臺灣博碩士論文加值系統

本論文提出一個改良的智慧型高頻交易預測模型的比較性研究，我們整合運用遺傳演算法的新穎適應性函數的設計及新的預測規則而建構一個較有效的使用高頻交易資料而實現的股價預測系統。透過台灣股市70個交易日盤中即時揭露資訊的資料，並分成兩個週期進行交互驗證，實驗結果顯示我們提出的改良的方法能夠超越之前Huang及Li的方法所產生的最佳預測結果；此外，我們的方法也能夠在減少些許預測能力的狀況下而大幅降低預測結果的標準差，因此顯示該方法能夠提供較穩定的預測效果。這些結果顯示本文所提的方法在使用高頻交易的資料上可改良先前的運用遺傳演算法所建構的財務預測模型的強健性及預測能力，我們希望這些研究結果能拓展股市投資於高頻交易領域的相關研究。

In this thesis, we present a comparative study of our proposed financial prediction models with other methods previously reported for high frequency trading (HFT). Using several alternative designs of the fitness functions, along with an additional prediction rule we propose, we employ the genetic algorithms (GAs) to construct a new version of the prediction models for stock price in the context of high frequency environment. We use seventy days of high frequency price data of several stocks, which comprised two periods of temporal validation for the evaluation of our models. The experimental results show that our proposed models can improve not only the effectiveness, but also as the robustness of the prediction for stock price in HFT. We therefore expect that these results are able to enrich this line of research and real-world applications for high frequency trading and computational finance.

摘要 II
致謝 IV
圖目錄 VII
表目錄 IX
1. 導論 1
1.1 研究背景 1
1.2 研究目的 2
1.3 研究架構圖 3
2. 文獻探討 4
2.1 高頻交易相關文獻 4
2.2股票價量關係 5
2.3 遺傳演算法及其運用於股市預測相關文獻 6
3. 研究方法 9
3.1 效能指標 9
3.1.1評估指標 13
3.1.2 適應性函數演化指標 14
3.2台灣股市交易機制之集合競價 15
3.2.1集合競價交易範例 16
3.3進場訊號及價格預測 21
3.3.1成交價與最佳買價及最佳賣價之法則 21
3.3.2 最佳五檔委託量權重法則 23
3.3.3最佳一檔委買量、委賣量壓力法則 24
3.4 遺傳演算法 27
3.4.1 基因編碼 28
3.4.2 解碼 29
3.4.3 選擇及淘汰方法 29
3.4.4 交配 29
3.4.5 突變 31
4. 研究結果 32
4.1 資料來源 32
4.2 研究區間 32
4.3 時間驗證(Temporal Validation) 33
4.4 演算法演化設定 34
4.5比較基準 35
4.6實驗結果 35
4.6.1 Model1於2015/09/15~2015/11/11期間的表現 36
4.6.2 Model2於2015/09/15~2015/11/11期間的表現 41
4.6.3 Model1於2015/11/18~2016/01/12期間的表現 46
4.6.4 Model2於2015/11/18~2016/01/12期間的表現 51
4.6.5 效能平均值與其標準誤差 56
5. 結論 59
6. 參考文獻 60

