An Investing Decision Support System Based on Double-Loop Learning Concept

[[abstract]]臺灣股市的散戶投資人比例甚高，但散戶在專業不足與資訊不對稱的情況下， 投資績效往往不如預期，實務上雖有一些投資輔助軟體與網站可以支援投資人的需 求，其支援層面多以提供市場與技術分析資訊為主，很少有針對提升投資人決策模 式提供支援。事實上金融市場的演變與眾多的國內外經濟與產業因素息息相關，是 一個極端複雜的體系，要改善投資人對此複雜系統的了解，必須使用更有效的方法。 本研究基於雙環路學習的概念，建構一平台讓投資人可以運用決策模擬法達到改善 心智模式的效果。使用者可以在此虛擬環境中，將各種設想的投資策略透過歷史交 易資料測試其投資績效，藉由此平台模擬測試結果將投資報酬率、進出場時間點以 及判斷買賣標準的技術指標等，經由圖報表的呈現回饋給使用者，讓使用者可以運 用所呈現的模擬結果資訊來重新修正自身投資策略與規則，進而思考其投資理念與 觀點是否有缺失或不足的地方，來達到雙環路學習的成效。[[abstract]]Individual investors occupied high proportions in Taiwanese financial market. Individual investors are generally inferior in information and analytical proficiency. Although there are some software and websites providing information to support to investors, existing support on investors’ decision models are rare in the practice. Financial market is an complex and dynamic system where numerous factors are jointly intertwined. It requires more effective approaches to improve investors’ mental models on investment decisions. This study proposed a new decision support approach by providing a double-loop learning mechanism to decision makers. Based on the proposed concepts, this study constructed a platform through which investors can test their strategies and evaluate the performances in simulations under historical market data. Users can learn from the information their mistakes and adjust their mental models accordingly. The information feedback about the effectiveness of one’s mental model is essential to achieve the double-loop learning and thus improve their mental models.[[note]]碩