股票的買賣在台灣是非常盛行的,經由技術分析來選股就顯得非常的重要。技術分析中最常見的移動平均線(MA),往往會受到極端值的影響而產生偏差,尤其台股市場是非常容易受到外在因素而造成巨大的變化,故如何找尋一較具穩健性的技術性分析工具便成為本文的研究重心之一。 變異係數在統計分析中常用以表示資料的分散程度,在財經分析中則常用以代表投資工具的風險大小。但是,變異係數是截斷式資料(cross section data)的分析工具,而股價卻屬時間序列資料(time series data),如何將變異係數的概念應用於時間序列的動態資料分析,並藉此來衡量股價的動態風險變化以利投資人選擇合意投資標的,逐成本文的另一研究重心。 此外,「選擇」「好的」投資標的時,不論是「選擇」或是「好的」,都會因人的主觀意識而有所不同。而模糊理論常用以解決因人類主觀意識的不同而產生的爭議問題。故在本文,我們也將模糊理論納入分析方法之一。 在本文中,我們定義了移動變異係數(moving coefficient of variation),並以其大小來表示股票的動態風險;同時,我們結合了模糊評判方法,建立一個價量綜合模糊評判法則來幫助我們篩選股票。最後,我們以台灣五家上市公司(奇美、友達、仁寶、台積電、統一)近三年的股票價量資料,來驗證我們方法的實用性。
Stock trading is very popular in Taiwan and it seems to be very important to choose a stock by technical analysis. The commonly seen, Moving Average (MA) in technical analysis is often affected by external factors to a huge change. How to pick a tool of steadily technical analysis becomes one of the researchful topics of this article. The coefficient of variation is often used to indicate the dispersion of data and represented to the risk degree of investment tools in financial analysis. The coefficient of variation is the tool of analysis for cross section data, but the price of stock belongs to time series data. How to apply the coefficient of variation’s concept to dynamic data analysis of time series and measure the dynamic risk change of stock price for investors to choose a satisfied investment target is getting another researchful topic of this article. Therefore, when choosing a good investment target, it would be different because of people’s subjective consciousness no matter “choosing” or “goodness”. And the fuzzy theorem is used to solve the disputable problem of different for human’s subjective consciousness, so we should bring the fuzzy theorem into one of the analysis section in this article. In this article, we define the moving coefficient variation and use it to represent the dynamic risk change of stocks. At the same time, we combine the way of fuzzy judgments and establish a comprehensive index of fuzzy judgment based on price and turnover volume of stocks to help us to choose stocks. Finally, we used data of stock price and turnover volume of five companies in Taiwan(CMO, AUO, Compal Electronics, TSMC and President Chain Store Corporation) in current three years to verify the the practicability of our approach