How to use a take-profit order

A take-profit order closes a position automatically once the market reaches a price in your favour: above your entry if you are long, below it if you are short. Set alongside a stop loss, the two define the whole range of a trade before it starts.

It is the more neglected of the pair. Enormous attention goes into where a trade should be abandoned and very little into where it should be harvested - and the second decision moves a strategy's results just as far as the first.

It is a limit order, which changes things

Where a stop becomes a market order and takes whatever price exists, a take profit is normally a limit order: it fills at your price or better, never worse. That asymmetry is quietly in your favour. A favourable gap straight through your target can fill above it.

The same mechanic has a cost. A limit order that is never reached is never filled, and a market can touch your level on the chart without your order executing - the relevant price is the one on the side of the market that closes your position, not the one drawn on the screen. A long position is closed at the bid; watching the ask reach your target proves nothing.

Where to put it

Two broad approaches. Derive the target from your risk - a multiple of the stop distance, so a $5 stop gives a $10 target at 1:2 - or take it from the market, placing it below the resistance a rally is likely to struggle at.

The second usually produces stranger numbers and better ones. A target at 1:2.4 because that is where sellers previously appeared is more defensible than a target at exactly 1:3 because 1:3 sounded right. Whichever you use, the ratio it produces has to be judged against the hit rate it needs, which is the subject of risk-to-reward and expectancy.

Partial exits, followed through

Rather than one target, many traders close a position in pieces. Here is what that does to the numbers on a 100-share position bought at $100 with a stop at $95, so $5 of risk a share, and $500 at risk in total.

1. Opened

   Position

   100 shares

   Entry

   $100

   Stop

   $95

   Risk (1R)

   $500

   One decision still to make: where, and in how many pieces, this gets closed.

2. First target: half out at $110

   Closed

   50 shares

   Gain

   $10/share

   Realised

   +$500

   Still open

   50 shares

   One R banked. The remaining half now has a much smaller downside relative to what has already been taken.

3. Remainder closed at $115

   Closed

   50 shares

   Gain

   $15/share

   Realised

   +$750

   Still open

   none

   Total realised $1,250 against $500 of initial risk.

The trade produced +2.5R. Had the whole position run to $115 it would have been +3R; had it reversed after $110 it would still have banked +1R. Partial exits do not improve outcomes, they narrow them.

Why partial exits are a trade rather than an upgrade

They feel like risk management and they are really a redistribution. Closing half at the first target guarantees something on the trades that reverse and caps you on the ones that run - and for strategies whose profitability depends on a handful of outsized winners, capping those is expensive.

Whether it helps is an empirical question about your own results, not a principle. The one thing it definitely does is make the strategy harder to evaluate, because the realised R of every trade is now a weighted average of several exits rather than one number.

The arithmetic of taking profit too early

This is the most consequential paragraph on the page. Suppose a strategy is designed around a 40% win rate, 2R winners and 1R losers. Its expectancy is (0.4 × 2) − (0.6 × 1) = +0.2R per trade. It works.

Now suppose the trader gets uncomfortable whenever a position moves into profit and routinely closes at +0.5R, while still letting losers reach the full −1R. The win rate is unchanged. The expectancy becomes (0.4 × 0.5) − (0.6 × 1) = −0.4R.

The strategy has not been slightly degraded. It has been turned from profitable into badly losing, by a habit that feels like prudence every single time it happens.

Cutting winners short and letting losers run their full distance is the same mistake twice, and only one half of it feels like a mistake.

Planned target, actual average winner

A trader running a 3R target whose realised winners average 1.4R is not running a 3R strategy. Their expectancy should be calculated on 1.4R, because that is what the method actually returns.

Both figures are on your own statement, and the gap between them is usually the single most informative number a discretionary trader can look at.

Linking the two orders

A stop and a target on the same position should be linked so that filling one cancels the other. Platforms usually call this OCO - one cancels the other - and handle it automatically for orders attached to a position.

It matters when they are placed separately. Sell 100 at $120 as a target and sell 100 at $90 as a stop, unlinked; the target fills, and the stop is still sitting there. If the market later falls to $90 it can execute against nothing, opening an unintended short. Seven of our hundred brokers name OCO explicitly among their order types, and most of the rest manage attached orders automatically - but it is worth knowing which your platform does.

Distant targets have a running cost

A target several weeks away means a leveraged position held for several weeks, and every night of that incurs a swap or financing charge. A gross gain of $400 that cost $90 in financing is a $310 trade.

This is a real argument against very distant targets on leveraged products, and it is entirely absent from the chart. Compare the net result of a close target reached quickly against a distant one reached slowly before assuming the larger number is better.

You do not need one

Plenty of strategies use no fixed target at all, exiting instead on a technical signal, a trend reversal, elapsed time, or a trailing stop that follows the market up and closes the trade after a reversal.

Fixed targets suit methods with a high hit rate and a defined move in mind. Trailing exits suit methods that need the occasional very large winner. Choosing between them is a strategy decision, not a preference.

Questions people ask about take-profit orders