Research On Asset And Liability Management Of Life Insurance Companies In China

The important features of modern insurance is the equal importance of underwriting and investment business.However,caused by competition and decline of underwriting profit,investment profits have increasingly become the important pillar of life insurance company.Life insurance funds has the characteristics of long maturity,its allocation is subjected by company's liabilities.Modern life insurance company mainly adopt the method of assets and liabilities management(ALM).Based on the anticipate of assets yield,liability cash flows and other variables,under the safety and liquidity requirements,ALM match the assets and liabilities in quantity and term structure characteristics to improve the life insurance funds investment returns and achieve the goal of profit maximization.The capital allocation is the key of ALM.Compared with the pure asset model,ALM method take many practical factors into consideration,such as the stochastic of liabilities,policy constraints,tax and transaction cost.Which is the best strategy framework for some long-term investors who need to repay liability in the future and obtain certain investment targets.Multi-stage stochastic programming method is an important branch of operational research,which can be used to make the dynamic stochastic decisions.This article based on the multi-stage stochastic programming approach to study the construction of asset liability management model that suit our country's life insurance companies.And use the framework and multi-stage stochastic programming ALM model for empirical research,which demonstrates the practicability and validity of the model,and achieved the purpose of this study.This article is composed of seven chapters:The first chapter mainly introduced the research background and significance,the relevant domestic and foreign literature,and expounds the research ideas,methods and innovation points of this article.The second chapter firstly analyzed the feasibility and necessity of life insurance fund investment and combed the principles and channels of our country's life insurance fund investment.After the review the life insurance fund investment?history and current situation.this chapter analyzed the risk of life insurance type and level of fund investment,and the adoption of assets liabilities management method.The third chapter reviewed the ALM theory and several technologies,expounded the new developments of ALM model,especially the model based on stochastic program.And built the ALM framework that suitable to China's life insurance companies.Finally,built a life insurance company multi-stage stochastic programming ALM model and framework,discussed the objective function?constraints and solution method.The fourth chapter mainly studied assets scenario generation model and method,after compared the existing assets scenario generation model and scenario tree structure method,this paper proposed an assets yield forecast mode which suitable for China's financial market,and studied asset scenario generation methods and steps.The fifth chapter mainly studied liability scenario generation model and method.Based on the insurance actuarial theory and the model point method.In this chapter,we constructed the stochastic variation equation of all kinds of insurance liabilities,such as insurance policy quantity,death payment surrender payment on the considered of policy heterogeneity.In addition,we construct a model of liability insurance with both additional bonus and cash bonus account,and build a liability scenario generation model of premium,payments,and actuarial reserves.Chapter 6 focuses on China's life insurance company ALM application of multi-stage stochastic programing,which is a multi-stage linear stochastic programming model with simple resources.The liquidity risk,investment risk and bankrupt risk are represented by the violation degree of constraints.To obtain the Wealth growth stability requirements,we use the net asset after punishments at each decision point as sub-object functions,added all the sub-object functions by weights as the model's main object function.After compared the results of different payment forms of life insurance and participating life insurance ALM model and the results of different forms of the objective function,the results proved the stability,practicability and effectiveness of the model.Chapter 7 mainly summarizes the main research work and achievements,future research direction.The innovation of this article mainly includes the following 3 parts:This paper constructed a dynamic stochastic ALM model,which suitable for China's life insurance companies.The method fully considered the characteristics of assets and liabilities of life insurance company in China.Model also contains several practical problems in lifeinsurance company management such as balance sheet types,transaction costs,regulatory constraints.Empirical results found that time payment and official participation rate can achieve a higher level of total asset and equity,which have guiding significance to the life insurance company's actual management.We built a liabilities dynamic model based on cash dividend mechanism.The existing research of insurance ALM are almost aimed at traditional life insurance products,there is little participating life insurance ALM research based on cash bonus payment.However different bonus payments have different impact on cashflow and allocation decision.In this article,we compared different accounts allocation decisions.The results shows that single premium accounts tend to allocate more fixed-income assets,such as bank deposits and bonds.In addition,the participating life insurance accounts allocation strategy tend to be more aggressive,the proportion of equity assets is relatively higher,which can obtain higher returns in the same situation.We took the stationarity of wealth growth into consideration to construct the objective function.When the literature use wealth maximization as the objective function,it only considered the final wealth maximization of the investment planning period,whcin may increase the volatility of wealth growth process.For this reason,we use the net asset after the punishment at each decision point as sub-object function,and added all the sub-object function by weight as the model object function.The empirical research results in the sixth chapter show that the object function we proposed lead to a more flexible decision,and can effectively improve the account assets and accumulated wealth value.