Trend Following with ETFs: A Practical Approach
Trend following is a systematic investment methodology that seeks to capture sustained price movements across asset classes by buying assets in uptrends and avoiding or shorting those in downtrends. Exchange-traded funds have made this strategy accessible to individual investors and advisors alike, offering low costs, liquidity, and broad diversification. This article outlines a practical framework for implementing trend following with ETFs, covering universe selection, signal construction, position sizing, risk management, and execution.
Why ETFs Suit Trend Following
ETFs provide several structural advantages for trend-following portfolios. They trade intraday like stocks, allowing precise entry and exit at calculated signals. Expense ratios for broad index ETFs frequently fall below 0.10%, preserving returns in a strategy that may trade several times per year. Most importantly, the ETF universe spans equities, fixed income, commodities, currencies, and real assets, enabling the cross-asset diversification that gives trend following its risk-reducing properties. Futures-based commodity ETFs and managed futures ETFs now let retail-sized accounts participate in markets once reserved for institutional traders.
Constructing the ETF Universe
A robust trend-following universe should include 20 to 50 liquid ETFs spanning multiple asset classes. A representative list might include U.S. large-cap (SPY, VTI), developed international (EFA, VEA), emerging markets (EEM, VWO), sector funds (XLK, XLE, XLF, XLV), long-term Treasuries (TLT), intermediate bonds (IEF), investment-grade credit (LQD), high yield (HYG), gold (GLD), broad commodities (DBC), real estate (VNQ), and the dollar (UUP). Selection criteria should emphasize average daily volume above $10 million, tight bid-ask spreads under five basis points, and at least five years of history. Avoid leveraged and inverse products for core trend signals; their daily reset mechanics cause volatility decay that erodes long-horizon returns.
Designing the Signal
The workhorse signal is the moving average crossover or price-versus-moving-average rule. A common specification: hold an ETF when its adjusted close exceeds its 200-day simple moving average, exit when it closes below. Faster variants using 50-day or 100-day averages generate more signals but suffer whipsaws in choppy markets. Dual moving average systems—buy when the 50-day crosses above the 200-day, sell on the reverse—reduce false positives at the cost of lag.
Momentum ranking offers an alternative: each month, rank the universe by 12-month total return excluding the most recent month, hold the top quintile, and rebalance. Time-series momentum—comparing an asset’s own return to zero over lookback windows of 3, 6, and 12 months—can be blended into a composite score, with the average of three binary signals (each above or below zero) producing a position score between zero and one. Academic research, including Moskowitz, Ooi, and Pedersen’s work on time-series momentum, supports lookbacks between one and twelve months. Combining a 200-day filter with 12-month momentum ranking captures both absolute and relative trends, historically improving risk-adjusted returns versus either alone.
Position Sizing and Portfolio Construction
Equal weighting across active positions is the simplest approach: if the signal flags 12 of 30 ETFs, allocate 1/12 of capital to each. Volatility targeting refines this by scaling positions inversely to realized volatility—using 20-day or 60-day standard deviation of daily returns—so that each position contributes roughly equal risk. A target of 10% annualized volatility per position, capped at 25% of capital, prevents any single holding from dominating portfolio risk. Portfolio-level volatility targeting, adjusting total exposure to maintain a 10–12% target, further smooths the equity curve.
In practice, allocate signal scores across the universe: a full position for scores above 0.75, half positions between 0.5 and 0.75, and cash otherwise, then normalize weights to the desired gross exposure. This graduated approach reduces turnover compared with binary rules while retaining responsiveness.
Risk Management
Stop-loss rules complement trend signals. A trailing stop of 8–10% below the highest close since entry, or a hard stop tied to two times average true range, limits drawdowns when trends reverse abruptly. Portfolio-level circuit breakers—cutting gross exposure by half if the portfolio falls 10% from its peak—provide additional protection. Correlation monitoring matters: in crises, ostensibly diverse assets converge, so cap exposure to any single asset class at 40% and to any single ETF at 15%. Finally, paper-test any rule set for at least five years of data, including 2008, 2020, and 2022, before committing capital, and document rules in writing to prevent discretionary overrides during drawdowns.
Execution and Rebalancing
Choose a rebalance frequency—monthly or weekly—and stick to it. Monthly rebalancing on the last trading day of the month, executed near the close, typically captures most of the available trend premium with manageable turnover of 200–400% annually. Use limit orders to control slippage, particularly in less liquid commodity and international funds. In taxable accounts, favor ETFs with high turnover inside tax-advantaged accounts and harvest losses on exited positions. Track results against a benchmark such as a 60/40 portfolio to evaluate whether the strategy delivers its intended crisis alpha.
Practical Pitfalls
Whipsaw losses in range-bound markets are the strategy’s chief cost; expect 30–40% of trades to lose money even when the system works over full cycles. Survivorship bias creeps in when backtests draw on today’s ETF list, so test with funds that existed at each historical date. Data quality matters—use total-return adjusted prices, since distributions distort raw closes. Finally, resist the temptation to add discretionary filters after a losing streak; the edge comes from consistent application through full market cycles.
Trend following with ETFs is neither a prediction engine nor a guarantee, but a disciplined process for participating in persistent moves while limiting exposure to sustained declines. Its value emerges from breadth of markets, consistent signals, and rigorous risk control—executed with the low-cost, liquid building blocks that ETFs provide.
