Degree
Master of Science in Finance
Department
Department of Finance
School
School of Business Studies (SBS)
Date of Submission
Fall 2023
Supervisor
Dr. Fawad Ahmed, Assistant Professor, Department of Finance, School of Business Studies (SBS)
Submission Type
Research Project
Document Type
Restricted Access
Pages
42
Keywords
DT models, GRU, LSTM, Neural network architectures, Stock price prediction, SVM
Abstract
This study investigates the effectiveness of LSTM and GRU neural network architectures in predicting stock prices using five years of daily price data from four major public companies in Pakistan. The dataset is divided into training and testing sets, and the performance of LSTM and GRU models is compared against SVM and DT models. The findings demonstrate that LSTM and GRU models outperform SVM and DT models, yielding higher average returns and better risk-adjusted performance. Moreover, the analysis of holding period returns indicates that the LSTM and GRU models provide predictions closely aligned with the actual returns experienced by investors. Based on these results, it is recommended that investors and traders utilize LSTM and GRU models for stock price prediction, Furthermore, it is recommended that further research be conducted to explore the potential of other models, such as deep learning models, for predicting stock prices.
Recommended Citation
., M. (2023). Stock Price Prediction using Neural Networks and Machine Learning Models: A case of Pakistan (Unpublished graduate research project). Retrieved from https://ir.iba.edu.pk/research-projects-msfin/2
The full text of this document is only accessible to authorized users.
