Implications of the Silicon Valley Bank collapse on the management of Held to Maturity (HTM) assets and overall bank liquidity risk management

The demise of Silicon Valley Bank (SVB) in March 2023 is a game changer for financial institutions in terms of the future of liquidity risk management. A major contributor to SVB’s collapse was its management of Held to Maturity (HTM) assets that were booked at amortised cost, rather than market value. As a result, there is increased regulatory scrutiny of banks with regard to HTM assets, with implications in terms of governance, controls, ongoing monitoring, asset pool composition/diversification, liquidity risk stress testing and behavioural modelling. A key observation will be the need to diversify asset composition for high quality liquid asset (HQLA) requirements that meet stricter stress testing standards.

CONTEXT

The collapse of Silicon Valley Bank (SVB) in March 2023 was a significant event in the financial services industry. It has significant implications for the future of liquidity risk management and global accounting standards. SVB was a major lender to technology startups and venture capital firms, and its failure sent shockwaves through the industry. The key factors that contributed to SVB’s collapse are:

• Excessive concentration in Held to Maturity (HTM) assets: SVB had a significant concentration of its asset portfolio in Held to Maturity (HTM) assets. HTM assets are generally valued at amortised cost rather than market value, as the intention is to hold the assets till maturity so that any short-term dislocations between the book and market values dissipate. Assets that have a market value treatment are classified as Available for Sale (AFS). SVB reclassified a big portion of its assets from AFS to HTM to avoid recognising unrealised losses for multiple quarters. There was a lack of understanding of risks stemming from market dislocations crystallising and significantly impacting short term liquidity. The rising interest rate environment (i.e. monetary tightening) led to HTM assets losing significant value and this put a lot of strain on SVB’s finances

• Concentration of loans in the Technology Sector: SVB’s loan portfolio was heavily concentrated in the technology sector, making it particularly vulnerable to the sector’s cyclical downturns. When tech valuations fell, SVB faced increased risk of loan defaults

• Run on the Bank: In March 2023, as concerns about SVB’s financial stability grew, a few large uninsured depositors began withdrawing their funds and this quickly spiralled into a major run on SVB. This rapid outflow of cash further depleted SVB’s liquidity position, leading to its rapid demise

• Risk Governance: The board of directors at SVB failed to ensure that senior management were accountable for the management of interest rate and liquidity risks. This was also a key factor leading to the firm’s demise

In the following sections, we focus primarily on HTM assets and liquidity risk management, discussing the implications of the SVB collapse for the industry, the key challenges faced by financial institutions and how the industry should be responding.

IMPLICATIONS FOR THE INDUSTRY

The collapse of SVB has significant implications for the industry in terms of regulatory scrutiny of HTM assets, the expectation of enhanced disclosures and a push towards better liquidity risk stress testing and behavioural modelling. The aim is to reduce the overall risk and contagion (i.e. ripple effects) that the collapse of a major institution can have on the broader economy.

Increased regulatory scrutiny

In the wake of SVB’s collapse, regulators such as the Federal Reserve [1] have increased their scrutiny of banks in terms of their management of interest rate risk and their classification of assets as either Held to Maturity (HTM) or Available for Sale (AFS). The rising interest rate environment exposed fragilities in bank risk management, particularly with regard to concentration in long duration HTM assets. The concern is that banks may be classifying securities as HTM to avoid marking them to market, which would then expose unrealized losses. Regulators will now demand stricter risk management practices and more diversified asset pools, especially high quality liquid assets (HQLAs) combined with diversified funding pools that are not concentrated in uninsured deposits.

Enhanced disclosure, governance and monitoring

Regulators will require greater transparency from banks on the composition of their HTM asset pools, including governance and controls in terms of ongoing monitoring of their asset flagging process. There is a push towards a formal boundary, with a more prescriptive statement of intent and ability to hold the HTM securities till maturity. Regulators want to reduce the ability for arbitrage between accounting treatments to classify securities at amortised cost vs. market value. Greater transparency would also help investors to better understand the risks associated with banks. Furthermore, market participants will have better certainty over the availability of HQLAs across the maturity spectrum, so they can better manage and optimise their balance sheets.

The need to significantly improved liquidity risk stress testing and behavioural modelling

There is an increased expectation from regulators for banks to significantly improve their liquidity risk stress testing and behavioural modelling capabilities, focusing on:

• The formulation of different scenarios for portfolio adjustment as interest rates increase, particularly the ability to liquidate HTM assets during periods of stress across entities, sectors and countries

• The concentration of funding sources, particularly unsecured deposits and unsecured wholesale funding; including assessing factors such as instrument types, markets, currencies, term structure and market access

• Assessment of the propensity of asset reclassification from AFS to HTM, especially during high interest rate environments

CHALLENGES AND THE WAY FORWARD FOR BANKS

There are different challenges facing banks in liquidity risk management, following the collapse of SVB. The key challenges and how the industry should respond are discussed below.

