摘要:Since its introduction, the Lee Carter model has been widely adopted as a means
of modelling the distribution of projected mortality rates. Increasingly attention is
being placed on alternative models and, importantly in the ¯nancial and actuarial
literature, on models suited to risk management and pricing. Financial economic
approaches based on term structure models provide a framework for embedding
longevity models into a pricing and risk management framework. They can include
traditional actuarial models for the force of mortality as well as multiple risk fac-
tor models. The paper develops a stochastic longevity model suitable for ¯nancial
pricing and risk management applications based on Australian population mortality
rates from 1971-2004 for ages 50-99. The model allows for expected changes aris-
ing from age and cohort e®ects and includes multiple stochastic risk factors. The
model captures age and time e®ects and allows for age dependence in the stochastic
factors driving longevity improvements. The model provides a good ¯t to historical
data capturing the stochastic trends in mortality improvement at di®erent ages and
across time as well as the multivariate dependence structure across ages.