01
Self-directed Project
2026
Quantamental Equity Research Model
Project Frame:
Built a point-in-time quantamental equity research model for long-term U.S. stock selection, using approximately 30 years of historical data across around 2,600 U.S.-listed equities.
Snapshot:
- Performance
- preliminary / illustrative / self-directed backtest: 14.2% net CAGR, 16.9% volatility, -27.4% max drawdown, 0.74 Sharpe
- Benchmark-relative
- 2.4% annualized alpha, 0.94 beta, 0.43 information ratio, 6.5% tracking error
- Factor Validation
- 0.028 Rank IC, 0.35 ICIR, 0.58% monthly top-bottom spread
- Implementation
- 18% monthly turnover, >$50m median ADV, 15.0% gross CAGR, 14.2% net CAGR
- Robustness
- Positive alpha in 3/4 market regimes, 1.5-2.0% early OOS alpha, overlay reduced peak drawdown by approximately 5 percentage points
- Limitation
- Preliminary OOS evidence; possible survivorship, delisting, vendor coverage, residual look-ahead, period, and multiple-testing risks.
Quantamental Research Equity Selection Factor Model Point-in-Time Data Portfolio Analytics
Key Work:
- Cleaned and structured historical U.S. equity data across price history, liquidity measures, corporate actions, and company-level fundamentals.
- Built point-in-time-aware controls to reduce look-ahead bias, stale-price distortion, incomplete-month effects, and survivorship-related risks.
- Classified stocks by GICS industry groups and developed industry-specific valuation rules based on metric applicability and business model differences.
- Designed seven interpretable signal families to identify valuation mispricing, business quality, momentum, market recognition, growth runway, capital allocation, and reinvestment quality.
- Constructed cross-sectional factor rankings and applied risk, liquidity, and valuation overlays before portfolio formation.
- Evaluated the model using portfolio performance, benchmark-relative performance, factor validation, implementation, and robustness metrics.