EUVIMEDEUVIMED entdecken
Crossref

Forecasting Stock Price and Exchange Rate Volatility Using a Comparative Approach of Geometric Brownian Motion and Neural Networks

Rouzbeh Zarezadeh, Artin Beytari, Mohammadreza Ghorbanian, Hossein Badiei

Accounting, Finance and Computational Intelligence · 2026

DOI: 10.61838/jafci.334 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Accounting, Finance and Computational Intelligence, 2026. Verfasst von Rouzbeh Zarezadeh und 3 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This study aims to compare the predictive power of the Geometric Brownian Motion (GBM) model and neural networks in forecasting volatility of Tehran Stock Exchange price index and exchange rate. This applied quantitative study used monthly data of the Tehran Stock Exchange price index and official exchange rate from 2012 to 2023. Volatility presence was first confirmed using the ARCH test. Volatility modeling was then performed using a numerically solved GBM framework based on stochastic differential equations and the Fokker–Planck approach, and its performance was compared with an online-learning neural network model. Model accuracy was evaluated using the coefficient of determination (R²) and root mean squared error (RMSE). For the stock price index, the GBM model achieved R² = 0.5751 and RMSE = 0.059, outperforming the neural network (R² = 0.438; RMSE = 0.448). For the exchange rate, GBM yielded R² = 0.5245 and RMSE = 0.026, compared to R² = 0.410 and RMSE = 0.447 for the neural network. Both research hypotheses were statistically supported, confirming the superior predictive performance of the GBM model. The Geometric Brownian Motion model demonstrates significantly higher predictive accuracy and explanatory power than neural networks in forecasting stock market and exchange rate volatility, offering a more reliable analytical framework for financial forecasting and investment decision-making.

Zitieren

APA

Zarezadeh, R., Beytari, A., Ghorbanian, M., & Badiei, H. (2026). Forecasting Stock Price and Exchange Rate Volatility Using a Comparative Approach of Geometric Brownian Motion and Neural Networks. Accounting, Finance and Computational Intelligence. https://doi.org/10.61838/jafci.334

Vancouver

Zarezadeh R, Beytari A, Ghorbanian M, Badiei H. Forecasting Stock Price and Exchange Rate Volatility Using a Comparative Approach of Geometric Brownian Motion and Neural Networks. Accounting, Finance and Computational Intelligence. 2026. doi:10.61838/jafci.334

BibTeX
@article{Zarezadeh2026,
  author = {Zarezadeh, Rouzbeh and Beytari, Artin and Ghorbanian, Mohammadreza and Badiei, Hossein},
  title = {Forecasting Stock Price and Exchange Rate Volatility Using a Comparative Approach of Geometric Brownian Motion and Neural Networks},
  journal = {Accounting, Finance and Computational Intelligence},
  year = {2026},
  doi = {10.61838/jafci.334}
}
Kostenlos starten →