基于深度学习和进化计算的外汇预测与投资组合优化
|
摘要
利用深度学习和进化计算技术来分别实现对外汇价格的预测与投资组合优化.首先,利用循环神经网络建立汇率预测模型,用来预测外汇产品的价格并计算期望收益率.接着建立了一个双目标的投资组合模型,即最大化期望收益率与最小化风险.为了更接近真实的外汇交易市场,该模型中允许买空与卖空,并考虑了点差对收益的影响.基于多个外汇产品的期望收益率与投资组合模型,利用多目标进化算法来搜索出最优的投资组合.在多个外汇产品的真实历史数据上的结果表明,该方法能够实现在外汇交易市场中的盈利.
The techniques in deep learning and evolutionary computation were adopted to forecast the exchange rate and to optimize the portfolio respectively.Firstly,recurrent neural network is used to build an exchange rate forecasting model,to forecastg the price of instrument and to calculate its expected yield.Then,a bi-objective portfolio model is built,i.e.,maximizing the expected yield and minimizing the risk.For approximating the real market,the proposed model could allow long and short selling,and could also consider the influence of spread.Based on the expected yields of multiple instruments and the proposed portfolio model,a multi-objective evolutionary algorithm was adopted to search for the optimal portfolio.According to the back test on the historical data of multiple instruments,it was verified that the proposed approach could make profit in the exchange market.
The techniques in deep learning and evolutionary computation were adopted to forecast the exchange rate and to optimize the portfolio respectively.Firstly,recurrent neural network is used to build an exchange rate forecasting model,to forecastg the price of instrument and to calculate its expected yield.Then,a bi-objective portfolio model is built,i.e.,maximizing the expected yield and minimizing the risk.For approximating the real market,the proposed model could allow long and short selling,and could also consider the influence of spread.Based on the expected yields of multiple instruments and the proposed portfolio model,a multi-objective evolutionary algorithm was adopted to search for the optimal portfolio.According to the back test on the historical data of multiple instruments,it was verified that the proposed approach could make profit in the exchange market.
引文
[1]STEPNICKA M,CORTEZ P,DONATE J P,et al.Forecasting seasonal time series with computational intelligence:on recent methods and the potential of their combinations[J].Expert systems with applications,2013,40(6):1981-1992.
[2]AGHION P,BACCHETTA P,RANCIRE R,et al.Exchange rate volatility and productivity growth:the role of financial development[J].Journal of monetary economics,2009,56(4):494-513.
[3]GALESHCHUK S.Neural networks performance in exchange rate prediction[J].Neurocomputing,2016,172:446-452.
[4]LIU L X,WANG W J.Exchange rates forecasting with least squares support vector machine[C]//Proceedings of the 2008 International Conference on Computer Science and Software Engineering.Hubei:IEEE,2008:1017-1019.
[5]KUAN C M,LIU T.Forecasting exchange rates using feedforward and recurrent neural networks[J].Journal of applied econometrics,1995,10(4):347-364.
[6]马骊,李阳,樊锁海.改进人工鱼群算法在外汇预测和投资组合中的应用[J].系统工程理论与实践,2015,35(5):1256-1266.
[7]INTHACHOT M,BOONJING V,INTAKOSUM S.Artificial neural network and genetic algorithm hybrid intelligence for predicting Thai stock price index trend[J].Computational intelligence and neuroscience,2016:3045254.
[8]DINH T T H,VU V T,BUI T L.A multi-objective ensemble learning approach based on the non-dominated sorting differential evolution for forecasting currency exchange rates[C]//Proceedings of the 2016Eighth International Conference on Knowledge and Systems Engineering.Hanoi:IEEE,2016:96-102.
[9]SKOLPADUNGKET P,DAHAL K,HARNPORNCHAIN.Portfolio optimization using multi-objective genetic algorithms[C]//Proceedings of the 2007 IEEE Congress on Evolutionary Computation.Singapore:IEEE,2007:516-523.
[10]BRANKE J,SCHECKENBACH B,STEIN M,et al.Portfolio optimization with an envelope-based multiobjective evolutionary algorithm[J].European journal of operational research,2009,199(3):684-693.
[11]李章晓,宋薇,张兴义.基于带决策者偏好多目标优化的证券组合投资研究[J].合肥师范学院学报,2015,33(3):37-40.
[12]CHO K,MERRINBOER B V,BAHDANAU D,et al.On the properties of neural machine translation:encoder-decoder approaches[J].ar Xiv preprint ar X-iv:1409.1259,2014.
[13]TIAN Y,CHENG R,ZHANG X,et al.An indicator based multi-objective evolutionary algorithm with reference point adaptation for better versatility[J].IEEEtransactions on evolutionary computation,2018,22(4):609-622.
[14]LECUN Y,BENGIO Y,HINTON G.Deep learning[J].Nature,2015,521:436-444.
[15]GRAVES A,MOHAMED A R,HINTON G.Speech recognition with deep recurrent neural networks[C]//Proceedings of the International Conference on Acoustics,Speech and Signal Processing.Vancouver:IEEE,2013:6645-6649.
[16]SUTSKEVER I,MARTENS J,HINTON G.Generating text with recurrent neural networks[C]//Proceedings of the 28th international conference on machine learning Bellevue,2011:1017-1024.
[17]GRAVES A,SCHMIDHUBER J.Framewise phoneme classification with bidirectional LSTM and other neural network architectures[J].Neural networks,2005,18(5):602-610.
