| For the traditional commercial banks whose main source of income is interest margin of deposit and loan, the loan allocation will directly affect commercial bank's operating income. In the face of the basic completion of the reform of the interest rate market and the gradual reduction of loan spreads, the increasing of non-performing loans ratio caused by the mismatch of loans and other issues are more serious. As of the end of 2015, China's domestic commercial banks non-performing loan ratio is 1.67%, increasing by 0.08 percentage points compared with the end of 2014. Therefore, carrying out the study of loan allocation is of great significance for the commercial bank, specifically, reasonable loan allocation will enable commercial banks to achieve the maximum profit while keeping the overall risk manageable. However, most of the existing research focus on the "incremental" loan allocation problem, and ignore the huge "stock" loan risk which is prone to black swan events.This academic dissertation is consists of four parts. The first chapter is the introduction part in which the nature of the research on loan allocation problem based on exponential spectral measures of risk, significance of this topic and the background of this topic is discussed, meanwhile, the existing research on loan portfolios optimization is discussed in detail. In the second chapter, the Markowitz mean variance model, VaR model and CVaR model are introduced, which lays the theoretical foundation for the establishment of the exponential spectral model in the third chapter. The chapter three is the method of loan allocation based on exponential spectral measurement. In this chapter, we introduce the shortcoming of existing research on commercial loan allocation model, that is, ignoring the considering the constraint of tail risk control and the stock of loans which cause huge losses to banks. Then, we build the model of exponential spectral model, which overcome the drawback of CVaR model for measuring the tail risk. The fourth chapter is the risk measure Em loan allocation model and the empirical analysis based on index of spectral, through loan data of a commercial bank credit database for empirical research, loan allocation model is established and verified based on exponential spectral risk measure EM model is effective and superior, and the increment and stock lending allocation model based on incremental model efficiency improvement.The main work of this paper:(1) We have established a loan allocation model based on exponential spectral risk measure EM to configure the tail risk of loan portfolio, and to give the risk weight of the tail risk, the greater the risk loss, the greater the weight. By comparing the model with the CVaR model and the geometric model, this paper shows that the model is more strict to the tail risk control.(2) We set exponential spectral measures of risk, value at risk VaR and bank available positions as the constraint conditions, set 0-1 programming as a decision making tool, meanwhile we take into account the stock of loans loan portfolio and increment loan portfolio overall goal of return and risk to establish a loan portfolio allocation model based on the stock and increment loan. Empirical analysis shows that the risk of "stock+incremental" loan portfolio allocation model is lower than the existing research, at the same time, the earning of "stock+ incremental" loan portfolio allocation model is higher that the existing research in which incremental loan is only considered.The main innovation and characteristics of this paper:(1) we measure the tail risk of commercial bank loan portfolios by exponential spectral risk, changing the existing researches fail to control tail risk.(2)we set up "stock+incremental" loan allocation model which control the risk of incremental and stock loan, changing the existing research which neglect of the huge tail extreme risks, and reducing the risk of the loan portfolio, enhancing the proceeds of the loan portfolio. |
PDF Full Text Request