Modeling And Analysis Of Asset-liability Management For Life Insurance Companies Of China

Because the historical reasons that China government controlled on the level of interest rates, and the asset-liability management of China's life insurance companies was forward, the new situation which the external economic environment was frequently changed exacerbated the assets-liability risk of China's life insurance companies. The former asset-liabilities management technology assumed that the interest rate was fixed, and can't be able to manage the asset-liability risk effectively, so that in the 1990s caused the serious interest rate loss. The former management method didn't consider any restraints either. Managers need to conduct assets and liabilities under the uncertainty of future changes. To enhance the management ability of assets and liabilities, this paper presents the application of stochastic programming model for assets and liabilities. Stochastic programming model for assets and liabilities can have a very good inspection of the uncertainty of income from assets and liabilities. It can take advantage of the restraints and the risk limits to choose the allocation of assets and liabilities in accordance with the long-term interests of the life insurance companies.Nowadays, most of the domestic researches of asset-liability management are on the qualitative research. Quantitative studies mostly focus on the interest rate risk management of assets and liabilities, such as Cash flow model and the Immune model. Both of them treat the interest rates as random variables, can manage the risk of interest rates, but don't consider other constraints of assets and liabilities in the allocation of assets and liabilities.The main text is divided into four chapters on content and viewpoint is:Chapter 1 mainly introduces the background for the research, the purpose and significance, the methods. It is thought that the risks of assets and liabilities of China's life insurance companies are increased, but the life insurance company's assets and liabilities management technology is relatively backward. Under this background, this article proposes to apply stochastic programming model in the asset-liability management of China's life insurance companies.Chapter 2 briefly describes the overall framework of asset-liability model, including the scenario generation and stochastic programming model. The scenario generation considers the uncertainty of the economic environment and applies the historical data to predict the future possible scenarios. Using the scenarios and the goal of the life insurance companies as input data, the stochastic programming model will choose the optimal allocation of assets and liabilities. There are several stochastic programming: Expected Value Programming, Simple Linear Programming with Compensation , Chance-constrained Programming, Dependent-chance Programming and Markov Decision Programming. Simple Linear Programming with Compensation is able to deal with the breach of restrictive conditions, so this paper use this model to manage assets and liabilities, and subsequently cites the success of the model application.Chapter 3 analyses the requirements of life insurance companies to establish asset-liability management model. To achieve the objective of profit maximization, it requires matching assets and liabilities, ensuring investment security and mobility of assets. In reality the main factors affecting the assets and liabilities are interest rates, mortality, surrender rate, the cost rate and the rate of inflation. Regulatory and other constraints will also affect the assets and liabilities. When constructing model need to consider the requirements of the model, it requires theoretical basis, the maneuverability and the ability to reflect the life insurance companies face environment. The data of the model needs to meet adequacy, relevance and timeliness. After proposing these requirements, this article established a Simple Linear Programming with Compensation of China's life insurance companies. To maximize profits at the same time, the model can take into account the capacity to bear risk, asset allocation constraints, asset accumulation constraints etc.Chapter 4 analyses the asset liability management model constructed in last chapter. The attention of application Model should be paid to the types of assets and liabilities, life insurance business management methods, the expansion of models and the ways of handling liabilities plan. Then the application of the model uses the time series of bonds, stocks, funds and bank deposits to generate scenario. Assuming capital gains subject to normal distribution, stochastic simulation technology is used to get the next two scenarios. These future scenarios as the future economic environment predictive value input model. The model is saluted to obtain the value of asset allocation and closing asset value. Finally the stochastic asset-liability model is compared with the immune model, the stochastic asset-liability model can deal with the risk-adjusted gain, transaction costs and allow binding constraint on the contrary better, but because the management of assets and liabilities is of complexity and variability, so model still have some limiting factors.The main innovation of the paper are:(1) In this paper, using the method of quantitative research manage assets and liabilities, the stochastic programming model can comprehensively consider the risks and benefits of assets and liabilities and then make decisions. (2) The concept of transaction costs has been added to the model, when selling and buying assets needs to pay a fee. By adding the concept of transaction costs, the stochastic asset-liability programming model is more realistic. In addition, the asset-liability management model also allows violation of certain constraints; the model can be weighed against the contrary of constraints and the allocation effect of assets and liabilities.The shortcoming of the paper are:(1) For the complexity and difficulty of calculating the model, model did not consider the exchange rate risk, so the model can not have assets in foreign currency assets. (2)Because the collection of life insurance company's internal data is difficult, so the data of cash flow and liabilities is assumed.The assumption makes that the results of model was not aimed.