Siddhant Shah

I develop systematic approaches to investment challenges—designing momentum strategies, optimizing sector portfolios, and building institutional AI platforms that integrate research, risk, and portfolio governance in real time. My work spans quantitative research, financial modeling, emerging markets, and AI/ML in capital markets, bridging precision analytics with business insight. Skilled in Python, SQL, Bloomberg, and advanced portfolio modeling, I craft data-driven solutions that accelerate smarter investment decisions.

Research • Machine Learning • Data Analytics • Python, R, SQL

Current Role

See All Experience

Boston University, Department of Computer Science

Oct 2024 - Present

Research Assistant
  • Developed proprietary trading strategies with mean-variance optimization across 9 sector ETFs, implementing risk-constrained portfolio construction and performance analytics (Sharpe ratios, drawdowns, transaction costs)
  • Published 3 research papers on trading strategies in leading journals (Stocks & Commodities Magazine, MLAIJ)
  • Achieved 19.9% annualized returns on crude oil momentum strategies and 68.65% returns on cross-asset Bitcoin trading over multi-year backtests

Skills: Python, R, SQL, Pandas, NumPy, scikit-learn, PyTorch, Backtesting, Portfolio Optimization

Published Research

See All Publications

Momentum-Based Trading Strategies in Crude Oil ETFs And Futures

Siddhant Shah, Eugene Pinsky

Technical Analysis of STOCKS & COMMODITIES

Research on momentum-based trading strategies in crude oil ETFs and futures, developing long-short models yielding up to 19.9% annualized returns over an 18-year testing period.

ResearchMomentum StrategiesCrude Oil

The Silver Lining of Daily Bitcoin Trading

Siddhant Shah, Eugene Pinsky

Technical Analysis of STOCKS & COMMODITIES

Strategy leveraging overnight silver returns to predict Bitcoin price movements, exhibiting lower drawdowns in a 10-year backtest with 68.65% annualized returns.

CryptocurrencyBitcoinSilver

Estimating the Accuracy of a Bagged Ensemble

Siddhant Shah, Eugene Pinsky

Machine Learning and Applications: An International Journal (MLAIJ)

Probabilistic framework to reduce computational overhead in model fine-tuning, using various distributions to estimate Random Forest performance with less than 3% relative error.

Machine LearningRandom ForestsEnsemble Methods

Featured Projects

See All Projects

PySpark vs KDB+/q Performance Analysis

High-performance financial analytics system comparison achieving 50-300x performance improvements with KDB+/q over traditional systems for time-series operations. Microsecond-level query response and 5-8x memory compression.

F1 Lap Time Prediction and Feature Analysis

End-to-end machine learning framework for F1 lap time prediction using real-time telemetry data, achieving 94.8% R² through advanced feature engineering with track curvature, elevation profiles, and driver performance metrics.

Finlatics - Business Analyst Experience Program

Introduction to working as a Business Analyst with MS Excel & Power BI

Technical Expertise

Quantitative Finance

  • Portfolio Optimization & Risk Management
  • Algorithmic Trading Strategy Development
  • Statistical Arbitrage & Alpha Generation
  • Backtesting & Performance Attribution

Machine Learning & Data Science

  • Ensemble Methods & Random Forests
  • Time-Series Analysis & Forecasting
  • Feature Engineering & Model Optimization
  • Statistical Modeling & Hypothesis Testing

Programming & Tools

  • Python (Pandas, NumPy, scikit-learn, PyTorch)
  • R (Statistical Analysis & Visualization)
  • SQL, PySpark, KDB+/q
  • Git, LaTeX, Advanced Excel/VBA

Certifications & Education

  • MS in Applied Data Analytics (Boston University)
  • BS in Mathematics & CS (Chennai Mathematical Institute)
  • CFA Level 1 Candidate (CFA Scholarship Recipient)
  • Bloomberg Market Concepts Certified