Interest rates,as core variables in financial markets,have crucial impacts on the sound operation of life insurance com-panies.This paper calculates the Hurst exponent of domestic interest rate sequences with different maturities and designs Wald statistical tests and shuffle sequence tests to identify the significant long memory characteristics of interest rate sequences.Build-ing upon this,the continuous-time long memory interest rates are modeled using the fractional CIR model,with parameter estima-tion conducted through neural network indirect inference.Moreover,nested random simulation methods are employed to study the impact of interest rate long memory on the assessment of interest rate risk economic capital for life insurance companies.The re-search findings reveal that interest rate sequences with maturities of one month and above exhibit significant long memory charac-teristics.Neglecting interest rate long memory would underestimate the mean reversion speed and volatility of interest rates.Fur-thermore,interest rate long memory increases the probability of extreme events,heightening the potential risk of losses for life in-surance companies and increasing the tail risk of economic capital assessment.Life insurance companies should meticulously ana-lyze and incorporate the long-term dependency structure of interest rates when conducting economic capital assessment to ensure sufficient foresight into potential tail risks.This will enable them to maintain financial stability and long-term value creation capa-bilities in the face of complex and volatile market environments.
关键词
寿险公司/利率风险/长记忆性/经济资本/间接推断
Key words
life insurance company/interest rate risk/long memory/economic capital/indirect inference
维普
万方数据