The Up To Data Evolutional Study Of The Financial Risk Existence And Measurement

With the globalization and integration of economy, the innovations of finance and improvement of technology, basic and structural changes have taken place in the financial market of the world. At the same time, faced with stronger fluctuation and more system risks, management of the financial risks have become key abilities for financial institutions and industrial and commercial engineering and modern financial theories. The process of the financial risk management is very complicated, generally speaking, it includes risk identification, risk measurement, technical selection and implementation of risk management and risk control four stages. The four aspects connect up properly with each other and each is very important. But among the four aspects, the risk measurement and control are the most important parts. The original purpose for creating the financial derivative tools is to hedge the financial risk, and in faith the tools can hedge the risk caused by the varieties such as interest rate, exchange rate and time etc for people. However, due to the varieties, complications and level effects, etc, of the financial derivative tools, it can provide more opportunities for speculating, and thus it can provide more complicate financial risk. With the increasing of the financial derivative tools, the virtual financial market scale begins more and more enormous. All those increase the difficulty of measuring the financial risk and one has to need much more and better measuring methods. The pricing problem of financial derivative tools is involved in both hedging the financial risk and business of the virtual financial market. If one overestimates the price, the financial bubble comes into being and it can affect the stability of the society, even can bring up the financial crisis. Moreover, since the price of the financial derivative tools is not stable by the impact of many financial datum marks (such as interest rate, exchange rate, price of assets), the reasonable pricing of the financial derivative tools can be seen in generally as the measurement methods of the financial risk. In fact, because of the fluctuation of the price of the financial derivative tools, one can gives the reasonable price and farther measurement and management of the risk by describing the possible movement of the price. Hence we mainly investigated some important financial derivative tools including financial bonds (Zero Coupon Bond, Mixed Bond), financial swaps, financial options and financial forward and financial future. We mainly consider the pricing formula about those financial derivative tools. So we especially investigated the European option price under the every day price control. For the financial forward and future contracts, we gave the formula of financial future price under the stochastic interest rate, and gave the corresponding ways of risk measurement. There are some traditional methods for the financial risk measurement such as the simple arithmetic method, Mean—Square method and the VaR method, etc, which was pulled out and developed ten years ago. Those methods have their respective advantage such as being understood easily and computed simply etc, especially the VaR method gets much more application for its advantages of computing correctly. However, these methods havemany shortcomings such as failing to reflect the direction of the return rate deviated from the expectation, they can not provide an accurate measurement of the lost and they have few applications. Since the βcoefficient method which came from the classical CAPM theory can overcome those shortcomings with the recent development, it has got extensive application. This article systematically introduces the theories of the origins of the βmethod, the classical CAPM model, the property and character of the βcoefficient and the methods for estimating and forecasting the βcoefficient. Of course what we pay more attention to is this method's development in recent years—the time variety of βcoefficient. Since the βcoefficient is not always be constant, and the traditional methods with constant βvalue can't interpret much phenomenon commendably and completely in finance, scholars generally agree that the βcoefficient is time varying. However, it is a difficult problem on how to get the βcoefficient of time varying, and there is not the best method currently which can be generally accepted. In this text, we simply introduce several models that can get the βcoefficient and interpret economic phenomenon. These models mainly include: Conditional CAPM model, threshold CAM model and SS model. One advantage of applying the model of time-varying βis that it accords with the practice, another is that it expands the original methodological application. It not only can help people to measure the risk but also help people to acquaint the reasons why causing the risk, and with the knowledge about the risk the investors can invest well. We introduce the result about those at the end of part two. In addition, in view of β-method may has different exhibitions in different market environments, we made a empirical studies about somestocks in Shanghai and Shenzhen markets of stocks by using new models in part three of this article. We validate the conclusion mentioned above by our investigation. Furthermore, we are aware of that the understanding correctly of the systemic risk of the bull and bear markets can help invertors to invest well. Since the model we bring out overcome the shortcomings of other models which be applied to this problem, it can increase the ability of understanding the real phenomenon. Besides those well known methods mentioned above which can measure the financial risk, there are many tools can measure the risk better when characterizing the financial market for example, if one can makes a correct model about the financial assets, the conclusion has got can help one measure the financial risk better. Hence, it is important for making an accurate model and reasonable risk measurement by using sufficiently the known information to obtain the parameter estimation. In general, sufficient and continuous data is the precondition in the most empirical studies about the estimation of the parameter. However, what one always meets when dealing with the data in practice is the problem that some data are missing. Usually there are two methods be used for this problem: When the data is enough and the remainder is consecutive, we cast the missing data; When the data is not enough, we use some other messages to find the missing data, for example, by linear interpolation methods. But the two methods have shortcomings respectively, for instance it need much data, compute incorrectly and can not use the message sufficiently. In the last part of this article, we hoped to improve on the parameter estimation method with missing data. We investigate how to use messages enough that has been obtained to estimate the parameter of time series model when it has missing data. By using EM algorithm, we gave the parameter estimation of ARMA(1,1) model. We also gave the calculate process when...