Heterogeneity In Stock Price Forecasting-Based on the ARIMA-GARCH Model And PCA-LSTM Model

Authors

  • Sirui Cheng

DOI:

https://doi.org/10.54097/me8jhv37

Keywords:

Stock forecasting; deep learning; heterogeneity; ARIMA model; LSTM.

Abstract

Financial theory suggests stock prices are mainly influenced by factors such as interest rates, market behavior, technical indicators, and firm value. Traditional approaches to stock price forecasting have been augmented by machine learning algorithms and time series models. Deep learning has experienced rapid development in the field of time series analysis and is becoming more mature. Therefore, this article delves into two prominent prediction methods: the ARIMA-GARCH model and Long Short-Term Memory network to compare their prediction performance and to analyze the heterogeneity in the effectiveness evaluation. The combination of the ARIMA and GARCH models makes it possible to account for both short-term fluctuations and long-term trends. And LSTM networks can capture the temporal relationships and diversified features of data, making it a popular choice for financial time series analysis. This study examines the forecasting of daily closing prices of representative individual stocks in Shanghai and Shenzhen A-shares and analyzes the factors that may cause heterogeneity in the prediction results. In addition to being affected by the chosen forecasting indicators, the differences in the forecasting results of individual stock prices can also be related to the company's economic information. Differences in the market structure of the industry and the effects of policy implementation can also lead to differences in forecasting results. These heterogeneity analyses provide references and suggestions for selecting models and variables in the financial field to improve prediction effectiveness.

Downloads

Download data is not yet available.

References

Hiemstra C, Jones J D. Testing for Linear and Nonlinear Granger Causality in the Stock Price‐Volume Relation. Journal of Finance, 1994, 49(5): 1639-1664.

Vapnik V. The nature of statistical learning theory. Springer Science & Business Media, 2013.

Atiya A, Talat N, Shaheen S. An Efficient Stock Market Forecasting Model Using Neural Networks. International Conference on Neural Networks.1997, 4: 2112-2115.

Xie X K, Wang H. Recurrent Neural Network for Forecasting Stock Market Trend. International Conference on Computer Science, 2016.

Sutskever I, Vinyals O, Le Q. Sequence to sequence learning with neural networks. In Advances in neural information processing systems, 2014, 3104-3112.

Nelson D, Pereira A, Oliveira R. Stock Market’s Price Movement Prediction with LSTM Neural Networks. International Joint Conference on Neural Networks, 2017, (4): 1419-1426.

Di Persio L, Honchar O. Artificial neural networks approach to the forecast of stock market price movements. International Journal of Economics and Management Systems, 2016, 1: 158-162.

Fischer T, Krauss C. Deep learning with long short-term memory networks for financial market predictions. European Journal of Operational Research, 2018, 270(2): 654-669.

Hotelling H. Analysis of a Complex of Statistical Variables into Principal Components. Journal of Educational Psychology, 1933, 417-441.

Kim H Y, Won C H, Forecasting the Volatility of Stock Price Index: A Hybrid Model Integrating LSTM with Multiple GARCH-Type Models. Expert Systems with Applications, 2018, 103: 25-37.

Downloads

Published

29-03-2024

How to Cite

Cheng, S. (2024). Heterogeneity In Stock Price Forecasting-Based on the ARIMA-GARCH Model And PCA-LSTM Model. Highlights in Science, Engineering and Technology, 88, 39-46. https://doi.org/10.54097/me8jhv37