Estimation And Application Of VaR Based On Copulas Of Price Risk In Invetory Financing

In recent years, Commercial banks have been developing new business models to alleviate the intensifying competition between commercial banks and the saturation of traditional business. Financing has been a problem restricting the development of SMEs in China, and therefore, as the main body of financing, commercial banks have been playing a very important role.After the outbreak of the global financial crisis national financial institutions and commercial banks have adopted the credit crunch policy. Both "BaselⅢ" and CBRC proposed higher requirements on capital ratios and liquidity and put forward higher risk regulatory standards on the new business of commercial banks, including supply chain finance.As the core business of supply chain finance, inventory financing has been playing an irreplaceable role. From the existing practices and theoretical researches, the risk management of inventory financing has been the key constraints of its development. Although commercial banks could lower down the financing risk because of the self-liquidating pledge and shared risk by the logistics enterprises in inventory financing, this multi- participants'behavior and the existence of inventory pledge undoubtedly increase the difficulty of the business risk management. The risk measurement of inventory financing is an important part of its risk management.The quality, value and liquidity of the pledge are most concerned by lenders and logistics enterprises in inventory financing. If the standard pledge is selected, the value risk then becomes the major risk in the business. Under the government's macro-regulation on economy, the value risk of the pledge is mainly reflected in the price risk. Different kinds of pledge with their randomly-fluctuating prices make the control of the business difficult.In the pledge portfolio, the price risk is caused by price volatility of pledge portfolio which is under the influence of the risk factor. The thoughts of integrated risk management can be brought in, if the price volatilly of a single pledge of is taken as a separate risk. Therefore, from the angle of commercial banks, in roder to consider price risk Measurement of the pledge portfolio ininventory financing, the article makes full use of existing technology which can integrate financial risk.This paper adopts the basic modeling idea to build the VaR calculating models through Copula theory. The idea that "BaselⅡ" presents is how to calculate the VaR of the portfolios. In "BaselⅡ", the models used in the simulations belong to the Dynamic Stochastic General Equilibrium family of the last generation, and the data we used in the models must be over three years. Using a Copula, single risks that are estimated separately can then be combined in a joint distribution that preserves the characteristics of the single risks. It has revealed to be a powerful tool to solve the financial problem.In the empirical research, the paper assumes the pledges (1#copper & AOOaluminum) as standard products. Then the realistic data can be obtained from the National Spot Exchange to compute the models. Two non-parametric estimation methods are adopted, namely the empirical distribution function and kernel density estimation,to establish the joint distribution function between the yield rates of pledge portfolio. First, KS, JB and Lilliefors testing methods are used to verify that neither of yield rates of the two obeys the normal distribution. Then kernel density function (Gaussian) is employed in the kernel density estimation of the sample distribution so as to determine the marginal distribution of the pledge. Based on empirical distribution function of the yield rate of pledge portfolio, we have t-Copula. Frank-Copula and normal-Copula as the model to be selected to describe the joint distribution of the sample. Finally, by calculating the Squared Euclidean Distance, t-Copula is chosen as the one to describe the joint distribution of the 1#copper & AOOaluminum.In the research, Matlab is used to form the joint distribution on the two confirmed pledge on the basis of plentiful historical data, and then Monte Carlo method is adopted to simulate the process to obtain the calculation results-the VaR of the price risk of the pledge combination in inventory financing. Copula-VaR is more accurate compared with the traditional VaR. With the confidence level as 99%. the days when VaR failed is less than 90% and 95%. indicating that a high confidence level will bring less risk. The simulation process shows that the accuracy of results will be improved as the quantity of pseudo random numbers increases.