Quantitative Analyst
During my internship, I played a pivotal role in developing and optimizing a data-driven investment strategy for Tunisian stocks. Key contributions include:-Project Leadership: Spearheaded the project’s planning and methodology, ensuring alignment with quantitative finance principles and data science best practices.-Data Preprocessing & Analysis: Led data cleaning efforts, addressing missing/extreme values, and conducting exploratory analysis on 6 stocks from various sectors, using statistical methods to understand market behavior.-Leveraging Technical Indicators: Applied comprehensive correlation analysis to identify the most effective technical indicators for forecasting each stock's price movement. This allowed for the selection of indicators with the highest predictive power, ultimately enhancing the accuracy of our neural network models.-Advanced Forecasting Models: Worked with a team member to design and implement 6 LSTM and Bi-LSTM neural networks, successfully predicting closing prices with a high degree of accuracy (average MAPE of 1.43%) over a 10-day forecast horizon.-Monte Carlo Approach: Applied a Monte Carlo methodology to ensure robustness against the inherent stochastic behavior of both the neural networks and optimization algorithms, leading to more reliable predictions and robust models.-Portfolio Optimization: Applied the Mean-MAD Optimization Framework to determine optimal portfolio allocations based on predicted stock prices and historical stock prices, rigorously back-testing against benchmarks to ensure robustness and profitability.
