保險公司和退休金提供者所遭受的長壽風險日益嚴重，資本市場上的證券化商品是目前長壽風險的風險管理方法，而通常死亡率的證券化商品會連結死亡率指數，但由於曝險人口和避險人口的統計上的不匹配產生的基差風險，使得避險端無法完美避險，因此在使用死亡率連結的證券化商品時應該要有效地降低基差風險。本文主要提供一個架構去建構跨國死亡率模型，根據瑞士再保公司在2003年發行的死亡率債券所連結的死亡率指數，我們選用英格蘭與威爾士、法國、義大利和瑞士這四個國家的資料去建構模型，在跨國死亡率之時間效應考慮去趨勢和VAR模型來分析，跨國死亡率之世代效應則應用共整合分析的方法，本文選用誤差衡量的方法MAPE和RMSPE比較原始的方法和本文研究方法的配適和預測死亡率的準確性。 Pension plans and Annuity providers are subjects to the threat of longevity risk. Most popular way to mitigate the risk is to trade mortality-linked securities in capital markets. In the mortality-linked securities, basis risk often relates to mismatches in demographics between the exposed populations and the hedging population. We must reduce the basis risk to manage the longevity risk effectively. This paper introduces a new framework for modeling the mortality rates for a pair related populations to aim consistent mortality forecasts. We propose an Lee-Carter with cohort effect model to fit the mortality rates which incorporate Time effects and Cohort effects. A VECM model is derived by the cohort parameters in multi-countries. We deal with time parameters by detrend method and conduct the VAR model by detrended data. This study illustrates the fitting and forecasting accuracy in new and original methods.