1. Improving the control framework and ongoing monitoring for HTM assets

Following SVB’s collapse, regulators are taking a more critical look at how banks classify their HTM securities. Banks will be expected to apply stricter criteria for determining whether an asset meets the HTM criteria, ensuring that only securities with a genuine intention to hold to maturity are classified as such.

Banks should ensure they have:

• Strong governance in terms of policies and procedures around ongoing monitoring of asset classifications (HTM vs AFS)

• Robust asset flagging controls at inception to enable ongoing monitoring of any changes in the classification (i.e. data lineage). The completeness, accuracy and timeliness of asset flagging are very important from a monitoring perspective

Banks should leverage technology to enhance their HTM portfolio management capabilities and manage HTM risks more effectively. Data analytics tools and cloud-based technologies should be employed to identify and monitor potential risks and optimise HTM portfolio composition. While the traditional HTM concept of holding assets to maturity remains relevant, the focus should be on adopting a more nuanced and risk-aware approach that balances the potential benefits of HTM with the need for transparency and effective risk management.

2. Effectively managing HQLA composition and diversification

The effective management of a bank’s HQLA asset pool, including HTM vs. AFS assets impacts the Liquidity Coverage Ratio (LCR) – i.e. whether the bank has sufficient HQLAs to meet outflows during a 30-day period of stress. This requires dynamically managing the HQLA asset pool, based on changing economic and market conditions.

With this in mind, banks should ensure that they:

• Set limits on HTM assets as a percentage of total assets to control the concentration of the asset pool in HTM assets.

• Monitor the overall level of diversification of the asset pool, including Level 1 and Level 2A HQLA assets.

• Monitor the Level 1 assets in a more granular manner to assess the split between reserve balances, short term treasuries, long term treasuries and mortgage securities and their associated asset classifications (HTM vs AFS).

• Monitor the Level 2A assets in a more granular manner to assess the split between government agency debt and agency mortgage-backed securities and their associated asset classifications (HTM vs AFS).

• Understand and manage their HQLAs across the yield curve spectrum (short term vs long term) and the overall duration of these assets. A big focal point is the exposure to long duration assets and propensity to transfer assets from AFS to HTM during high interest rate environments.

• Assess the impact of HTM concentration (amortised cost) on IFRS9 provisions.

• Assess the impact of HTM concentration on the leverage ratio and regulatory capital requirements.

Diversification is important for banks when managing their HQLA pool, with Environmental, Social and Governance (ESG) considerations becoming increasingly important. Banks should consider broadening their HQLA asset pool to include:

• Green bonds whose proceeds are ringfenced primarily for environment projects.

• Sustainability linked bonds (SLBs) where the coupon paid is linked to an issuer’s achievement of sustainability linked Key Performance Indicators (KPIs) for example achievement of carbon emissions targets.

3. Refining liquidity risk stress testing approaches

Banks should refine their internal liquidity risk stress tests to assess the impact of a rising interest rate environment through:

• Incorporating different scenarios in their internal liquidity adequacy assessment process including asset portfolio adjustments, carefully evaluating the ability to liquidate assets during periods of stress (e.g. short term vs. long term and HTM vs. AFS assets). The Fundamental Review of the Trading Book (FRTB) [2] provides a good starting point to assess liquidity horizons for different asset classes.

• Understanding the impact on the LCR by assessing whether the HQLA pool composition is actually sufficient to cover outflows for a sustained 30-day period of stress. Concentration in HTM securities is a major cause for concern in a rising interest rate environment as these are booked at amortised cost and not market value.

• Assessing the diversification of the funding pool during periods of stress through:

  • Evaluating the overall share of uninsured deposits that are prone to high drawdown risk.

  • Revisiting the assumed drawdown assumptions for deposits during periods of stress.

  • Understanding the relative stickiness of retail and wholesale deposits (including operational and non-operational deposits) and the split between them.

  • Understanding the relative stickiness of financial and non-financial deposits and the split between them.

4. Developing behavioural models for HTM asset reclassification

Behavioural models are statistical models that can be used to predict the future behaviour of a security based on its historical behaviour. These models can be used to identify securities that are more likely to experience significant price movements, which can then be reclassified into a higher risk category.

Banks should develop models that focus on the propensity of AFS to HTM reclassification during high interest rate environments which are driven by factors such as:

• Share of uninsured deposits.

