整合隱藏式馬可夫模型與演化策略於股市交易策略最佳化

Integrated Hidden Markov Model and Evolution Strategies to Optimize Stock Trading Strategies

10.6846/TKU.2017.00825

王聖方

隱藏式馬可夫模型 ； 演化策略 ； 選股 ； 擇時 ； 資金配置 ； Hidden Markov Model ； Evolution Strategies ； Selection ； Timing ； Capital Allocation

淡江大學資訊管理學系碩士班學位論文

2017年

碩士

張應華

繁體中文

銀行低利率時代來臨，日本甚至已進入負利率的時代，因此愈來愈多人另尋其他管道投資。其中股票是投資者熟悉的一種投資理財工具，舉凡電視、報章雜誌和網際網路，皆有不少人在熱烈討論。一般投資者常根據技術指標來進行股票擇時的操作，如利用KD值、MA指標、RSI…等來決定股票的買賣時機，但經常出現各指標間互相矛盾的買賣訊號，或是預測訊號與股票實際漲跌狀況差異甚大。 本研究整合隱藏式馬可夫模型與演化策略，建立一股市投資決策最佳化系統。利用隱藏式馬可夫模型，透過觀測值的變化來預測股票漲跌狀態。使用風險每變動一單位，投資人獲利程度的累積變化做為進場時機的依據，亦即做為何時該重新調整投資人手上投資組合的判斷，買賣交易策略則透過三個常用的技術指標：KD、MA和RSI來預測股票是否該買或該賣。結合進場訊號、買賣訊號與演化策略，共同演化出最佳資金配置與選股策略以幫助投資人做出正確的決策，以獲取優渥的超額報酬。

The low interest rate has come, and Japan has even entered the era of negative interest rates, so there is getting more people looking for other ways to invest. One of investment and financial instruments is stocks, there are many discussions from TV to internet. General investors use technical indicator to make stock timing decision, for example: KD, MA, and RSI. However, the signals from technical indicator are different from indicator to indicator. This paper integrated Hidden Markov Model and Evolution Strategies to build a stock investment strategies system. By Hidden Markov Model, it can observe the observation sequence to forecast the stock will rise or fall, which is the hidden state. In the timing part, this paper use cumulative “changes in return of per change in risk” to forecast when to adjust the portfolio. This paper use three common technical indicator, KD, MA, and RSI to make the trading decision. Finally, combing the signal of adjusting portfolio, the signal of trading signal and genetic algorithm, the best capital allocation will be evolved, and that is helpful to investors make correct decision.