[1].佘珮琦，2007，〈我國證券市場交易制度介紹〉，《證券暨期貨月刊》，第二十五卷，第四期，5-17頁。
[2]. 蔡佩雯，2013，〈逐筆交易及縮短撮合循環秒數概述〉，《證券暨期貨月刊》，第三十二卷，第十二期，14-20頁。
[3].Huang, C.-F. and Li, H.-C. (2017). “An Evolutionary Method for Financial Forecasting in Microscopic High-Speed Trading Environment,” Computational Intelligence and Neuroscience, Volume 2017 (2017), Article ID 9580815, 18 pages.
[4].Zhang, M., Hanna, M., Li, J., Butcher, S., Dai, H. and Xiao, W. (2010). “Creation of a Hyperpermeable Yeast Strain to Genotoxic Agents Through Combined Inactivation of PDR and CWP Genes,” Toxicological Sciences. 113(2), 401-11.
[5].Jarnecic, E. and Snape, M. (2010). “An Analysis of Trades by High Frequency Participants on the London Stock Exchange,” Working Paper.
[6].Jovanovic, B. and Menkveld, A.J. (2010). “Middlemen in Limit-order Markets,” Working Paper.
[7].Frino, A., Lepone, A. and Mistry, M. (2011). “The New Breed of Market Participants: High Frequency Trading: Evidence from the Australian Stock Exchange,” Working Paper.
[8].Jarrow, Robert A. and Protter, Philip. (2011). “A Dysfunctional Role of High Frequency Trading in Electronic Markets,” Johnson School Research Paper Series No. 08-2011.
[9].Ying, C. C. (1966). “Stock Market Prices and Volumes of Sales,” Econometrica, 34, 676-686.
[10].Karpoff, J. M. (1987). “The Relation Between Price Changes and Trading Volume: A Survey,” Journal of Financial and Quantitative Analysis, 22, 109-126.
[11].Saatcioglu, K. and Starks, L. T. (1998). “The Stock Price-volume Relationship in Emerging Stock Markets: The Case of Latin America,” International Journal of Forecasting, 14(2), 215-225.
[12].Gervais S., Kaniel R. and Mingelgrin, D. H. (2001). “The High-Volume Return Premium,” The Journal of Finance, 56(3), 877-919.
[13].莊家彰、管中閔(2005)。〈台灣與美國股市價量關係的分量回歸分析〉。《經濟論文》, 33,4, 379-404.
[14]. 高士軒(2008)。《價量關係：量是否為價格發現的先行指標》。碩士論文，逢甲大學，台中市。
[15]. 李雪. (2010)。〈牛市與熊市中股票價格和成交量互動關係對比分析-基於對上證綜指的實證研究〉。《江蘇商論》(14), 30-31.
[16]. Allen, F. and Karjalainen, R. (1999). “Using Genetic Algorithms to Find Technical Trading Rules,” Journal of Financial Economics, 51, 1999, 245-271.
[17]. Lin, W., Ming, D. and Phua, H. P. K. (2001). “Neural Network with Genetically Evolution Algorithms for Stocks Prediction,” Asia-Pacific Journal of Operation Research, 18(1), 103-108.
[18]. Korczak, J. and Roger, P (2002). “Stock Timing Using Genetic Algorithms,” Applied Stochastic Models in Business and Industry, 18, 121-134.
[19]. Jiang, R. and Szeto, K. Y. (2003). “Extraction of Investment based on Moving Average: A Genetic Algorithm Approach,” Computational Intelligence for Financial Engineering, Hong Kong, 403-410.
[20]. 陶宏德(1995)。《應用遺傳演算法達成知識之最適 化— 以臺灣股市技術分析指標為例》，交通大學資訊管理研究所碩士論文。
[21]. 邱昭彰、李安邦(1998)。〈遺傳演算法在發展股市投資專家知識規則之研究〉，《資管評論》第八期，民國 87 年 12 月，頁 21-38。
[22]. 許智明(2003)。《運用遺傳演算法搜尋最佳化技術指標之台灣股市實證研究》，雲林科技大學資訊管理研究所，碩士論文。
[23]. 連立川、葉怡成(2007)。〈以遺傳神經網路建構台灣股市買賣決策系統之研究〉，《資訊管理學報》。
[24]. Huang, C.-F., Li, C.A., Chang, B.R. and Li, H.-C. (2015). “A Comparative Study of Intelligent Pairs-trading Models for Investment,” The Proceedings of 2015 International Conference on Innovation, Communication and Engineering (ICICE 2015), Paper ID 1894.
[25]. Huang, C.-F., Chang, C.-H., Kuo, L.-M., Lin, B.-H., Hsieh, T.-N. and Chang, B. R. (2012). “A Genetic-search Model for First-day Returns Using IPO Fundamentals,” International conference on machine learning and cybernetics (ICMLC), 5, 1662–1667.
[26]. Huang, C.-F. , Hsu, C.-J., Chen, C.-C., Chang, B.R. and Li, C.-A. (2015). “An Intelligent Model for Pairs Trading Using Genetic Algorithms,” Computational Intelligence and Neuroscience, Volume 2015 (2015), Article ID 939606, 10 pages.
[27]. Huang, C.-F., Hsieh, T.-N., Chang, B.-R. and Chang, C.-H. (2011). “A Comparative Study of Stock Scoring Using Regression and Genetic-based Linear Models,” Proceedings of 2011 IEEE International Conference on Granular Computing (GrC 2011), 268-273.
[28].台灣證券交易所。〈集中市場交易制度介紹〉。http://www.tse.com.tw/zh/page/trading/introduce.html#01
[29].詹場、胡星陽、陳建宏(2003)。〈台灣證券市場揭示狀態之資訊含量〉，《證券市場發展季刊》第十五卷，第四期：1–36頁。
[30]. 郭修旻、李秀雯(1999)。〈股票市場波動性與總體經濟波動性及市場交易量之關係─臺灣市場實證研究〉，《中國工商學報》 二十一期(1999年10月)，pp.249–265.
[31]. Holland, J. (1975), Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence, University of Michigan Press, Ann Arbor.
[32]. Goldberg, D. E. and Deb, K. (1991). “Comparative Analysis of Selection Schemes Used in Genetic Algorithms,” Foundation of Genetic Algorithms, 69-93, 1991.
[33].De Jong, K. A., Spears, W. M. and Gordon, D. A. (1993). “Using Genetic Algorithms for Concept Learning,” Machine Learning, 13, 161-188, 1993.
[34].Huang, C.-F. (2012). “A Hybrid Stock Selection Model Using Genetic Algorithms and Support Vector Regression,” Applied Soft Computing, 12(2), 807-818.
[35].張美燕(2006)。《股價升降級距與最佳五檔揭示對股市之影響》，中原大學企業管理研究所學位論文。
[36].陳筑音(2006)。《最佳五檔資訊內涵》，政治大學財務管理研究所碩士論文。