depositors, with potentially $300 billion of insured deposits at risk. The study suggests that recent declines in bank asset values have significantly increased the fragility of the US banking system to uninsured depositor runs.
The study proposes several medium-run regulatory responses to an uninsured deposit crisis. One response is to expand even more complex banking regulation on how banks account for mark to market losses, but this may not address the core issue at hand consistently. Alternatively, banks could face stricter capital requirements, which would bring their capital ratios closer to less regulated lenders. However, there have been debates on the sufficiency of progress made on bank capital requirements after the 2007 financial crisis.