Research Of Stock Price Trend Forecast Based On Time Series Analysis

With the rapid development of china's capital market and the great increment of people's income,more and more people put their money into the stock market hoping to keep and increase the value of their wealth.Although investing in stock market can bring you a great fortune,it is also very risky.The nonstop up and down movement of its price puzzles investors.They realize that the capital market is very difficult to understand.They desperately need a theory to explain the reason why the prices change.They want to understand the mechanism of the price movement.And they need scientific forecast methods which can guide their investment to avoid risk and get good return.Because of the great potential economic interest,the forecast of stock price which attracts the research of many scholars and professionals,in spite of its difficulty.In 1970's,Fama proposed the famous Efficient Market Hypothesis which concluded that the movement of stock price is a random walk,the price of stock is impossible to predict.Nobody can get extra return by means of analyzing the history of price and public information.However,recent researches show that this theory contradicts the facts.To better explain the movement of stock price,Peters proposed the Fractal Market Hypothesis which emphasizes the influence of information acceptance and time scale.And this theory points out that fractal structure exists in all stable markets.The price of stock is predictable in some degree.In the presupposition that stock price is predictable,the methods of stock price forecast are systematically introduced.They are traditional fundamental analysis and technical analysis,and modern forecast methods of ARIMA and GARCH model which are based on time series analysis are also introduced.Of course,because of the complexity of stock price movement,we can not predict the future price exactly with one single simple model.In this dissertation,the innovation is the empirical analysis of trend of Shanghai stock index based on ARIMA model and GARCH model.We find that the constructed models have excellent short-term prediction ability.They are useful to note the investors the risk and the opportunity of buying and selling.They have certain practical values.