UK Quant Momentum Strategy

Quantitative momentum is a systematic, rules-based approach to capturing the momentum factor premium. It uses multiple lookback periods, risk adjustments, and statistical ranking to select stocks with strong recent performance.

Different lookbacks capture different trend durations. Composite (1M, 3M, 6M, 12M weighted) is more robust than single lookback. The 6-month period typically receives highest weight as it's most robust historically.

Momentum divided by volatility (like Sharpe ratio). Rewards high returns but penalizes volatility. Stock with +15% return and 20% vol beats +20% return with 40% vol. Creates smoother, more persistent momentum signal.

30 stocks is optimal for UK quant momentum. Too few is concentrated risk. Too many dilutes alpha. 30 provides good diversification while maintaining meaningful momentum exposure.

Weighting by inverse volatility so each stock contributes equal risk. Lower vol stocks get higher weight. Creates more stable returns than equal weight. Standard for quant strategies.

Sharp reversals where losers rebound and momentum stocks fall (like 2009 Q1). Can cause 20-30%+ drawdowns. Mitigate with regime filter (200 SMA), crash protection (10% decline trigger), and quality combination.

If FTSE 100 > 200 SMA = Bull market, full momentum exposure. If below = Bear market, reduce to 50% or defensive only. Simple but effective protection against momentum crashes in bear markets.

Very significant. Stamp duty 0.5% + commissions + spreads ≈ 0.8-1% round-trip. With 30-50% monthly turnover, annual costs 3-6%. Gross momentum ~6-10%, net ~3-6%. Must minimize turnover.

Optimize on rolling window (5 years), test on next period (1 year), roll forward, repeat. Prevents overfitting to single period. More realistic out-of-sample performance estimate than single backtest.

Score = 0.7×Momentum_Z + 0.3×Quality_Z. Quality = ROE, low debt, earnings stability. Select stocks scoring well on both. Reduces crash risk while maintaining momentum exposure.

Composite = 0.1×Mom_1M + 0.2×Mom_3M + 0.4×Mom_6M + 0.3×Mom_12M. Z-score within universe. Divide by volatility z-score for risk adjustment. Rank by final score. Select top 30.

Predict persistence (which momentum stocks continue), learn optimal weights (adaptive signal combination), classify regimes (timing). Models: Random Forest, XGBoost. Ensemble with traditional (60/40 split) for robustness.

Trade 5% of ADV: 30 stocks × £50M average ADV × 5% = £75M. FTSE 350 strategy: £100-300M capacity. Larger needs more stocks, slower execution. Monitor realized impact vs estimates.

Position: max 5% single stock. Sector: max 30%. Factor: momentum beta 0.8-1.2, market beta 0.9-1.1. Drawdown: 15% triggers 50% reduction. Daily monitoring, weekly review, monthly comprehensive report.

Cluster stocks by correlation, allocate between clusters, then within. Reduces concentration in correlated positions. More robust than traditional risk parity. Handles non-stationary correlations better.