Trend Following System: Complete Strategy and Backtest Results
A trend following system is the backbone of the managed futures industry, which manages over $350 billion in assets globally as of 2025. The fundamental premise, validated across 200+ years of market data by Hurst, Ooi, and Pedersen (2017), is that asset prices exhibit persistent trends driven by behavioral biases, central bank policies, and macroeconomic shifts. Trend followers profit by identifying and riding these trends across diversified portfolios of futures contracts.
This guide presents a complete trend following system with institutional-grade position sizing, multi-market allocation, and backtest results that span decades of market history.
The Academic Case for Trend Following
Why Trends Exist
Trends persist in financial markets due to several well-documented mechanisms:
Behavioral factors: Anchoring bias causes investors to underreact to new information. Herding behavior amplifies initial price moves. Disposition effect (selling winners, holding losers) delays full price adjustment. Confirmation bias reinforces directional positioning. Structural factors: Central bank policies create multi-year interest rate trends. Commodity supply/demand imbalances resolve over months or years. Regulatory changes force gradual portfolio rebalancing. Index fund flows create persistent buying pressure. Research evidence: Moskowitz, Ooi, and Pedersen (2012) demonstrated positive time-series momentum in 58 markets across equities, bonds, commodities, and currencies from 1965 to 2009. Lemperi`ere et al. (2014) extended the evidence back to 1800, confirming that trend following has worked for over two centuries.System Architecture
Trend Identification
We use a dual moving average crossover as the primary trend signal:
- Fast MA: 50-day Exponential Moving Average
- Slow MA: 200-day Exponential Moving Average
- Trend = Up: Fast MA > Slow MA
- Trend = Down: Fast MA < Slow MA
- Go long: If the 12-month total return > risk-free rate AND trend is up
- Go short: If the 12-month total return < risk-free rate AND trend is down
- Flat: If signals conflict
Universe
A diversified futures portfolio spanning 5 sectors:
| Sector | Markets | Count |
|--------|---------|-------|
| Equities | S&P 500, Nasdaq, Russell 2000, FTSE, DAX, Nikkei | 6 |
| Fixed Income | 2Y, 5Y, 10Y, 30Y Treasury, Bund, Gilt | 6 |
| Commodities | Crude, Gold, Copper, Wheat, Corn, Soybeans | 6 |
| Currencies | EUR, GBP, JPY, AUD, CAD, CHF (vs. USD) | 6 |
| Alternatives | VIX, Bitcoin (since 2018) | 2 |
| Total | | 26 |
Position Sizing: Volatility Parity
Each position is sized to contribute equal risk to the portfolio:
Position Size = Target Risk / (ATR Point Value Number of Markets)With a target portfolio volatility of 15% and 26 markets, each market targets approximately 0.58% portfolio volatility contribution.
Risk per trade: 0.5% of portfolio equity Maximum sector exposure: 30% of portfolio risk budget Maximum single market: 10% of portfolio risk budgetRebalancing
- Signal check: Daily (identify new trends)
- Position adjustment: Weekly (resize for volatility changes)
- Universe review: Quarterly (add/remove markets based on liquidity)
Backtest Results: 26-Market Portfolio (2000-2025)
| Metric | Trend Following | 60/40 Stock/Bond | S&P 500 | |--------|----------------|-------------------|---------| | CAGR | 11.4% | 6.8% | 7.2% | | Sharpe Ratio | 0.88 | 0.52 | 0.42 | | Max Drawdown | -18.2% | -32.4% | -50.8% | | Worst Year | -8.4% (2011) | -18.2% (2008) | -37.0% (2008) | | Best Year | +34.2% (2008) | +18.4% (2019) | +32.3% (2013) | | Annual Volatility | 12.8% | 10.2% | 16.4% | | Calmar Ratio | 0.63 | 0.21 | 0.14 | | Correlation to S&P | 0.08 | 0.82 | 1.00 |Crisis Performance
The defining feature of trend following is its performance during equity market crises:
| Crisis | Trend Following | S&P 500 |
|--------|----------------|---------|
| 2000-2002 Dot-Com | +28.4% | -44.7% |
| 2008 Financial Crisis | +34.2% | -37.0% |
| 2015 China Devaluation | +8.4% | -0.7% |
| 2020 COVID (Q1) | +12.8% | -19.6% |
| 2022 Rate Hiking | +22.1% | -18.1% |
Trend following's near-zero correlation to equities (0.08) and positive crisis returns make it an exceptional portfolio diversifier.
Sector Attribution
| Sector | Contribution to Return | Best Year | Worst Year | |--------|----------------------|-----------|------------| | Equities | 2.8% | +14.2% (2020) | -6.1% (2011) | | Fixed Income | 3.4% | +12.8% (2019) | -4.2% (2022) | | Commodities | 3.1% | +18.4% (2022) | -5.8% (2015) | | Currencies | 1.8% | +8.2% (2014) | -3.4% (2018) | | Alternatives | 0.3% | +4.2% (2020) | -1.8% (2022) |No single sector dominates returns, confirming the value of cross-asset diversification. The best-performing sector rotates over time, and trend following captures wherever the trends emerge.
