Stock Portfolio Optimization

Stock Portfolio Optimization

Dr. (Mrs.) J. R. Pansare1, Swaraj Dhayarkar2, Swarali Deshpande3, Rohan Mate4, Aakanksha Patil5

1Modern Education Society’s Wadia College of Engineering, Department of Computer Engineering, SPPU Pune, India

1jayshree.pansare@mescoepune.org 2sdhayarkar7@gmail.com 3swaralideshpande03@gmail.com 4rohanmate17@gmail.com 5aakankshapatil85231@gmail.com

Abstract— This project focuses on the development of a portfolio optimization system that leverages the Sharpe Ratio for risk-adjusted return maximization, combined with machine learning models, specifically Long Short-Term Memory (LSTM) networks, to enhance investment decision-making. By utilizing historical stock data, the system predicts future returns and optimizes portfolio allocation to maximize the Sharpe Ratio. The project integrates deep learning techniques for return prediction and uses optimization algorithms to adjust the portfolio’s asset weightings. Through this approach, the system achieves an efficient balance between risk and return, offering a robust solution for investors. The inclusion of Monte Carlo simulations, backtesting and graphical visualization further ensures the reliability and performance of the optimized portfolios. This application provides investors with a comprehensive tool for real-time portfolio management, allowing for informed decisions based on predictive models and financial metrics.

Keywords— Portfolio Optimization, Sharpe Ratio, Risk-Adjusted Returns, Long Short-Term Memory, Machine Learning, Stock Prediction, Asset Allocation, Monte Carlo Simulation, Back-testing, Financial Modelling.