• Holdings in higher duration long term assets (e.g. Treasuries and mortgage securities).

• The leverage ratio.

Developing behavioural models for HTM security reclassification is a complex challenge, as it requires banks to collect and analyse large amounts of historical data. Additionally, banks need to develop robust algorithms that can capture the complex relationships between different market factors.

Machine learning (ML) can play a key role in developing behavioural models for HTM security reclassification. ML algorithms can be used to identify patterns in historical data that are not easily detectable by traditional methods. Additionally, ML algorithms can be used to automatically update models as new data becomes available.

With this in mind, banks should:

• Invest in data collection and analysis: Banks need to invest in systems and processes that can collect and analyse large amounts of historical data. This data can then be used to develop and train behavioural models.

• Develop expertise in ML: Banks need to develop expertise in ML to create and maintain effective behavioural models. This expertise can be acquired through effective training and development programs.

• Ensure that behavioural models are effectively validated: An independent model validator should evaluate whether the behavioural model is fit for use by assessing conceptual soundness, design, implementation, usage, ongoing monitoring, data management practices, documentation quality and the robustness of the control framework. There should be effective challenge of the model to ensure that the model is in line with industry best practice and the validation documentation should stand the test of internal audit and regulatory scrutiny. This will help to ensure that behavioural models are consistent and accurate.

5. Enhancing liquidity risk change management and governance processes

A failure of effective risk governance was also a major contributor to SVB’s collapse. Banks should enhance their liquidity risk change management across the lifecycle covering requirements definition, user acceptance testing (UAT), production deployment and ongoing process improvement from data acquisition through to reporting. Furthermore, banks should also improve their overall risk governance processes. To achieve the above, banks should focus on the following:

• Improve data governance: Banks should ensure that the completeness, accuracy and timeliness of data feeding from source systems to liquidity risk models and reporting systems is robust. Clear data ownership and data quality monitoring are key in this regard.

• Enhance requirements definition and testing processes: Banks should ensure there is clear requirements traceability for changes to liquidity risk management systems, models and reporting platforms, with robust test plans and acceptance criteria prior to production deployment.

• Enhance liquidity risk policies/procedures: Banks need to uplift existing liquidity risk policies and procedures to include HTM asset monitoring and enhancements to liquidity risk identification, assessment, mitigation and governance. This should also include enhancements to stress testing, with a clear articulation of management actions during periods of stress.

• Facilitate senior management understanding of liquidity risk: Banks will need to ensure that risk managers continually educate senior management through the delivery of structured workshops to drive an understanding of:

  • The key strategic levers (e.g., portfolio composition and risk management activities) that influence HTM asset allocations and liquidity risk metrics.

  • The key elements of behavioural model design, assumptions/limitations, implementation, usage and performance.

• Improve reporting and drilldown of liquidity risks: Banks should create dashboards for the ongoing monitoring of HTM related risks, to drive risk and portfolio management decisions. The reports should facilitate explanations for day-on-day movements of key metrics such as the LCR, leverage ratio and regulatory capital.

Overall, as part of the change lifecycle, banks should leverage technology and automation to enhance their overall liquidity risk management capabilities. This includes utilizing data analytics tools, stress testing software and automated reporting systems to improve efficiency and risk oversight.

SUMMARY

The collapse of SVB has major implications for banks in terms of liquidity risk management. Concentration in Held to Maturity (HTM) assets booked at amortised cost rather than market value, combined with a major run on uninsured deposits and failings in risk governance were major contributors to SVB’s demise. This has led to increased regulatory scrutiny, enhanced disclosure expectations and a need to significantly improve liquidity risk stress testing and behavioural modelling capabilities.

To address these challenges, banks should:

• Improve the control framework and ongoing monitoring for HTM assets.

• Effectively manage HQLA composition and diversification.

• Refine liquidity risk stress testing approaches.

• Develop behavioural models for HTM asset reclassification.

• Enhance liquidity risk change management and governance processes.

By taking these actions, banks can significantly mitigate overall liquidity risks and enhance their management of HTM assets, thereby strengthening their financial resilience, enabling them to navigate market disruptions more effectively and protect the overall stability of the financial system.

HOW 4MOST CAN HELP

Founded in 2011, 4most has grown to become one of the leading independent credit risk, market risk and actuarial consultancies in Europe and the Middle East. 4most’s team of risk experts can help banks with liquidity risk and HTM asset monitoring, risk governance, control assessment, gap analysis, data management, stress testing, behavioural model development/validation, documentation enhancement and delivering customised risk training.

For further questions regarding liquidity risk and HTM implementation, please don’t hesitate to contact us.

Ram Ananth – Head of Market Risk: ram.ananth@4-most.co.uk