Silicon Valley Bank: Interest Rate and Liquidity Risk within Asset Liability Management

Politicians and pundits have been casting blame and pointing fingers at a variety of sources that contributed to the collapse of Silicon Valley Bank (SVB), which also resulted in contagion at other financial institutions with similar balance sheet structures. Rather than point blame, it seems like a better idea to examine the way that SVB monitored and managed interest rate risk (IRR) and liquidity risk utilizing their internal asset liability management (ALM) processes. This article will evaluate SVB’s 2022 10-K and other sources to identify some of the key interest rate and liquidity risk drivers that led to their failure.

Banks operate under specific regulations requiring senior management to assess the interest rate risks embedded in their institution’s balance sheet. Most banks charter an Asset Liability Management Committee (ALCO) to monitor and manage these risks. Larger and more complex banking organizations often form a tactical ALCO which might meet weekly and a strategic ALCO which might meet monthly. An ALCO at a smaller bank might only meet quarterly depending on the nature and degree of IRR at their institution.

The explanation of why SVB collapsed is pretty straight forward. As interest rates rose, they incurred substantial unrealized losses on their securities portfolio, which was primarily invested in relatively safe (from a credit risk perspective) U.S. Treasury securities and Agency-guaranteed residential mortgage-backed securities (RMBS). Accounting standards did not require the bank to recognize these losses in current earnings, but they did reduce the bank’s overall level of tangible equity capital. Banks are required to maintain certain levels of capital to protect against losses to depositors, but at some point, many of the bank’s customers realized the bank’s capital cushion had been substantially eroded and withdrew their deposits. Frenzied social media posts encouraging depositors to withdraw their funds appears to have exacerbated the bank’s difficulties in retaining customer balances. Ultimately this run on SVB’s deposits destroyed all confidence in the bank’s ability to manage its funding needs and the state of California closed the institution.

There were a lot of moving parts that led to SVB’s failure, many of which were readily apparent by examining the bank’s public disclosures. Let’s look at SVB’s 2022 10-K to see if it’s possible to identify disclosures that should have indicated problems related to liquidity and interest rate risk.

Let’s start by evaluating SVB’s market risk management framework described in the 10-K. Market risk also incorporates exposures to changes in equity values, foreign exchange rates and commodity prices, but we will focus on their interest rate risk exposures.

“Interest Rate Risk Management – Market risk is defined as the risk of adverse fluctuations in the market value of financial instruments due to changes in market interest rates. Interest rate risk is our primary market risk and can result from timing and volume differences in the repricing of our rate-sensitive assets and liabilities, widening or tightening of credit spreads, changes in the general level of market and benchmark interest rates and the shape of the yield curve. Additionally, changes in interest rates can influence the rate of principal prepayments on mortgage securities, which affects the rate of amortization of purchase premiums and discounts. Other market risks include foreign currency exchange rate risk and equity price risk (including the effect of competition on product pricing). These risks and related impacts are important market considerations but are inherently difficult to assess through simulation results. Consequently, simulations used to analyze the sensitivity of net interest income (NII) to changes in interest rates will differ from actual results due to differences in the timing and frequency of rate resets, the magnitude of changes in market rates, the impact of competition, fluctuating business conditions and the impact of strategies taken by management to mitigate these risks.”

The main tools that banks use to measure and manage interest rate risk are net interest income (NII) forecasting and economic value of equity (EVE) modeling. Let’s look at SVB’s disclosures on interest rate risk measurement:

“Model Simulation and Sensitivity Analysis – Both EVE and NII measures rely upon the use of models to simulate cash flow behavior for loans and deposits. These models were developed internally and are based on historical balance and rate observations. As part of our ongoing governance structure, each of these models and assumptions are periodically reviewed and recalibrated as needed to ensure that they are representative of our understanding of existing behaviors. A specific application of our simulation model involves measurement of the impact of changes in market interest rates on the EVE. EVE is defined as the market value of assets, less the market value of liabilities. Another application of the simulation model measures the impact of changes in market interest rates on NII assuming a static balance sheet, in both size and composition, as of the period-end reporting date. In the NII simulation, the level of market interest rates and the size and composition of the balance sheet are held constant over the simulation horizon. Simulated cash flows during the scenario horizon are assumed to be replaced as they occur, which maintains the balance sheet at its current size and composition. Investment portfolio cash flow is based on a combination of third-party prepayment models and internally managed prepayment vectors depending on security type. Based on a historical deposit study of our clients, we make certain deposit balance decay rate assumptions on demand deposits and interest-bearing deposits, which are replenished to hold the level and mix of funding liabilities constant. These assumptions may change in future periods based on changes in client behavior and at management’s discretion.”

Recall the concept of NMD. If the customer has the right to withdraw these funds, are they considered to be a stable funding source? Terms like sticky or core deposits are often used to describe these funds. Banks conduct statistical studies to estimate how stable/sticky/core these funds over long-term rate cycles. These studies indicated that SVB deposits were sticky in the long term, but did not address depositor behaviors in the short term when faced with a rapid deterioration in the bank’s financial condition.

The interest-bearing deposits may have a contractual maturity but when a bank is perceived as failing then customers will still demand their money and are willing to pay any contractual breakage penalty for the return of their principal. In addition, a disclosure in the 10-K showed that a substantial portion of time deposit balances were uninsured, making them even less sticky:

This analysis has attempted to provide a measured view of the unique factors at SVB that unfolded over the past few months as opposed to what is being portrayed by politicians and the media. The 2022 10-K contains a great deal more information than has been evaluated here, but the point was to highlight the key risk drivers associated with interest rate and liquidity risk. It seems to be pretty straight forward – SVB did not fully understand the impact of unrealized losses in their investment portfolio, and uninsured depositors lost faith in the bank and pulled their money out. The Fed plans to release an assessment report by May 2023 that will most likely address other issues that are not covered here such as capital adequacy, liquidity stress testing, and the scope of regulations for mid-sized banks. While credit risk does not seem to be a major factor at this point, many of SVB’s customers were start-up tech ventures with unique credit needs, and without a bank like SVB to underwrite their borrowing needs it is unclear how they will continue to fund their operations and service existing debt, potentially leading to higher rates of delinquencies and defaults. It would be a fairly certain bet that the Fed’s report will be a lot longer than these nine pages!

¹ AFS stands for Available-for-Sale and HTM stands for Held-to-Maturity. These are accounting terms and banks have to classify individual securities as AFS or HTM. The following is from the SEC’s website: “HTM securities, which management has the intent and ability to hold until maturity, are carried at amortized cost. AFS securities are carried at fair value and unrealized gains and losses are reported as net increases or decreases to accumulated other comprehensive income (AOCI).”

The Federal Reserve passed reforms to bank capital and liquidity rules in October 2019 that allowed banks with assets under $250 billion to opt out of including Accumulated Other Comprehensive Income (AOCI) from regulatory capital – an option that SVB exercised for their financial reporting.

² https://www.spglobal.com/marketintelligence/en/news-insights/latest-news-headlines/svb-signature-racked-up-some-high-rates-of-uninsured-deposits-74747639

About the Author: Ken Kapner