This project let me merge my love for analytics with my curiosity about financial markets.
I built an Algorithmic Trading Simulator — a web-based application that tests trading strategies using real market data from the yfinance API.
The system allows users to simulate both active and passive investment strategies, evaluating them side by side against the NIFTY 500 benchmark.
To make the simulation realistic, I implemented a backend database in MySQL with normalized tables to handle corporate actions like stock splits and bonus shares.
One of the most challenging and rewarding parts was coding the tax calculation engine, tailored for Indian regulations — including STCG (20%) and LTCG (12.5%) with loss carry-forward logic.
On top of that, I built a performance dashboard that visualizes XIRR, cumulative returns, drawdowns, and compares how each strategy performs over time.
The project runs entirely through a Streamlit interface, combining the power of Python’s analytical stack with a clean, interactive UI that investors can actually use.
What I liked most was seeing how quantitative logic and real-world data come together to build something functional, insightful, and fast.
Tech Stack
- Languages: Python, SQL
- Libraries: Pandas, yfinance, NumPy, Matplotlib
- Frameworks & Tools: Streamlit, MySQL, Power BI (for summaries)
- Techniques: Backtesting, Financial Modeling, Portfolio Analytics
Key Learnings
- Designing end-to-end analytical applications with a database backend
- Integrating live data feeds into analytical models
- Applying quantitative finance principles in a real coding environment
- Balancing technical accuracy with clear visualization and usability
