Research On Portfolio Strategy Based On VaR And Genetic Algorithms

Markowitz portfolio theory proposed by marking the birth of modern investment theory, for the first time from the capital gains rate and risk are two angles to evaluate the portfolio, reflecting the people enough attention to investment risks, rather than just the pursuit of high returns. In fact, since the seventies of last century, the collapse of the Bretton Woods system, the major developed countries in order to quickly develop and gradually financial deregulation, the market appeared a large number of derivatives, the feed does not further increase the financial risks. The urgent need for a more comprehensive portfolio theory to guide investors in understanding the financial markets from a variety of angles, an accurate estimate of the benefits and risks of investment activities, which are required to support the traditional financial theory and new measurement methods. Therefore, portfolio theory has also been improved.VaR as a monetary amount to quantify the degree of risk in a portfolio of assets, the market is a low predictive value quantile portfolio value changes the probability distribution of a given period, and thus VaR is a downside financial risks (losses) of a monetary measure. All financial instruments portfolio VaR we can hold the bank obtained by calculating a single measure of the bank’s market risk of all positions. In addition to the Basel Committee to allow qualified financial institutions can use VaR method to estimate their capital adequacy ratio, and promoting the establishment of international standards VaR measurement. On the other hand, financial institutions fully aware of the importance of risk management, you need to configure the appropriate funds based on the risk profile of its operations in order to ensure safety. So the introduction of VaR financial institutions were put on the agenda of the risk management and strengthen theoretical study of VaR is significant for financial institutions. This paper introduces the traditional three VaR calculation methods, including Delta-normality, historical simulation and Monte Carlo simulation method. And carried out backtesting test and found three methods can not be measured by back testing, so the need for VaR calculation method of correction. CAMP in the classic, β coefficient is to assume the same. However, a large number of documented stock β system with time-varying, whether it is economic growth and other macroeconomic factors, or their own accounting variables and other factors have an impact on the β coefficients. Therefore, the study of β coefficients, when degeneration is an important factor that must be taken into account, the paper classical CAPM model state transition model, using the Kalman filter estimate β coefficient, variable VaR β coefficient calculation based on risk more accurate measurements. And introduced into the portfolio, building a dynamic portfolio of various types of investor’s investment practice guidance. Therefore, the practical value of this study is obvious.Markowitz portfolio theory emphasizes the variety of combinations of securities by the correlation reduces the risk of multiple securities portfolio, respectively, with a description of the expected return rate of future earnings, with a standard deviation describe future risk level. The standard deviation may reflect a combination of risk that may occur, but can not describe the possibility of risk, VaR is specified at a certain confidence level of expected losses and the confidence level is set to reflect the investor’s risk appetite, the easier to investors investment choices. Therefore, based on the estimated VaR can be used as a supplement to the classical Markowitz theory. By adding in the Markowitz portfolio theory VaR constraint conditions can be more rigorously calculate portfolio weights. Thus Based on VaR and multi-objective genetic algorithm portfolio strategy research, has great theoretical significance.