Mathematical and statistical assessment of the solvency of financial institutions

金融机构偿付能力的数学和统计评估

• 批准号: RGPIN-2016-05002
• 负责人: Boudreault, Mathieu
• 金额: $ 1.09万
• 依托单位: Université du Québec à Montréal
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2019
• 资助国家: 加拿大
• 起止时间: 2019-01-01 至 2020-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=675599
• 关键词: Mathematical; statistical; assessment; solvency; financial

Insurance companies, (investment) banks, pension plans and governments are exposed to several risks that can threaten their ability to protect the savings of millions of Canadians. Therefore, this research program is intended to assess the solvency of financial institutions in three different areas where they could fail.******First, we will look at how life insurance companies manage exposure to equity and interest rate risks in a popular line of business known as segregated funds (variable annuities). Using mathematical models that simulate realistic scenarios, the intent is to measure how common risk management strategies typically work and how we can improve risk management whenever modelling assumptions do not match reality. We would like to raise awareness among actuarial researchers on the mathematical aspects of managing these products. We also seek to provide general guidance to insurers and regulators as to how risk management can be improved to ensure that the savings of retirees are better protected.******Second, corporate bonds are very commonly used by banks, insurers and pension plans to manage their liabilities. But the lessons we learned from 20082009 is that even investment-grade bonds can be hit hard by credit risk. Therefore, we seek to improve the assessment of credit risk in these bonds. We propose statistical methods to accurately estimate the creditworthiness of corporations using prices of credit derivatives that are highly liquid and traded by experts (e.g. credit default swaps). We would like finance researchers and risk managers to use this approach as a complement to popular ratings-based methods. Given that trillions of dollars are traded in corporate bonds, the potential impact can be enormous, especially for Canadian insurers and pension plans.******Few people know that hurricanes have made landfall in Canada more than 25 times since 1950. They provoked storm surges, heavy rainfall and floods causing hundreds of millions of dollars in damage. Very recent climatology research has shown that hurricanes tend to migrate towards the poles (north or south), indicating that in a not-so-distant future, the risk posed by these storms is likely to increase over Eastern Canada. We will design and analyze actuarial and statistical models of hurricane risk in light of the connections between the frequency and intensity of these events. The impacts of this dependence relationship in catastrophe risk modelling and management have been rarely investigated, and this should help reduce the societal impacts of future hurricanes over the country.******In the past, the latter three risks have had an important impact on banks, insurers, governments, pension plans, etc. Therefore, this research program is meant to develop tools to assess and manage these risks with an eye to increasing the solvency of financial institutions and in turn, protect the savings of the public.

保险公司、（投资）银行、养老金计划和政府都面临着一些风险，这些风险可能威胁到它们保护数百万加拿大人储蓄的能力。因此，本研究计划旨在评估金融机构在三个不同领域的偿付能力，这些领域可能会失败。******首先，我们将了解人寿保险公司如何管理被称为隔离基金（可变年金）的流行业务线中的股票和利率风险。使用模拟现实场景的数学模型，目的是衡量常见的风险管理策略通常是如何工作的，以及当建模假设与现实不匹配时，我们如何改进风险管理。我们希望提高精算研究人员对管理这些产品的数学方面的认识。我们亦会就如何改善风险管理，以确保退休人员的储蓄得到更好的保障，向保险公司和监管机构提供一般指引。******其次，银行、保险公司和养老金计划非常常用公司债券来管理它们的负债。但我们从2008年至2009年的经验教训是，即使是投资级债券也可能受到信贷风险的严重打击。因此，我们寻求改进这些债券的信用风险评估。我们提出了统计方法来准确估计企业的信誉，使用高流动性和专家交易的信用衍生品（如信用违约掉期）的价格。我们希望金融研究人员和风险管理人员使用这种方法，作为流行的基于评级的方法的补充。考虑到数万亿美元的公司债券交易，潜在的影响可能是巨大的，尤其是对加拿大的保险公司和养老金计划。******很少有人知道，自1950年以来，飓风在加拿大登陆超过25次。它们引发了风暴潮、暴雨和洪水，造成了数亿美元的损失。最近的气候学研究表明，飓风倾向于向两极（北方或南方）迁移，这表明在不久的将来，这些风暴对加拿大东部造成的风险可能会增加。我们将根据这些事件的频率和强度之间的联系，设计和分析飓风风险的精算和统计模型。这种依赖关系对巨灾风险建模和管理的影响很少被调查，这应该有助于减少未来飓风对该国的社会影响。******在过去，后三种风险对银行、保险公司、政府、养老金计划等都产生了重要影响。因此，本研究计划旨在开发工具来评估和管理这些风险，以提高金融机构的偿付能力，进而保护公众的储蓄。