運用市場輪廓物理力量於金融市場交易行為之發現

Abstract

金融市場是充滿變化的動態環境，市場價格之變化是由於市場參與者之交易行為而變動，認為市場具有其行為規則。本研究提出一個創新的方式來推斷市場邏輯和市場變化之知識規則。研究採用倒傳遞類神經網路和灰色系統方法來計算隨機指標(KD)、指數平滑異同移動平均線(MACD)、資金流量指標(MFI)、價值區間擺動因子(VARF)及定量市場輪廓指標，以期許透過物理力量的總合評判，評估影響台灣加權指數期貨市場其市場邏輯與市場結構變化之知識規則。實驗結果顯示，透過此研究模型衡量市場長線動量的趨勢變化，有效提升預測模型之準確率及獲利能力，顯示市場價值動能趨勢對於當日短線之價格變化有其重要性。此外，將本研究之實驗組與對照組的隨機交易進行比較，發現實驗組更能從市場中獲得利潤，亦表示金融市場並非完成符合隨機漫步理論。

The financial market is a dynamic environment full of rapid changes. Changes in market prices occur primarily due to the trading behavior of market participants. However, the participants must follow established market rules. This study proposes a novel approach to extrapolate market logic and knowledge rules. This study applies the back-propagation neural network (BPNN) and grey system methods to compute stochastic (KD), moving average convergence-divergence (MACD), money flow index (MFI), value area rotation factor (VARF) and quantitative market profile data to extrapolate the market logic and knowledge rules that influence the Taiwan capitalization weighted stock index (TAIEX) futures market structure via an integral physical quantities assessment. The experimental results show that using the proposed model to measure the momentum of value area over the past few days improves the prediction model accuracy and profit. This implies that the momentum trend over the past few days is important for predicting the price direction in one day. This study compares the experimental group with random trading and finds the proposed model obtains more profit than random trading. Therefore, financial market trading patterns are not random.