Forecasting stock market behavior in BRICS economies using artificial neural machine learning models

Shrikant Panigrahi, Gagan Kukreja and Sumathi Kumaraswamy

Journal of Business and Socio-economic Development, 2025, vol. 6, issue 1, 70-89

Abstract: PurposeThis study aims to forecast the stock market behavior of BRICS nations (Brazil, Russia, India, China and South Africa) using advanced machine learning models. The focus is on identifying market trends, predicting future index prices and analyzing returns.Design/methodology/approachThe research applies long short-term memory (LSTM) neural networks and seasonal auto-regressive integrated moving average with exogenous variables (SARIMAX) to analyze daily stock index data from 2010 to 2023. Forecast accuracy is assessed using standard performance metrics, including root mean square error (RMSE), mean absolute error (MAE) and mean absolute percentage error (MAPE).FindingsLSTM model results produced positive return forecasts for Brazil, India and South Africa, while Russia and China exhibited negative returns or stagnant trends. SARIMAX forecasts predominantly indicated neutral or negative returns. India and China demonstrated higher prediction accuracy, with lower error rates across all metrics. The study reveals notable differences in forecasting performance between the two models.Research limitations/implicationsThe comparative results presented here support the growing body of evidence suggesting that neural networks are more suitable for forecasting in complex financial systems, especially when market conditions are influenced by both macroeconomic factors and investor behavior.Practical implicationsThe results suggest that investors seeking data-driven forecasting in emerging markets should prioritize markets like India and China, where LSTM shows high reliability. For policymakers, the findings reinforce the importance of improving data transparency and regulatory stability to improve market predictability and investor confidence.Originality/valueThis study contributes to the literature by combining deep learning and traditional time series forecasting methods to evaluate post-crisis stock market dynamics in emerging economies. The insights offer actionable insights for investors and policymakers aiming to optimize portfolios and design informed investment strategies in BRICS markets.

Keywords: BRICS; Stock market forecasting; LSTM; SARIMAX; Neural networks; Machine learning; Financial markets (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
https://www.emerald.com/jbsed/article/6/1/70/12714 ... et-behavior-in-BRICS (text/html)
https://www.emerald.com/jbsed/article-pdf/6/1/70/1 ... d-05-2025-0134en.pdf (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eme:jbsedp:jbsed-05-2025-0134

DOI: 10.1108/JBSED-05-2025-0134

Access Statistics for this article

Journal of Business and Socio-economic Development is currently edited by Professor Allam Hamdan

More articles in Journal of Business and Socio-economic Development from Emerald Group Publishing Limited
Bibliographic data for series maintained by Emerald Support ().