Research And Application Of Genetic Algorithm And Grey Forecast Model

The institutional investors to invest in quantitative satisfied. Because of quantitative investment income stability, risk control, etc. Quantitative investment involves the quantitative stock selection, quantitative matching, quantitative timing, statistical arbitrage, asset allocation and so on every link of market transactions. Quantitative investment has become a very important part of market investment.This article is divided into five chapters. First,this paper introduces the background of financial quantitative trading, and expounds the quantitative trading in Chinese stock market has huge development potential. Then this paper introduces and deeply at three market trading model- genetic algorithm, the grey system, matching trading. In the complicated stock market, using the genetic algorithm can find out the largest influence, and the highest correlation factor, and according to the factor to pick out the best stock. The traditional genetic algorithm to deal with the problem of large amount of calculation convergence speed too slow, even appear the problem such as local convergence. So if you want to have a very good result on the stock market using of the genetic algorithm. you need to improve it. This paper introduces the fitness function. This make the original genetic algorithm to improve the convergence rate greatly. The empirical analysis, the accumulated earnings of the stock market is also very good. Grey prediction GM(1, 1) model is proposed, which is mainly aimed at "poor information and small sample" system prediction model. According to China’s national conditions, if you want to take advantage of the grey prediction model to predict a good stock market, you must improve it in the process of building model. Variations are limited within 10% of the stock market. This paper offered two improvements. First, the recursive relation model is improved and the algebraic recursive equation is constructed. Second, proximate formula into three adjacent, according to the automatic optimal weighting formula introduced their corresponding coefficient. This paper makes an empirical analysis to this method. Then, this paper made indepth at pairs trading strategy. Pairs trading strategy has become a kind of common trading strategy in the market. Two similar movements of foreign currency to the euro and the British pound as paired trading object, and through the least-square method and the cointegration relationship against the euro and sterling did detect the internal connection of the two kinds of foreign exchange. Finally, this paper presents a grey genetic algorithm model building and empirical analysis.This paper mainly combines the genetic algorithm and grey system theory. First, this paper using genetic algorithm to choose the trading of stocks. Then, based on grey model to choose trading time. Application of fully shows that the method is feasible.