Scaling in Trading: How Scaling In and Scaling Out Change Risk and Reward

Scaling in means building a position in several entries instead of one; scaling out means closing it in several exits. Both change two numbers you can compute before the trade: how much you lose if the stop is hit, and how much you make if the target is reached. The worked examples below, all hypothetical, show that neither technique lowers risk or raises gains by itself: each one trades part of one for part of the other.

Scaling in: adding to a position

Adding to a winning trade (pyramiding) raises the average entry price. If the stop stays where it was, the money at risk grows faster than the position.

The rule that follows: after every add, recompute the risk from the new average price and move the stop so that the total stays within the amount you set for one trade (see position size). Adding to a losing trade (averaging down) does the opposite: it lowers the average price but increases the position just as the market proves the entry wrong, and the risk only stays bounded if the stop does not move.

Three ways to scale in

Traders who scale in do it in one of three ways, and they do not carry the same risk.

• Equal parts as the trade works. The plan sets the full size in advance and buys it in equal slices at set intervals in your favor. The Turtle rules did this with Units: one Unit at the breakout, then one more every half N (half the average true range), up to four, with the stop of every Unit raised by half N at each add. In the rules' crude oil example, the stops move from 25.90 to 27.70 as the adds come in at 28.90, 29.50 and 30.10. More on our Turtle Traders page.
• A pyramid. Each add is triggered by a profit, and the slices can grow. The narrator of Reminiscences of a Stock Operator, the book about Jesse Livermore, describes his rule for cotton: buy in slices of 10,000, 10,000 and 20,000 bales, each only after the market had moved his way, and get out entirely if the first 10,000 or 20,000 bales showed a loss.
• Averaging down. Each add comes after the price has moved against you. The average price falls, but the position grows exactly while the market says the entry was wrong, and the stop is the only thing that caps the loss.

The first two only add to a trade that already shows a profit; the third adds to one that shows a loss. That is the whole difference, and it is why the stop must be recomputed after every add, whichever method you use.

Scaling out: taking partial profits

Here the trade is 200 shares bought at $50 with a stop at $48, so the initial risk (1R) is $400. Two plans are compared: hold everything to a $56 target, or sell half at $52 and move the stop on the rest to $50, the entry price.

Hold all: 200 shares to $56 or $48. Scale out: 100 sold at $52, stop on the other 100 moved to $50. With 1R = $400, the results are −$400, −$400 or +$1,200 for holding, and −$400, +$200 or +$800 for scaling out. Hypothetical, before costs. Scaling out turns the middle scenario from a loss into a gain, and gives up one third of the profit in the best one. Which plan earns more depends only on how often each scenario happens: holding everything comes out ahead when the full run to $56 happens more than 1.5 times as often as the reversal after $52, because it gains 1R more in the first case and gives up 1.5R in the second.

Moving the stop after a partial exit

Moving the stop to the entry price after the first exit is what makes the middle row positive: the half still open can no longer lose. A trailing stop goes one step further and follows the price, which locks in part of the open gain but can close the trade on an ordinary pullback. Each choice is a new trade-off between the losses you avoid and the gains you cut, and the same table method applies: write the scenarios, compute the result of each in R.

What scaling costs

Each extra entry or exit is one more order, and every order pays the spread and any commission. Two entries and two exits cost twice as much as one of each, and the R values in the table above are before those costs. On a long-term trade the difference is small; on a short trade with a close target it can take a large share of the gain, the same problem our scalping page measures.

Scaling also needs a position large enough to split. You cannot sell half of one contract or of one share. The Turtle rules show the limit in their own example: on a $100,000 account, the size for one Unit of heating oil came out at 1.688 contracts and was rounded down to 1, which leaves nothing to scale. On a small account, check that each slice is at least the minimum order size before you plan on several.

Is it true that 97% of day traders lose money?

The figure comes from one study of the Brazilian equity futures market: among individuals who began day trading between 2013 and 2015 and persisted for at least 300 days, "97% of them lost money". It is a measured result for that market and period, not a universal law, and it does not separate traders who scaled in or out from those who did not. Scaling changes the shape of each trade's result; it does not create an edge where there is none.

Exits and targets are compared on risk reward ratio and stop loss order; more definitions are in trading concepts.