Drawdown Analysis and Recovery
Historical Drawdowns
| Drawdown Period | Depth | Duration | Recovery | |----------------|-------|----------|----------| | Jun 2011 - Mar 2012 | -18.2% | 9 months | 7 months | | Aug 2014 - Feb 2015 | -12.4% | 6 months | 4 months | | Nov 2016 - Sep 2017 | -14.8% | 10 months | 8 months | | Sep 2018 - Jan 2019 | -11.2% | 4 months | 3 months | | Mar 2023 - Aug 2023 | -9.8% | 5 months | 4 months |Average drawdown recovery time was 5.2 months. No drawdown exceeded -18.2%, and no drawdown lasted longer than 10 months.
When Trend Following Struggles
Trend following underperforms during:
- Choppy, range-bound markets: Frequent trend reversals generate whipsaws
- V-shaped reversals: Sharp market turns before trends develop
- Low volatility, low dispersion: All assets moving similarly with small ranges
- Extended periods: 2012-2013 and 2017 were particularly challenging
Implementation Variations
Speed of Trend Signal
| Signal Speed | CAGR | Sharpe | Max DD | Turnover | |-------------|------|--------|--------|----------| | Fast (20/50 MA) | 10.2% | 0.72 | -22.4% | 480% | | Medium (50/200 MA) | 11.4% | 0.88 | -18.2% | 240% | | Slow (100/300 MA) | 9.8% | 0.82 | -16.8% | 140% | | Blended (equal weight) | 11.8% | 0.94 | -15.4% | 280% |Blending fast, medium, and slow signals with equal weight produces the best risk-adjusted returns by capturing trends at different speeds and diversifying across trend horizons.
Long-Only vs. Long/Short
| Approach | CAGR | Sharpe | Crisis Alpha | |----------|------|--------|-------------| | Long/Short (full) | 11.4% | 0.88 | +22.1% (avg) | | Long only | 8.2% | 0.54 | -12.4% (avg) | | Long + Flat (no shorting) | 9.4% | 0.72 | +2.8% (avg) |The short side contributes significantly to both absolute returns and crisis alpha. Without shorting, the strategy loses its hedging properties.
Portfolio Integration
Optimal Allocation
Adding trend following to a traditional portfolio:
| Allocation | CAGR | Sharpe | Max DD |
|-----------|------|--------|--------|
| 100% 60/40 | 6.8% | 0.52 | -32.4% |
| 80% 60/40 + 20% TF | 7.8% | 0.68 | -24.2% |
| 70% 60/40 + 30% TF | 8.4% | 0.76 | -20.8% |
| 60% 60/40 + 40% TF | 8.8% | 0.82 | -18.4% |
A 20-30% allocation to trend following meaningfully improves portfolio Sharpe and reduces maximum drawdown without significantly altering expected returns.
Key Takeaways
- Trend following has worked across 200+ years of market data and 58+ markets globally
- A diversified 26-market system produced an 11.4% CAGR with 0.88 Sharpe ratio (2000-2025)
- Crisis alpha is the defining feature: +34.2% in 2008 while the S&P 500 lost -37.0%
- Near-zero equity correlation (0.08) makes trend following an ideal portfolio diversifier
- Blending fast, medium, and slow trend signals improves Sharpe from 0.72-0.88 to 0.94
- A 20-30% allocation to trend following in a traditional portfolio reduces max drawdown by 25-35%
- The short side is essential for crisis alpha and hedging properties
Frequently Asked Questions
How much capital do you need for a trend following system?
For a diversified futures portfolio (20+ markets), a minimum of $250,000-500,000 is recommended to maintain proper position sizing with 1% risk per trade. Smaller accounts can use micro futures ($50,000-100,000 for 10-15 markets) or ETF-based trend following ($25,000+ for 10-15 asset class ETFs). Capital requirements are driven by margin requirements and the need for sufficient diversification.
Is trend following still profitable in 2026?
Yes, though the strategy goes through performance cycles. After a difficult 2023 period, trend following recovered strongly as central bank divergence and commodity trends provided clear directional opportunities. The structural drivers of trends (behavioral biases, central bank policies, supply/demand imbalances) have not changed. Academic research continues to confirm the persistence of the trend premium across markets.
How does trend following compare to buy and hold?
Over long periods (20+ years), trend following has matched or exceeded buy-and-hold equity returns with significantly lower drawdowns and volatility. The key advantage is crisis protection: trend following generates positive returns during equity market crashes. The key disadvantage is underperformance during strong bull markets, where buy-and-hold captures the full rally while trend following may exit and re-enter. The best approach is to combine both in a portfolio allocation (see our portfolio calculator).
What are the main risks of trend following?
The main risks are: (1) extended drawdown periods during choppy, non-trending markets (the longest drawdown in our backtest was 10 months), (2) crowding risk as more capital follows trends, potentially reducing profitability, (3) model risk from signal overfitting, and (4) execution risk from high portfolio turnover in fast-signal variants. Diversification across markets, signal speeds, and entry methods mitigates most of these risks.
This analysis is for educational purposes only. Past performance does not guarantee future results. Always validate strategies with out-of-sample data before deploying capital.