Master's Thesis at the University of Basrah Explores Predicting Stock Investment Returns and Risks Using Neural Networks

A master's thesis at the College of Administration and Economics, University of Basrah, explored predicting stock investment returns and risks using multi-layered neural networks. The study focused on a sample of banks listed on the Iraq Stock Exchange for the period 2012–2024.

The thesis, submitted by student Abdul Khaliq Mudhar Mahmoud, aimed to test the ability of multi-layered neural network models to predict stock investment returns and risks based on historical data.

The thesis included the development of a predictive model that enabled the researcher to estimate returns and risks for a six-year period, from 1/1/2025 to 31/12/2030

The thesis concluded that multi-layered networks can predict returns, while a hybrid model combining modern and traditional methods can predict risks.

The thesis recommended adopting these models to support investment decisions.