Lee-Carter
\li\-\ˈkɑrtər\
The Lee–Carter model is one of the most widely used statistical models for forecasting mortality rates and analyzing historical mortality improvements. It was first introduced in 1992 by Ronald Lee and Lawrence Carter.
Under this model the average trend in mortality improvements seen historically is assumed to continue. The model expresses mortality rates as a combination of an age shape (baseline mortality by age, averaged over time) and a time trend (mortality changes over time, captured by time series methods). In practice, the model is fit using historical data and the age-shape, age-sensitivity and time parameters are used to project mortality rates into the future. Projections can be made with either a deterministic or stochastic approach.
Variations of Lee-Carter have been developed to account for the golden-cohort effect.
Applications of Lee-Carter include annuity pricing, pension liabilities, swaps, hedges and insurance products.