Portfolio Investment Algorithm Based on LSTM-ARIMA Prediction Model
DOI: 10.23977/autml.2022.030104 | Downloads: 20 | Views: 942
Author(s)
Xianzhuo Chen 1, Xi Wei 1, Zhitian Su 1
Affiliation(s)
1 Heilongjiang University, Harbin, 150006, Heilongjiang, China
Corresponding Author
Xianzhuo ChenABSTRACT
When investing in bitcoin and gold, we need to be backed by predictions of future prices, which only require predictions of price data as of today. First, the ARIMA model is used for data prediction, but this model has certain limitations when dealing with nonlinear data, so the prediction results are not ideal. The LSTM model is a recurrent neural network model that can handle nonlinear data. Therefore, in this article, I use a combination of linear and nonlinear prediction models, the LSTM-ARIMA hybrid model, to predict the price of bitcoin and gold. According to the prediction results, the prediction error of the LSTM-ARIMA hybrid model has better results. Our LSTM-ARIMA hybrid model is able to predict the price of gold and bitcoin in the next 10 days. It uses the mean of predicted prices as price expectations and the variance of predicted prices as investment risk. In addition, we consider the expected risk of the forecasting model as investment risk. To determine portfolio investment strategies, we maximize post-investment value, minimize risk, and make investment decisions every day as the next day's asset holding state. In order to solve the optimal portfolio investment strategy, we use the particle swarm algorithm to solve the problem, which can ensure that each step of the solution is a local optimal solution. Finally concluded that the initial $1000 investment can be worth $130948420399.063 as of September 10, 2021. Next, we analyze the sensitivity of the portfolio investment strategy model to gold and bitcoin transaction costs. Plot the cumulative returns when αgold=1%, αbitcoin is 2%, 2.5%, 3%, and αbitcoin=2%, and αgold is 1%, 1.5%, and 2%, respectively. By observing the cumulative return graph, it can be found that the model has greater feedback and higher sensitivity to changes in bitcoin transaction costs, and the increase in transaction costs has a greater impact on the final return. However, the model is less sensitive to the transaction cost of gold and has strong model stability. Finally, we can draw the following conclusions: 1. The LSTM-ARIMA hybrid model outperforms the traditional time series prediction model in correlation coefficient prediction. 2. The particle swarm algorithm can converge to a local minimum at each decision. 3. With the support of prediction model and particle swarm algorithm, each step of the optimization of the portfolio investment model can provide the optimal solution, thus ensuring the upward trend of the overall income. 4. Prove that the model provides the best portfolio strategy through economic indicators such as the Sharpe ratio.
KEYWORDS
Trading Strategy, LSTM-ARIMA Hybrid Model, Particle Swarm, Investment Risk, Multi-Objective ProgrammingCITE THIS PAPER
Xianzhuo Chen, Xi Wei, Zhitian Su, Portfolio Investment Algorithm Based on LSTM-ARIMA Prediction Model. Automation and Machine Learning (2022) Vol. 3: 22-32. DOI: http://dx.doi.org/10.23977/autml.2022.030104.
REFERENCES
[1] Ponomare V E, Oseledets I, Cichocki A. Using Reinforcement Learning in the Algorithmic Trading Problem. Journal of Economic Dynamics and Control.2020.
[2] Gang Chen, Xiaomei Guo. Research on Oversampling Algorithm for Imbalanced Datasets Based On ARIMA Model. Proceedings of the 33rd China control and decision making conference. 2021:441-446.
[3] Troiano L, Mejuto Villa E, Loia V. Replicating a Trading Strategy by Means of LSTM for Financial Industry Applications. IEEE Transactions on Industrial Informatics, 2018:3226-3234.
[4] Kamble V.B., Deshmukh S.N.. Comparision between Accuracy and MSE, RMSE by Using Proposed Method with Imputation Technique. Oriental journal of computer science and technology, 2017, 10(04).
[5] Markowitz H. Portfolio selection. Journal of Finance.1952(7):77-91.
[6] Cai Biao, Zhu Xinping, Qin Yangxin. Parameters optimization of hybrid strategy recommendation based on particle swarm algorithm. Expert Systems with Applications, 2021,168:
[7] Qianhan Zhang. The Use of Sharpe Ratio in Portfolio Optimization.Proceedings of 4th International Conference on Economics and Management, Education, Humanities and Social Sciences (EMEHSS 2020). 2020:224-229.
[8] Javier Estrada. Maximum Withdrawal Rates: A Novel and Useful Tool. Journal of Applied Corporate Finance, 2017, 29(4).
Downloads: | 1616 |
---|---|
Visits: | 67651 |
Sponsors, Associates, and Links
-
Power Systems Computation
-
Internet of Things (IoT) and Engineering Applications
-
Computing, Performance and Communication Systems
-
Journal of Artificial Intelligence Practice
-
Advances in Computer, Signals and Systems
-
Journal of Network Computing and Applications
-
Journal of Web Systems and Applications
-
Journal of Electrotechnology, Electrical Engineering and Management
-
Journal of Wireless Sensors and Sensor Networks
-
Journal of Image Processing Theory and Applications
-
Mobile Computing and Networking
-
Vehicle Power and Propulsion
-
Frontiers in Computer Vision and Pattern Recognition
-
Knowledge Discovery and Data Mining Letters
-
Big Data Analysis and Cloud Computing
-
Electrical Insulation and Dielectrics
-
Crypto and Information Security
-
Journal of Neural Information Processing
-
Collaborative and Social Computing
-
International Journal of Network and Communication Technology
-
File and Storage Technologies
-
Frontiers in Genetic and Evolutionary Computation
-
Optical Network Design and Modeling
-
Journal of Virtual Reality and Artificial Intelligence
-
Natural Language Processing and Speech Recognition
-
Journal of High-Voltage
-
Programming Languages and Operating Systems
-
Visual Communications and Image Processing
-
Journal of Systems Analysis and Integration
-
Knowledge Representation and Automated Reasoning
-
Review of Information Display Techniques
-
Data and Knowledge Engineering
-
Journal of Database Systems
-
Journal of Cluster and Grid Computing
-
Cloud and Service-Oriented Computing
-
Journal of Networking, Architecture and Storage
-
Journal of Software Engineering and Metrics
-
Visualization Techniques
-
Journal of Parallel and Distributed Processing
-
Journal of Modeling, Analysis and Simulation
-
Journal of Privacy, Trust and Security
-
Journal of Cognitive Informatics and Cognitive Computing
-
Lecture Notes on Wireless Networks and Communications
-
International Journal of Computer and Communications Security
-
Journal of Multimedia Techniques
-
Computational Linguistics Letters
-
Journal of Computer Architecture and Design
-
Journal of Ubiquitous and Future Networks