運用SVM於MACD技術指標結合市場輪廓於臺灣加權股價指數之行為發現

Abstract

於市場交易中，投資者期望能逢低買進、逢高賣出以賺取超額的利潤，於是許多投資者藉由MACD指標的交叉，做為價格反轉時進出場的交易訊號，但MACD指標卻在價格波動較低的盤整時，造成指標鈍化形成錯誤的交易訊號，損失了頻繁交易的手續費。除了避免盤整時造成指標鈍化的發生，市場交易者的行為以及角色更是重要的ㄧ個環節，猶如Steidlmayer於1984年提出市場輪廓理論，認為市場價格的變動，是市場中各種類型的參與者的交易行為所造成，市場並非隨機漫步，而是具有行為模式和規則可被解讀。 本研究使用支持向量機(Support Vector Machine)進行學習預測，對臺灣加權指數期貨嘗試發掘MACD交叉時，歷史波動率是否能避免盤整時造成的交易虧損，又如何利用市場輪廓理論來辨別交易市場中的角色。並且利用統計分析方式以及純粹使用MACD指標進行比較，SVM的預測是否優於其於兩者。研究結果顯示，運用SVM於MACD技術結合市場輪廓理論，更優於傳統統計分析以及純粹使用MACD指標的結果，更能為投資者建立一良好的投資模型。

In the markets, investors wants to be the bargain hunter to make excess profits. So investor use technical indicator MACD as a signal that when should buy and when should sell the stock. But MACD indicator has a problem that when the market is during the correction period, the indicator will lose the predict accuracy, so that, the investor will lose the cost of trading. Except avoiding correction period, how to recognize the role of the investor is an important factor, Steidlmayer put forward market profile theory. The theory is that changes in market prices are caused by various types of market participants’ trading behavior. Market does not follow random walk, but it fluctuates with the behavior patterns and rules. This research use machine learning of Support Vector Machine try to find out, when MACD crossover, if the historical volatility can avoid the loss of trading cost during correction period, and how to use market profile to recognize the role of participator in the market use the weighted price index of the Taiwan stock exchange(TAIEX). And compare SVM to statistical analysis and the model merely using the MACD indicator. Result of this research shows that the model combining MACD and market profile with SVM has better performance than the model of statistical and merely using the MACD indicator.