How to Trail Stops in Futures Trading: Three Methods and Their Trade-Offs

A fixed profit target keeps you consistent but caps the trade when the market keeps going. A trailing stop lets a winner run but can hand back open profit on a routine pullback. Neither is correct in the abstract. Which one suits you depends on how your strategy behaves, and you find that out by testing.

Method 1: fixed-distance trail

You move the stop a set number of points behind price. Example: long ES at 5,800 with an initial stop at 5,790, a 10-point risk, or $500 per contract. The rule is that each time price makes a new 5-point gain, the stop moves up 5 points. At 5,805 the stop is 5,795, at 5,810 it is 5,800, and so on.

It's easy to automate and easy to review. Its weakness is that the same 5 points means different things on a quiet day and a fast one, so in a fast market you may get stopped by normal noise.

Method 2: ATR trail

ATR (Average True Range) measures recent bar size, so the trail distance adapts. Example: 5-minute ATR(14) on ES reads 8 points, and your rule is stop = highest high since entry minus 1.2 x ATR, updated at bar close. That's 9.6 points, which you round to a valid tick, such as 9.5 or 9.75. If ATR rises to 12, the trail widens to 14.4 points. The multiplier is a choice, and 1.2 is only an illustration. Test several values and compare results after costs.

Method 3: swing-structure trail

You trail the stop to just beyond the most recent swing low (long) or swing high (short). Example: long from 5,800, price makes a swing low at 5,802 and continues to 5,815. You move the stop to 5,800, below that swing. A later swing low at 5,810 moves it to 5,808. The logic is that if price breaks the last level where buyers stepped in, the move has changed character.

The catch is that swings are subjective. Define one in writing: for instance, a swing low is a bar whose low is lower than the two bars on each side. Without a definition you'll adjust it to feel comfortable, and then you're no longer testing a rule.

Fixed target or trail?

I'd pick by looking at the data from your own trades, not by gut feel. Log each trade's maximum favorable excursion, the farthest it went in your favor, and compare it to the target you used. If winners often travel much farther than your target, a trail may help. If they usually reverse right after your target, a fixed exit may be better. Many trades in a choppy session will do poorly under any trail.

Two contracts: partial exit plus trail

One common structure needs at least two contracts, or two micros. Hypothetical ES example with a 10-point stop:

1. Enter 2 ES at 5,800 with stops at 5,790. Total risk is 2 x 10 x $50 = $1,000 before costs and slippage.
2. At 5,810 (1R), sell one contract for +$500.
3. Move the stop on the second contract to 5,800 (breakeven, before costs).
4. Trail the second contract with your chosen method.

Be precise about the outcomes. Before 1R is reached, the worst case is the full $1,000 loss, and a stop can slip. After step 2 and 3, the trade is no longer at risk of a net loss unless the second stop fills badly, since the first contract has banked $500. Commissions come off either way. Taking half at 1R lowers variance, but it also gives up some of the upside on the contract you sold, so compare the structure against your single-exit version over the same trades.

Automating in NinjaTrader

NinjaTrader's ATM strategies can manage the stop for you. In the stop strategy settings you'll find Auto Breakeven (moves the stop to breakeven after a profit trigger) and Auto Trail (steps the stop as profit grows, with settings for the profit trigger, the stop distance and how often it moves). ES moves in 0.25 point ticks, so 4 ticks is 1 point. Check the exact menus and field names in the official NT8 guide at ninjatrader.com/support/helpguides/nt8, because they can differ by version.

One detail worth knowing: a simulated stop is held by the platform and not at the exchange. If your connection or computer fails, nothing is working that stop. A resting stop at the exchange has the opposite trade-off. Decide which risk you prefer, and test the template on a SIM account first. SIM fills don't show real slippage or queue position.

If you use any automation with a prop firm, read the firm's current rules first. Apex, for example, prohibits automated or algorithmic preset bots that place, modify or cancel trades without personal real-time review, per its prohibited activities page.

The mid-trade tightening problem

The usual way a trail goes wrong is that you change it halfway through a trade because you're nervous. If your rule says the stop sits at 5,800 and price dips to 5,802, you've made the decision in advance. Overriding it removes the rule you were testing, so your results no longer say anything about the rule. If you want to try a different trail, change it for the next batch of trades and log it as a different version.

Test each method on a SIM account or in replay across the same set of trades, record results in R multiples after costs, and keep the one your own data supports. YMI's documented bot behavior is described at /docs/kpl-bot-specs.