A comparative study of ARIMA, SVMs, and LSTM models in forecasting the Moroccan stock market
by Hassan Oukhouya; Khalid El Himdi
International Journal of Simulation and Process Modelling (IJSPM), Vol. 20, No. 2, 2023

Abstract: Modelling and forecasting time series constitute important financial research areas for academics and practitioners. The stock markets significantly impact investment decisions in many different economic activities. Forecasting the direction of stock prices is essential in planning an investment strategy or determining the right time to trade. However, stock price movements (trends) are noisy, nonlinear, and chaotic. Predicting the evolution of price series to improve the return on investment is difficult. In this research, we are interested in modelling and forecasting the daily prices of a primary index in the Moroccan stock market. To this aim, we propose a comparative study between the results obtained from applying two independent approaches in machine learning (ML) methods as the support vector regression (SVR), long short-term memory (LSTM), and the classical method autoregressive integrated moving average (ARIMA) models. The empirical results show that LSTM and SVR nonlinear models can yield slightly better performance accuracy than the linear model ARIMA on average for log closing price series (37 trading days).

Online publication date: Fri, 02-Feb-2024

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Simulation and Process Modelling (IJSPM):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com