Realistic Algorithmic Trading Review Using Python

This project presents a comprehensive and realistic simulation framework for algorithmic trading using the Python programming language. A central component of the work is an author-wise review of 50 academic and industry research contributions related to trading strategies, models, and performance evaluation. By utilizing the Faker library for synthetic author and title generation, and pandas for structured data manipulation, the system creates randomized yet credible ETF-based trade records. These synthetic logs include details such as buy/sell dates, quantities, trade outcomes (profit/loss/open), and performance percentages, effectively mimicking real-world financial activity. In parallel, the literature survey compiles diverse insights from existing research, identifying recurring challenges such as model overfitting, unrealistic backtesting assumptions, poor generalization in volatile market conditions, and a lack of interpretability in black-box models. Each issue is paired with its respective proposed solution, creating a structured and comparative view of the domain’s evolution. To enhance accessibility and visualization, all data outputs are displayed within a user-friendly web-based dashboard developed using Streamlit. This includes a dynamically generated author-wise review table, an Excel export option, and a pie chart summarizing the distribution of trading results. By merging practical simulation with a literature-backed evaluation, the project aims to provide a holistic, research-driven foundation for understanding, analyzing, and prototyping algorithmic trading systems.