Momentum Strategies¶
Strategies that profit from price trends continuing.
Overview¶
Momentum strategies exploit the tendency for winning stocks to keep winning and losing stocks to keep losing. This is one of the most well-documented anomalies in finance.
Why Momentum Works¶
- Behavioral: Investors underreact to news initially
- Herding: Success attracts more buyers
- Slow information diffusion: Not everyone gets news at once
- Analyst coverage: Forecasts adjust slowly
Strategies in This Section¶
| Strategy | Description | Complexity |
|---|---|---|
| Price Momentum | Classic 12-1 momentum | Basic |
| Industry Momentum | Sector rotation | Intermediate |
| Trend Following | Moving average systems | Basic |
| Momentum + Quality | Quality-filtered momentum | Intermediate |
Quick Example¶
Text Only
data:
source = "prices.parquet"
format = parquet
signal momentum:
// 12-month return, skip last month (12-1)
ret_12m = ret(prices, 252)
ret_1m = ret(prices, 21)
momentum = ret_12m - ret_1m
emit zscore(momentum)
portfolio main:
weights = rank(momentum).long_short(top=0.2, bottom=0.2)
backtest from 2015-01-01 to 2024-12-31
Key Considerations¶
Momentum Crashes¶
Momentum can experience sharp reversals: - 2009 momentum crash - Market turning points
Mitigation:
Text Only
// Add volatility scaling
vol_scale = where(vix > 30, 0.5, 1.0)
weights = base_weights * vol_scale
Turnover¶
Momentum strategies can have high turnover. Balance: - Rebalancing frequency - Transaction costs - Signal decay
Capacity¶
Works best with smaller positions due to: - Market impact - Signal crowding
Expected Performance¶
| Metric | Typical Range |
|---|---|
| Sharpe | 0.4 - 0.8 |
| Annual Return | 5% - 12% |
| Max Drawdown | 15% - 30% |
| Turnover | 200% - 400% |
Research References¶
- Jegadeesh & Titman (1993): "Returns to Buying Winners and Selling Losers"
- Carhart (1997): "On Persistence in Mutual Fund Performance"
- Asness et al. (2013): "Value and Momentum Everywhere"