标题： “隐藏到到期”会计的不易察觉的风险：关于存款人挤兑和银行韧性 显示英文标题

标题： The not-so-hidden risks of 'hidden-to-maturity' accounting: on depositor runs and bank resilience

摘要： 我们构建了一个基于资产负债表的模型来捕捉挤兑风险，即在压力下从流动性缓冲中筹集资本的潜在能力降低，这由存款人的审查驱动，并进一步因提款而引发的抛售行为加剧。 该模型受到2023年3月硅谷银行（SVB）崩溃的启发，我们将模型应用于评估其违约前资产负债表脆弱性的积累情况。 更一般地，我们分析了资产负债表的哪些特征对于银行业监管机构充分评估挤兑风险和韧性是关键的。 通过将SVB的资产负债表时间序列数据引入我们的模型，我们能够展示资金来源和相应资产构成的变化如何使SVB容易受到挤兑风险的影响，因为它们越来越多地依赖于持有至到期的会计准则，掩盖了证券组合中的重估损失。 接下来，我们制定一个易于处理的优化问题，以解决持有至到期资产的指定问题，并量化银行在不进行重估的情况下持有这些资产的能力。 通过将该模型校准到SVB的资产负债表数据，我们揭示了该银行在2020年至2022年期间的资金风险和隐含的风险承受能力，直至其倒闭。 最后，我们通过First Republic Bank、US Bancorp和PNC Financial Services Group Inc.的资产负债表验证了我们的模型。 显示英文摘要

摘要： We build a balance sheet-based model to capture run risk, i.e., a reduced potential to raise capital from liquidity buffers under stress, driven by depositor scrutiny and further fueled by fire sales in response to withdrawals. The setup is inspired by the Silicon Valley Bank (SVB) meltdown in March 2023 and we apply our model to assess the build-up of balance sheet vulnerabilities before its default. More generally, we analyze which characteristics of the balance sheet are critical for banking system regulators to adequately assess run risk and resilience. By bringing a time series of SVB's balance sheet data to our model, we are able to demonstrate how changes in the funding and respective asset composition made SVB prone to run risk, as they were increasingly relying on held-to-maturity accounting standards, masking revaluation losses in securities portfolios. Next, we formulate a tractable optimization problem to address the designation of held-to-maturity assets and quantify banks' ability to hold these assets without resorting to remarking. By calibrating this to SVB's balance sheet data, we shed light on the bank's funding risk and implied risk tolerance in the years 2020--22 leading up to its collapse. We conclude by validating our model on the balance sheets of First Republic Bank, US Bancorp, and PNC Financial Services Group Inc.