標題: 應用滾動加速修正偏誤複式區間決定台灣股票之適當買賣價格
Deciding the Proper Trading Price for Taiwan Stocks using Rolling-Bias Corrected and Accelerated Bootstrap Confidence Intervals
作者: 陳珈瑋
Chen, Chia-Wei
唐麗英
Tong, Lee-Ing
工業工程與管理系所
關鍵字: 複式模擬法;複式信賴區間;股票投資;Rolling;BCa Bootstrap;Confidence intervals;Share price estimate
公開日期: 2013
摘要: 股票買賣是一般民眾常使用的投資方式,由於股票市場資訊的取得日漸便利,資訊太多有時會讓投資者難以整理,不曉得合理之股票買價或是賣價,且成功的股票投資人如巴菲特、葛拉漢等人對股票都有各自的看法,有時會造成投資者的困擾。現有股票投資之相關文獻,大多是探討股票的買賣時機與技術的分析,少見有對股價設定買與賣價格的研究。股票之高點通常沒辦法維持長久,因此若能找出適當的股票買賣價格對投資者會有非常大的幫助。因此本研究主要目的是找出各股票適當的買賣價格。本研究首先針對財務指標(如:毛利率、本益比及股東權益報酬率等)選出具投資價值的股票,再使用加速修正偏誤複式區間估計方法來估計各股票的股價,然後以滾動的方式淘汰過去的資料,計算出適當的股票買賣價格,最後,以不同信賴度之信賴區間檢視投資報酬率、次數、成功率的變化。本研究篩選出十六支具投資價值得台灣股票,經過本研究之分析後,證實本研究之方法確實可有效降低股票投資風險及提高投資報酬率。
Stock market is the most common investment tool in the present time. Although it is easy to get the information of any share, the information sometimes is too much to determine the plausible bid price or selling price for stock investor. Some successful stock investors such as Warren Buffett, Graham, have their own way to make the decisions. Investors can not adopt all of their suggestions. Therefore, the objective of this study is to determine optimal investing share prices for Taiwanese stocks. First, the financial indicators (e.g. gross margin ratio, return on equity etc.) are used to pick up the potential shares, to avoid investment on the financial crisis company. Then, the rolling-bias corrected and accelerated (BCa) bootstrap confidence intervals are employed to estimate the bid price and selling price for these stocks. Finally, the sensitivity analysis of different (1-α)100% confidence intervals is conducted by varying the return on investment (ROI) and trading frequency. Sixteen Taiwan Stocks are selected to demonstrate the effectiveness of the proposed method. The analytical results show that the proposed method can effectively reduce the investment risk and promote the ROI.
URI: http://140.113.39.130/cdrfb3/record/nctu/#GT070153314
http://hdl.handle.net/11536/73730
Appears in Collections:Thesis