Control Theory Analysis For Stock Trading Decision

In this paper,we study the monitoring performance of the control chart when the controlled process obeys the fundamental assumption that process data are independent and identically distribution.We mainly consider the process data are autocorrelation and heteroscedasticity.In order to solve the problem for autocorrelation process,we apply ARIMA model,and combined with the GARCH model to deal with the problem for heteroscedasticity process.The study of this kind of problems have practical significance for the users of control chart.We mainly undertake the following two studies:First of all,we study the monitoring performance of control charts by using average run length.We discuss the influence for heteroscedasticity process on the conventional EWMA control chart in detail.Further,we modify the EWMA control chart,and compare the modified EWMA control chart with the conventional EWMA control chart by the numerical simulation method.At the same time,we study the application of stock market,and discuss the effect of different control charts to monitor the same stock data.The modified EWMA control chart can detect shifts earlier and more efficient.Secondly,based on the previous study,we use the control chart to analyze the trading decision of one stock.The stock day closing prices of E-Jiao are used for empirical analysis.There are significant correlation between data and variance change over time.These characteristics can be eliminated after we fit the appropriate time series model.We monitor the residuals to analyze the best trading time of the stock on line.From the above two problems,we know that the sensitivity of the control chart will reduce when the process data are autocorrelation and heteroscedasticity.By choosing an appropriate model to fit these data,we can get more effective monitoring effect by modify the control limits and monitor the residuals of the model.