A trend-following strategy for trading futures contracts in the daily timeframe and a brief discussion of the risks and limitations.
For all backtests in this article we used Norgate data. We highly recommend this data service (we do not have a referral arrangement with the company.) Updated: January 6, 2023.
Timeframe: Daily (continuous back-adjusted contracts).
Markets: 23 futures contracts (Brent Crude oil, Crude oil, Cocoa, E-Mini S&P 500, Euro-BTP Long-term, Dax, Euro STOXX 50, Euro-Bund, Gold, Feeder Cattle, Lean Hogs, Copper, Coffee, Live Cattle, Frozen Concentrated Orange Juice Grade A, Palladium, Platinum, Silver, Corn, 10-Year U.S. T-Note, Chicago SWR Wheat, Euro-Buxl, US Dollar Index).
Strategy type: Trend-following based on breakouts, stop and reverse.
Maximum positions: 23, long/short.
Position size: Based on stop-loss and maximum risk per position.
Trade entry: All trades are executed at the open of the next bar.
Backtest range: 01/3/2000 –12/30/2022
The strategy is not optimized for the highest annualized return. The entries and exits are based on price breakouts. The breakout lookback period is not optimized but based on the popular length used by trend-following CTAs.
Equity Curve, Drawdown Profile, and Yearly Returns
The annualized return is 12.8% and the maximum drawdown is 41%. Since 2000, there are seven down years with the largest loss of 8.3% in 2012. The strategy has provided convexity during the dot-com and GFC bear markets. From 2000 to 2002, the strategy gained a combined 37%, and for 2008 the strategy gained nearly 66%. In 2022, the strategy was up 20%.
Performance Summary and Comparison to SPY ETF Buy and Hold
| STRATEGY | SPY BUY AND HOLD | |
| CAGR | 12.8% | 6.2% |
| MDD | -41.0% | -55.2% |
| VOLATILITY | 18.8% | 19.8% |
| SHARPE | 0.68 | 0.31 |
| TRADES | 1247 | 1 |
| LONG TRADES | 639 | 1 |
| SHORT TRADES | 608 | – |
| AVERAGE TRADE | 14.0% | – |
| PROFIT FACTOR | 1.48 | – |
| WIN RATE | 21.8% | – |
| AVG. BARS IN TRADE | 78.8 | – |
| EXPOSURE | 60.5% | 100% |
The strategy Sharpe ratio is more than double the SPY ETF buy and hold Sharpe ratio for two-thirds of the exposure.
Monte Carlo Simulation
According to the Monte Carlo simulation results, there is a 5% chance the maximum drawdown will be larger than 57% and a 1% chance it will be larger than 66%. (Caveat emptor: Monte Carlo analysis has limitations.)
Average Monthly Returns Chart
March, May, June, and October have underperformed on average. May has been a negative month. There appears to be strong seasonality in the first two months of the year, on average.
Using a Different Group of Futures Contracts
In the backtest below we used 11 futures contracts that are included in at least one CTA index replication model of an ETF. All the other backtest parameters were kept the same.
Markets: Crude oil, E-Mini S&P 500, Gold, 10-Year U.S. T-Note, Ultra U.S. T-Bond, Ultra 10-Year U.S. T-Note, 2-Year U.S. T-Note, U.S, T-Bond, Eurodollar, Euro FX, and Japanese Yen.
Equity Curve, Drawdown Profile, and Yearly Returns
The CAGR of 12.1% is 70 basis points lower as compared to the backtest with the 23 contracts, but the maximum drawdown is significantly lower, 21.7% for the 11 contracts versus 41% for the 23 contracts.
In 2022. there was an outlier return of about 87% before transaction costs. The reason is the concentration on fixed-income futures and the sharp decline in this market in 2022, but also the trends in currencies. All contracts generated profit in 2022. The top performer was the 2-Year U.S. T-Note.
Caveat emptor: Choosing the markets to trade based on recent results may be the outcome of several biases, including but not limited to selection, data-mining, data-snooping, and hindsight. In general, these biases may lead to overconcentration in specific markets, for example, currencies or fixed-income, and could lead to large losses when these markets enter a consolidation period. A large number of markets guarantees that the impact on the long-term profitability of any tail or undesirable market events will be limited.
Rules of the strategy
The rules of the strategy are available for sale. Contact us for details.
The rules included below are sufficient for implementing the strategy in a backtesting or trading platform.
Risks of Trend-following Strategies
Trend-following strategies attempt to capture extended trends in markets also referred to as outlier trades. Due to implementing stop-losses, the win rate is small because the strategies exit with a loss until a trend starts. This style of trading can be demoralizing and could lead to a loss of discipline. Trend-following strategies, as the one analyzed in this article, are fundamentally simple; the alpha comes from their systematic approach to trading the markets.
In addition, trends and trend-following are not the same things and many confuse the two. A trend-following strategy will not find trends where there are none. From 2004 to 2016, or 13 years, the annualized return of the top 20 CTAs was 2.4%. However, the dispersion in CTA returns is high and not reflected in the averages. There is a high specific risk in choosing a CTA or a trend-following strategy.
Furthermore, the proper execution of trend-following strategies with many futures contracts requires sufficient capitalization. For about 20 contracts, the approximate initial capital is probably around a million dollars. Otherwise, due to insufficient capital, a strategy could miss a few profitable trends and that will negatively impact the Sharpe and MAR ratios.
Click here for a list of strategies.
Disclaimer: No part of the analysis in this blog constitutes a trade recommendation. The past performance of any trading system or methodology is not necessarily indicative of future results. Read the full disclaimer here.
Charting and backtesting program: Amibroker. Data provider: Norgate Data
If you found this article interesting, you may follow this blog via RSS or Email, or on Twitter
