Canonical Correlation Analysis About The Rise And Drop Of Stock And Its Impacted Factors

Canonical Correlation Analysis is the further development of the Principal Component Analysis and Factor analysis , Which is to study the interdependence Between the two groups' variables and translate the interdependent of two variables into studying the new interdependence of them. But do not discard the original variable message, These two new variables are from the linear combination of the first group variable and the second group, and both the number and content of the two variables may be different, so, Canonical Correlation Analysis has wide application in practical problems.From long term, the rise and drop of each stock is due to the common effect of some factors. The aim of this thesis is to build up a Canonical Correlation Analysis model about the rise and drop of stock and its impacted factors, ration how the change of each factor impact the rise and drop of stock, give some authentic analysis to the calculating result, and find out the main factors impacting the rise and drop of stock so as to direct how to invest in stock.With the help of stock analysis software, get some basic data such as opening price, closing price, the highest price, the lowest price, etc.; to get quantities of deriving data of stock by dealing with these data with the help of VF language program, such as the length of the upper and lower shadow, etc; to compare these data with the rise and drop of stock and have a correlation analysis with the help of SPSS software; and deal with some special data with Canonical Correlation Analysis methods in order to analyze the factors concerning the rise and drop of stock..This thesis has a tentative analysis about some factors affecting the rise and drop of stock, using the Bank of China (601988) as an example. This thesis uses a 100 days' average extent of daily rise and drop, rising extent, falling extent as variation of dependent variable unit, and uses the following as variation of independent variable unit: the extent of daily rise and drop of Shanhai Stock Index, the closing price of the day before, the daily transaction amount of stock, the closing price minus the average cost of stock, the length of the upper shadow of the day K line, the length of the lower shadow of the day K line, the high opening rate and the low opening rate of stock, etc. This thesis gives correlation analysis about the data of 270 days' trade from July 5, 2006 to July 13, 2007, and has tentative discussion about the result. The result is: Canonical Correlation Analysis method can reveal the relative relation between the rise and drop of stock and its relating factors so that this method can help direct stock invest.There are four parts in this thesis: the first chapter introduces the background, purpose, researching methods and main conclusion; the second chapter is about the basic thinking, steps and calculation of the Canonical Correlation Analysis; the third chapter is about the choosing of independent variable unit and dependent variable unit; and the last chapter gives out some authentic proofs about Canonical Correlation Analysis. At last, the thesis offers some advice to stock investors.