1. [3]周照偉, 鄭榮祿, 蔡賢亮, 楊崇宏, & 牟聖遠. (2015). 臺灣股市技術分析實證: 以隨機指標, 相對強弱指標, 指數平滑異同平均線指標及趨向指標為例. 高雄應用科技大學人文與社會科學學刊, 1(2), 119-133.
   連結：
2. [4]夏承億，民104，以共演化式遺傳演算法輔助動態股票投資決策分析，淡江大學資訊管理學系碩士班碩士論文。
   連結：
3. [5]許溪南, 何怡滿, & 張瓊如. (2012). KD 與 MA 技術指標在避險時機之應用: 以台指選擇權為例. 輔仁管理評論, 19(1), 27-46.
   連結：
4. [6]陳淑玲, 吳安琪, & 費業勳. (2011). 臺灣股票市場技術指標之研究─ 不同頻率資料績效比較. 東海管理評論, 12(1S), 187-225.
   連結：
5. [7]Ahrari, A., & Deb, K. (2016). An improved fully stressed design evolution strategy for layout optimization of truss structures. Computers & Structures, 164, 127-144.
   連結：
6. [8]Beyer, H. G., Finck, S., & Breuer, T. (2014). Evolution on trees: On the design of an evolution strategy for scenario-based multi-period portfolio optimization under transaction costs. Swarm and Evolutionary Computation, 17, 74-87.
   連結：
7. [9]Cai, Q., Zhang, D., Wu, B., & Leung, S. C. (2013). A novel stock forecasting model based on fuzzy time series and genetic algorithm. Procedia Computer Science, 18, 1155-1162.
   連結：
8. [10]Chang, Y. H., & Lee, M. S. (2016). Incorporating Markov decision process on genetic algorithms to formulate trading strategies for stock markets. Applied Soft Computing.
   連結：
9. [11]Chang, Y. H., & Wang, S. C. (2013). Integration of Evolutionary Computing and Equity Valuation Models to Forecast Stock Values Based on Data Mining. Asia Pacific Management Review, 18(1), 63-78.
   連結：
10. [12]Deb, K. (2005). Multi-objective optimization using evolutionary algorithms. Wiley.
    連結：
11. [13]Duda, R. O., Hart, P. E., & Stork, D. G. (2012). Pattern classification. John Wiley & Sons.
    連結：
12. [15]Gupta, A., & Dhingra, B. (2012, March). Stock market prediction using hidden Markov models. In Engineering and Systems (SCES), 2012 Students Conference on (pp. 1-4). IEEE.
    連結：
13. [16]Gorgulho, A., Neves, R., & Horta, N. (2011). Applying a GA kernel on optimizing technical analysis rules for stock picking and portfolio composition. Expert systems with Applications, 38(11), 14072-14085.
    連結：
14. [18]Hassan, M. R., & Nath, B. (2005, September). Stock market forecasting using hidden Markov model: a new approach. In Intelligent Systems Design and Applications, 2005. ISDA'05. Proceedings. 5th International Conference on (pp. 192-196). IEEE.
    連結：
15. [19]Hassan, M. R., Nath, B., & Kirley, M. (2007). A fusion model of HMM, ANN and GA for stock market forecasting. Expert Systems with Applications, 33(1), 171-180.
    連結：
16. [21]Jones, G. (1998). Genetic and evolutionary algorithms. Encyclopedia of Computational Chemistry, 2, 1127-1136.
    連結：
17. [23]Lin, X., Yang, Z., & Song, Y. (2011). Intelligent stock trading system based on improved technical analysis and Echo State Network. Expert systems with Applications, 38(9), 11347-11354.
    連結：
18. [24]Lin, S. K., Wang, S. Y., & Tsai, P. L. (2009). Application of hidden Markov switching moving average model in the stock markets: Theory and empirical evidence. International Review of Economics & Finance, 18(2), 306-317.
    連結：
19. [25]Nguyen, N., & Nguyen, D. (2015). Hidden Markov Model for Stock Selection. Risks, 3(4), 455-473.
    連結：
20. [26]Papailias, F., & Thomakos, D. D. (2015). An improved moving average technical trading rule. Physica A: Statistical Mechanics and its Applications, 428, 458-469.
    連結：
21. [28]Sahin, U., & Ozbayoglu, A. M. (2014). TN-RSI: Trend-normalized RSI indicator for stock trading systems with evolutionary computation. Procedia Computer Science, 36, 240-245.
    連結：
22. [30]Tanaka-Yamawaki, M., & Tokuoka, S. (2007). Adaptive use of technical indicators for the prediction of intra-day stock prices. Physica A: Statistical Mechanics and its Applications, 383(1), 125-133.
    連結：
23. [33]Zawidzki, M. (2016). Optimization of Multi-branch Truss-Z based on evolution strategy. Advances in Engineering Software, 100, 113-125.
    連結：
24. [34]Zhu, D. M., Xie, Y., Ching, W. K., & Siu, T. K. (2016). Optimal portfolios with maximum Value-at-Risk constraint under a hidden Markovian regime-switching model. Automatica, 74, 194-205.
    連結：
25. 中文文獻
26. [1]吳韻玲，民98，台灣房地產景氣領先指標預測力研究─HMM模型之運用，國立屏東商業技術學院不動產經營所，碩士論文。
27. [2]杜金龍，民99。最新技術指標—在臺灣股市運用的訣竅（增訂三版）。臺北：財訊出版。
28. 英文文獻
29. [14]Farshchian, M., & Jahan, M. V. (2015, November). Stock market prediction with Hidden Markov Model. In Technology, Communication and Knowledge (ICTCK), 2015 International Congress on (pp. 473-477). IEEE.
30. [17]Granville, J. E. (1963). New Key to Stock Market Profits. Prentice-Hall Inc., N.J.
31. [20]Ingle, V., & Deshmukh, S. (2016, August). Hidden Markov Model Implementation for Prediction of Stock Prices with TF-IDF features. In Proceedings of the International Conference on Advances in Information Communication Technology & Computing (p. 9). ACM.
32. [22]Lane GC. (1984). Lane’s Stochastics. Technical Analysis of Stocks and Commodities magazine 1984;2:87-90.
33. [27]Runarsson,T.,Sarker,R.,Jonsson, M.T.(2000).Constrained nonlinear integer programming, self-adaptation and evolution strategies.International Journal of Knowledge-Based Intelligent Engineering Systems,4,164-171.
34. [29]Somani, P., Talele, S., & Sawant, S. (2014, December). Stock market prediction using Hidden Markov Model. In Information Technology and Artificial Intelligence Conference (ITAIC), 2014 IEEE 7th Joint International (pp. 89-92). IEEE.
35. [31]Wilder Jr, J. W. (1986). The Relative Strength Index. Journal of Technical Analysis of Stocks and Commodities, 4, 343-346.
36. [32]Yeh, C. H. (2012). The profitability of moving average in Taiwan: A New Anomaly. International Journal of Business and Social Science, 3(20).