[18]HOCHREITER S,SCHMIDHUBER J.Long short-term memory[J].Neural computation,1997,9(8):1735-1780.
[19]CHUNG J,GULCEHRE C,CHO K H,et al.Empirical evaluation of gated recurrent neural networks on sequence modeling[J].ar Xiv preprint ar Xiv:1412.3555,2014.
[20]ZHANG X,DUAN F,ZHANG L,et al.Pattern recommendation in task-oriented applications:A multiobjective perspective[J].IEEE computational intelligence magazine,2017,12(3):43-53.
[21]刘广瑞,王庆海,姚冬艳.基于改进人工蜂群算法的多无人机协同任务规划[J].郑州大学学报(工学版),2018,39(3):51-55.
[22]SCHAFFER J D.Multiple objective optimization with vector evaluated genetic algorithms[C]//Proceedings of the First International Conference on Genetic Algorithms.Hillsdale:Association computing Machinery,1985:93-100.
[23]MARKOWITZ H M.Portfolio selection[J].Journal of finance,1952,7:77-91.
[2]AGHION P,BACCHETTA P,RANCIRE R,et al.Exchange rate volatility and productivity growth:the role of financial development[J].Journal of monetary economics,2009,56(4):494-513.
[3]GALESHCHUK S.Neural networks performance in exchange rate prediction[J].Neurocomputing,2016,172:446-452.
[4]LIU L X,WANG W J.Exchange rates forecasting with least squares support vector machine[C]//Proceedings of the 2008 International Conference on Computer Science and Software Engineering.Hubei:IEEE,2008:1017-1019.
[5]KUAN C M,LIU T.Forecasting exchange rates using feedforward and recurrent neural networks[J].Journal of applied econometrics,1995,10(4):347-364.
[6]马骊,李阳,樊锁海.改进人工鱼群算法在外汇预测和投资组合中的应用[J].系统工程理论与实践,2015,35(5):1256-1266.
[7]INTHACHOT M,BOONJING V,INTAKOSUM S.Artificial neural network and genetic algorithm hybrid intelligence for predicting Thai stock price index trend[J].Computational intelligence and neuroscience,2016:3045254.
[8]DINH T T H,VU V T,BUI T L.A multi-objective ensemble learning approach based on the non-dominated sorting differential evolution for forecasting currency exchange rates[C]//Proceedings of the 2016Eighth International Conference on Knowledge and Systems Engineering.Hanoi:IEEE,2016:96-102.
[9]SKOLPADUNGKET P,DAHAL K,HARNPORNCHAIN.Portfolio optimization using multi-objective genetic algorithms[C]//Proceedings of the 2007 IEEE Congress on Evolutionary Computation.Singapore:IEEE,2007:516-523.
[10]BRANKE J,SCHECKENBACH B,STEIN M,et al.Portfolio optimization with an envelope-based multiobjective evolutionary algorithm[J].European journal of operational research,2009,199(3):684-693.
[11]李章晓,宋薇,张兴义.基于带决策者偏好多目标优化的证券组合投资研究[J].合肥师范学院学报,2015,33(3):37-40.
[12]CHO K,MERRINBOER B V,BAHDANAU D,et al.On the properties of neural machine translation:encoder-decoder approaches[J].ar Xiv preprint ar X-iv:1409.1259,2014.
[13]TIAN Y,CHENG R,ZHANG X,et al.An indicator based multi-objective evolutionary algorithm with reference point adaptation for better versatility[J].IEEEtransactions on evolutionary computation,2018,22(4):609-622.
[14]LECUN Y,BENGIO Y,HINTON G.Deep learning[J].Nature,2015,521:436-444.
[15]GRAVES A,MOHAMED A R,HINTON G.Speech recognition with deep recurrent neural networks[C]//Proceedings of the International Conference on Acoustics,Speech and Signal Processing.Vancouver:IEEE,2013:6645-6649.
[16]SUTSKEVER I,MARTENS J,HINTON G.Generating text with recurrent neural networks[C]//Proceedings of the 28th international conference on machine learning Bellevue,2011:1017-1024.
[17]GRAVES A,SCHMIDHUBER J.Framewise phoneme classification with bidirectional LSTM and other neural network architectures[J].Neural networks,2005,18(5):602-610.
[18]HOCHREITER S,SCHMIDHUBER J.Long short-term memory[J].Neural computation,1997,9(8):1735-1780.
[19]CHUNG J,GULCEHRE C,CHO K H,et al.Empirical evaluation of gated recurrent neural networks on sequence modeling[J].ar Xiv preprint ar Xiv:1412.3555,2014.
[20]ZHANG X,DUAN F,ZHANG L,et al.Pattern recommendation in task-oriented applications:A multiobjective perspective[J].IEEE computational intelligence magazine,2017,12(3):43-53.
[21]刘广瑞,王庆海,姚冬艳.基于改进人工蜂群算法的多无人机协同任务规划[J].郑州大学学报(工学版),2018,39(3):51-55.
[22]SCHAFFER J D.Multiple objective optimization with vector evaluated genetic algorithms[C]//Proceedings of the First International Conference on Genetic Algorithms.Hillsdale:Association computing Machinery,1985:93-100.
[23]MARKOWITZ H M.Portfolio selection[J].Journal of finance,1952,7:77-91.