Trend following explained
Trend following is a rules-based approach that tries to stay with a price move after it has become visible. A rule might compare today’s price with a moving average, or compare a shorter average with a longer one. It does not need to explain why a price is moving. It defines observable conditions and responds when they occur.
The phrase covers many methods, not one standard strategy. A trend system needs an asset universe, a signal, an entry and exit convention, and a position-sizing rule. Two systems both called trend following can therefore behave very differently. Without the details, the label says little about what a backtest actually measured.
Trend rules are usually reactive. A moving average uses past prices, so a price must move before the average catches up. This delay can mean entering after an advance has already begun and exiting after a decline has started. That lag is not necessarily a coding error; it is part of using past data to identify a move. The trade-off is that a signal may avoid some short-lived noise while responding late to reversals.
A worked example
Imagine a fictional share trades at £100 and its 50-day average is £95. A simple illustrative rule enters when the close is above that average and exits when it closes below. If the share later closes at £90 while the average remains £96, the exit condition is met. These round prices are made up; the example is not a recommendation or a tested result.
This example leaves out important design choices. Does the rule check the close or intraday price? Is the order placed at that same close or the next available price? What happens if prices gap? What portion of cash is invested? A valid test must answer these questions before results are examined. Otherwise, the apparent rule can shift from one interpretation to another.
Why people study trends
A trend follower is interested in persistence: whether a move continues for some time after it has started. The idea is intuitive, but intuition is not evidence that a particular rule will work after costs or in a different period. A backtest can show what happened under explicit assumptions in historical data. It cannot guarantee persistence in future prices.
Trend systems can behave poorly when prices oscillate within a range. The rule may enter after a brief rise and exit after a brief fall, repeating small losses or costs. A system can also leave after a decline and miss part of a quick recovery. The path matters, not just the final direction. Maximum drawdown and time out of the market add context to a headline return.
A trend signal also does not predict a correction or bear market. It describes a condition derived from past prices. It can change only after the data changes, and the market may reverse before the signal responds. Treating a moving average as a forecast confuses a lagging summary with knowledge of the future.
How to evaluate a trend rule
Write the formula and timing precisely. Choose the instrument, data, lookback and exit rule before viewing the evaluation period. Compare the rule with a simple baseline over the same dates. Keep a held-out period separate from parameter selection. Include failed variants and account for the fact that trying many settings can produce an attractive result by chance.
Data quality matters. A test based only on companies still trading can omit delisted shares and flatter historical results. StrategyLens uses delisted prices where it can, while recognising that coverage is incomplete. A transparent description should say what data the test includes and what it cannot establish.
Costs matter to any frequent switching. StrategyLens simulations use the zero-cost basis under ADR 0033: no commission or slippage is charged, and the buy-and-hold comparison is charged the same. Real trading costs money, and real results can be lower, especially for rules that trade often. The zero-cost comparison is a stated limitation, not an estimate of what an investor would keep.
Trend following and moving averages
A moving average is one possible way to define a trend, not the definition of trend following itself. Some rules use price breakouts, some combine several time frames, and some adjust exposure based on volatility. Each added choice creates another decision that should be documented and tested without tuning to the final evaluation sample.
A common concern is that a slow rule reacts late while a fast rule changes direction frequently. There is no universally correct setting. Choosing a value because it produced the best historical result invites overfitting. A held-out test can check whether a choice made on earlier data also behaves similarly later, though one holdout cannot prove a durable effect.
Read results with their full path and limitations. A lower drawdown can accompany lower growth; a rule can also look calmer simply because it holds cash for some periods. Compare like with like and avoid turning a historical description into a claim that a strategy works. Trend following is a category of methods to study, not a promise.
This is research, not financial advice. Past results, real or simulated, do not tell you what will happen next.
An index can also look smoother than its components because it combines many securities. A trend signal on an index and one on an individual share are therefore not interchangeable experiments. The instrument, price source and treatment of distributions should be stated, especially when a comparison spans a long period. These details help a reader reproduce the question even if the data provider later changes its history.
What is trend following?
A rules-based approach that enters or stays invested after a price trend is established and exits when a defined signal says it has weakened or reversed.
Is trend following the same as momentum investing?
They overlap. Trend following often describes a rule for responding to an asset's own price path; momentum investing often describes selecting assets based on relative or absolute past performance.
What is a common weakness of trend following?
Signals lag, so a rule may enter after a move has begun and exit after a reversal. Repeated false signals can also occur in sideways markets.