WhitmanTrading

How to Scale Out of a Position

To scale out of a position, decide the exit levels and the fraction taken at each before entering, then follow them mechanically. Partial exits reduce the size held through the largest moves, which is where most methods make their money.

Scaling out means closing a position in parts as it works. It is common, it feels considerably better than a single exit, and on most methods it produces a smaller total result — which is worth knowing before deciding whether the trade is worth making.

Before you start

A decision about what happens to the stop on the remainder, made before the first exit. Moving it to break-even is the usual choice and it has consequences.

An honest view of what partial exits do to your expectancy. They reduce the size held through the largest moves, and the largest moves are where most methods earn.

The exit levels written down at the same time as the entry. Deciding to take half off while watching is not scaling out; it is an improvised exit with a name.

The steps

1. Write the levels and fractions before entering

A range-bound stretch of price with planned exit levels.
Prices and fractions, decided at entry. Illustrative chart - not real market data.

“Half at the first target, the rest on the trail.” Two numbers and a rule, written when the position does not exist yet.

2. Use few exits, not many

A slice of price data with two exit points.
Two parts, not five. Illustrative chart - not real market data.

Each partial exit pays part of a round trip — about 2% of the median bar range of 0.493 on this site’s shared series — and adds a decision point.

3. Decide the remainder’s stop in advance

A long-horizon price series with a revised invalidation.
Break-even, or unchanged. Decide beforehand. Illustrative chart - not real market data.

Moving the stop to break-even after the first exit is the common choice. It removes the losing outcome and increases the chance of being stopped out of a move that continues.

4. Take the first exit at a level, not at a feeling

A slow-moving stretch of price reaching a defined level.
A structural level, not a comfortable amount. Illustrative chart - not real market data.

A prior high, a measured target, a fixed multiple of the risk. Somewhere the market has a reason to react, rather than wherever the position has become uncomfortable.

5. Let the remainder run on the original plan

The first half of a price series continuing after a partial exit.
The remainder still has a plan. Illustrative chart - not real market data.

The rest is not free money to be managed loosely. It has a trail or a target, decided at entry, and it is followed the same way the first part was.

6. Record it as one trade

A section of a price series recorded as a single idea.
One idea, one entry in the journal. Illustrative chart - not real market data.

One entry, with an average exit price. Splitting it into two records inflates the trade count and makes the win rate describe something that did not happen.

7. Compare it against a single exit, on your own record

The first half of a price series compared two ways.
What would one exit at the same target have produced? Illustrative chart - not real market data.

For thirty trades, record what a single exit at your final target would have made. That comparison answers the question for your method rather than in general.

How to tell it worked

Levels and fractions were written before the entry, in every case.

At most 2 partial exits were used per position.

The remainder’s stop rule was decided in advance, not at the moment of the first exit.

And a single-exit comparison exists over at least 30 trades.

What it does to the arithmetic

A candlestick chart annotated with the round-trip cost of a switch.
Each partial exit pays part of the cost again. Illustrative chart - not real market data.

It raises the proportion of winning trades and lowers the average win. More positions finish positive because part was banked early; each finishes smaller because less was held through the move.

A section of a price series drawn without volume context.
And in a thin market each exit costs more. Illustrative chart - not real market data.

On a method whose results come from a minority of large moves, that is a direct cost. Those are exactly the trades where holding the full size mattered, and scaling out reduces the size held in precisely them.

Why it is still worth doing

Because a position you can hold is worth more than one you cannot. If taking half off is what allows you to stay in a trade for its full duration rather than closing the whole thing at the first uncomfortable bar, the expectancy lost is a fee for a behaviour gained.

And because break-even stops end the losing outcome. After the first exit and a stop moved up, the trade cannot lose — which changes how the remaining position feels to hold, and feelings are what actually determine whether plans get followed.

Both of those are real and neither is a mathematical improvement. Being clear about which one you are buying is the whole of doing this honestly.

Running the comparison on your own record

Two columns beside each trade: what you actually made, and what a single exit at your final target would have made. Both numbers, every trade, for thirty trades.

The difference is what scaling out costs you. It will be a number rather than an impression, and on most trend-based methods it is larger than people expect because the largest trades dominate the total.

Then decide with the figure in front of you. If the cost is small, scaling out is close to free and the behavioural benefit is worth having. If it is large, the honest options are a single exit or a smaller position that you can hold whole.

What the comparison rules out is the middle position — scaling out while believing it improves the result. It usually does not, and the record settles that in a month rather than in an argument.

The original data

Of the 24,971 unique videos in research/search-study-corpus.jsonl, 5 mention scaling in the title, at a median of 5,240 views across 5 channels, and only 20% are instruction-shaped. Taking profit appears in 84 at 15,628 and win rate in 161 at 13,711. The counts come from site/corpus_count.py.

A candlestick series with several gaps, the largest of them marked.
A gap can skip every planned partial exit. Illustrative chart - not real market data.

5 videos on scaling against 84 on taking profit. The specific technique most people actually use has a seventeenth of the coverage of the general subject, and almost none of it addresses what partial exits do to the result.

A stretch of price bars cut short at a decision point.
Half taken, the rest ran another 200%. Regret it? Illustrative chart - not real market data.

The answer to the question on that chart is that this is the known cost, arriving as expected. Scaling out reduces what you hold through the largest moves — and if that trade is unacceptable, the answer is to stop scaling out rather than to scale out and be disappointed by it.

When it fails

The failure is the unplanned partial exit, and it converts a rule into a comfort mechanism. The position is working, the gain is larger than usual, and half comes off — not at a level, but at the point where holding became uncomfortable. That decision gets made earlier each time, because the discomfort arrives sooner as the position size grows. Within a few months the method takes half off almost immediately on every trade, and the exits have no relationship to the market at all.

The second failure is many small exits. Costs and decisions both multiply.

A third is no rule for the remainder’s stop. It gets decided under pressure.

A fourth is recording it as several trades. Every statistic downstream is wrong.

A fifth is treating the remainder as free. It still needs a plan.

And a sixth is never comparing against a single exit. The cost stays invisible.

Take profit covers the exit decision in general. Entry and exit is where both halves get planned together. And position sizing is what the fractions are taken from.

What I actually do

The honest framing is that I do this because it makes a position easier to hold, not because it makes more money. Taking half off reduces what I keep in the moves that run furthest, and those are where the result comes from. It is a trade of expectancy for the ability to stay in the trade at all.

— Michael Whitman

This page is educational, not financial advice. Test every idea on your own charts before risking money.