Modelling and Forecasting of Mortality in Russia Using Actuarial Stochastic Models

This paper is devoted to modeling and forecasting mortality in Russia using actuarial stochastic methods. In the modern world, average life expectancy is gradually increasing and Russia is not an exception. As a result, pension funds and life insurance companies face longevity risk. In order to estimate the amount of future payments correctly, one should accurately predict future death rates. This study examines stochastic mortality models, such as modifications of the Lee-Carter model and functional models. Stochastic modeling of mortality rates in Russia was previously poorly studied. Actuarial stochastic models were estimated for Russia based on statistics 1970-2014. Data 2015-2018 was used for backtesting. It turned out that to explain mortality in Russia it is necessary to consider the influence of birth year, known as cohort effect. Forecast of the best model for Russia in respect of those who are older than 45 years was used for creating cohort mortality tables. According to the forecast, in 2020 complete future lifetime comprises 5.75 years for males born in 1939 and it is 6.57 years for females of the same birth year. It was also shown that Russia has its own specifics of mortality in comparison with England&Wales and the United States.