Trend Following

Trend following is a trading strategy that takes positions in the direction of recent price movement. It buys markets that have been rising, sells or sells short markets that have been falling, and holds each position until the price action indicates that the trend has ended. The approach makes no attempt to forecast turning points or to judge what an asset is worth. It reacts to prices alone.

The strategy is usually applied systematically across a wide range of futures markets, covering share indices, bonds, currencies and commodities. It is the main strategy of the managed futures industry, whose firms are known as commodity trading advisors. Trend following is a form of momentum investing based on each market's own past return, as distinct from cross-sectional momentum, which ranks assets against one another.

A trend-following system needs a rule to detect a trend, a rule to exit and a rule for position size. Common trend signals include the following.

• Moving averages. A moving average crossover buys when a short-term average of price rises above a long-term one and sells when it falls below.
• Breakouts. Breakout trading rules buy when the price exceeds its highest level of the past several weeks and sell when it drops below its lowest.
• Past return. A time-series momentum rule goes long if the return over a look-back period, often twelve months, is positive, and short if it is negative.

Exits are usually built into the same rule, or set as a trailing stop-loss order that follows the price. Most programs combine several signals and look-back lengths so that results do not hinge on one parameter.

Positions are scaled so that each market contributes a similar amount of risk. A widely used formula sets the quantity from the amount the trader is prepared to lose and the distance to the exit:

• $\text{Capital}$ is the size of the account.
• $\text{Risk fraction}$ is the share of capital risked on one trade.
• $\text{Stop distance}$ is the loss per unit if the exit is triggered, often set as a multiple of the market's recent average daily range.

With capital of $1,000,000, a risk fraction of 1 percent and a stop placed $5 from the entry price, the position size is $1{,}000{,}000 \times 0.01 / 5 = 2{,}000$ units, and the planned loss is $10,000. A more volatile market has a wider stop and therefore a smaller position. The same principle underlies volatility targeting at the portfolio level.

The idea of trading with the trend is old, and rule-based versions were developed for commodity futures in the mid-twentieth century, notably by Richard Donchian. In 1983 the Chicago traders Richard Dennis and William Eckhardt recruited and trained a group of novices, later known as the Turtles, in a breakout system, to test whether trading could be taught. The managed futures industry grew substantially in the following decades.

Academic support arrived later. A 2012 study by Tobias Moskowitz, Yao Hua Ooi and Lasse Pedersen documented time-series momentum across dozens of futures markets, and subsequent work extended the evidence back more than a century.

Trend following has a distinctive pattern of results. Most trades lose a small amount, because markets spend much of their time moving sideways and signals reverse. A minority of trades catch large, sustained moves and produce most of the profit. The distribution of returns therefore has positive skewness, the opposite of strategies that earn small steady gains and occasionally lose heavily.

The strategy has tended to do well during prolonged market declines, because it ends up short falling markets. Managed futures funds on average gained in 2008 and again in 2022, when shares and bonds both fell. This record has led investors to hold it for diversification. The protection is not dependable in sudden reversals, where there is no time for a trend to form.

Explanations draw mainly on behavioral finance. Investors may react slowly to new information, so prices adjust gradually, and later investors may chase the move and extend it. Other accounts point to participants who trade for reasons other than profit, such as central banks managing exchange rates or producers hedging, whose activity can sustain price moves. None of these explanations is settled, and the strength of the effect has varied over time.

• Whipsaws. In range-bound markets the strategy repeatedly buys high and sells low. Losing periods can last for years, as in much of the decade after 2009.
• Reversals. A sharp turn in a crowded trend causes losses across many positions at once. In March 2023 a sudden rally in government bonds after bank failures in the United States produced heavy losses for trend followers that were short bonds.
• Drawdowns. A deep maximum drawdown is normal, and many investors abandon the strategy before a recovery.
• Costs. Trading many markets with leverage incurs transaction and management fees that reduce net returns.
• Capacity and crowding. As assets in the strategy have grown, similar positions held by many funds may make exits more disorderly.

The approach is the conceptual opposite of mean reversion, and the two are often combined because their returns tend to differ.