| 摘要: | 證券投資買賣一直以來不論在學術界、業界或是一般大眾，都是會特別關注的議題，這幾年來更隨著FinTech的崛起，投資管理層面議題更顯重要。目前選股或選時相關研究較多針對某一層面進行探討，比較少有完整的投資組合與獲利風險評估方式，進行有系統或全面性討論實際交易所需要思考的各個層面，或是探討程式交易、投資組合與風險評估結合後的運作表現。本研究結合投資流程下各個部份，建立一套投資決策流程模型，並探討交易策略、部位大小和風格投資結合後績效表現，最後提供一個自動化報酬與風險評估平台，做為投資者在投資前買賣上參考。
本系統是以MultiChars、Python、Microsoft SQL Server、Django和Plotly進行實作，當中包含三個子系統，分別為程式交易策略最佳化子系統（最佳化技術面策略的參數）、投資組合管理子系統（建立投組分類）、獲利與風險評估子系統（提供投組績效結果）。
本研究在台股市場中選出具代表性且包含一定商品數量的中型100來驗證，交易策略選用能抓住長期發展趨勢的通道策略與均線策略，投組建立方式分別採用風格投資法下的價值面、規模面、動量面和波動面四種方式，進行分類，各別按照數值大小在當中平均分為五組，運用各種不同部位大小比例進行投資，最後在上述基礎下，比較選股選時（基本面加技術面）與選股（單純採用基本面）兩種投資方式之間績效成效。
根據實驗結果，投資方式採用選股選時不論是在哪一種類型風格投資法與哪一種類型的交易策略下，績效表現都優於單純採用選股來進行投資，整體報酬較優，而投資期間承擔風險相對較低，由此可證在選股選時模型結合後，進行投資交易與管理的可行性。
;Investment securities are always main topics in academic circle, industry or public. The issue of investment management is more and more important since FinTech has been discussed over the past few years. Most study in stock selection and market timing only surveyed in specific part, there was not only a few systematic and overall discussion in portfolio investment or profit and risk evaluation, but also a few study discussed the performance of combining program trading, portfolio investment and risk evaluation together. In this study, there are three research purposes. First, combining each process in investment securities to build an investment strategy process model. Second, discussing the performance of combining trading strategy, position sizing and style investment. Finally, providing an automatic return and risk evaluation platform for investors to use.
This study uses MultiCharts, Python, Microsoft SQL Server, Django and Plotly to implement this system. It includes three sub system, program trading strategy optimization sub system (optimizes the parameters of the technical analysis strategies), portfolio investment management sub system (creates the classification of portfolio investment) and profit and risk evaluation sub system (provides the performance of portfolio investment).
Mid-Cap 100 ETF in Taiwan Stock is representative and has enough number of stocks to verify this system. Channel strategy and moving average strategy are trend indicators in technical analysis, this study chooses them as experiment. Portfolio investment uses value, size, momentum and volatility effect in style investment. Portfolio in each effect is equally classified into five categories and invests different ratio of position sizing. Based on the fundamentals above, this study compares portfolio performance between the combination of stock selection and market timing (fundamental and technical analysis) and stock selection (fundamental analysis).
In this experiment, no matter which style investment and trading strategy, the performance of combination of stock selection and market timing is better than stock selection, the rate of return is higher and the risk is lower. This result proves that the combination of stock selection and market timing is feasible to use in investment trading and management. |
