Risk and position size

What a stop actually is: a price, and a rule about that price

Everybody says "I had a stop at 1.20". That sentence is only half a stop. The other half is the rule that says what counts as reaching 1.20, and which order gets sent when it does - and that half is where the money actually goes.

About 10 minutes to read. Free, no account.

The two halves of a stop

A stop is the decision, made before you are in the trade, about the price at which you were wrong. It has two parts and both have to be written down.

The price. One number. Not a feeling, not "around 1.20", not "if it really breaks down". A number you could read out to somebody else.
The rule. What event at that price makes you leave, and what order you send when it happens. Two traders with the identical price can exit days apart and at prices 3% apart because their rules differ.

Start by converting the price into a distance, because the distance is the number every other lesson in this module needs.

Stop distance, in currency and in percent

entry             =  1.2500
stop price        =  1.2000

distance          =  1.2500 - 1.2000   =  0.0500
as a percent      =  0.0500 / 1.2500   =  4.00%

That 4.00% is the only output of the stop that matters arithmetically. It is what the next lesson turns into a position size, and it is what you compare against the size of a move you expect.

Touching a level is not the same event as closing through it

A candle is a summary of four prices: open, high, low, close. A stop rule has to say which of them it watches, and the two common answers produce different trades out of the same chart.

On touch. You are out the moment any trade prints at or through 1.2000. This is what a resting stop order on an exchange does - it watches every tick, not the candle.
On close. You are out only if a bar finishes below 1.2000. A wick through the level that recovers before the bar ends leaves you in the trade.

Here is one bar where the two rules disagree completely.

One 4-hour candle, two rules

open   1.2300    high   1.2350
low    1.1920    close  1.2080

on touch:  low 1.1920 is below 1.2000  ->  out
           the wick went 0.0080 past the stop,
           which is 0.0080 / 1.2500  =  0.64% of entry

on close:  close 1.2080 is above 1.2000  ->  still in

Neither rule is correct in general, and the trade-off is symmetrical, which is the part people skip. On touch takes you out of trades that would have recovered - the whole family of complaints that begin "it stopped me out and then went straight up". On close holds you through moves that keep going, and the loss you eventually take is larger than the one you planned, because the close can be anywhere.

Whichever you choose, choose it before the trade. Switching from "touch" to "close" while price is sitting on your stop is not a rule change, it is the absence of a rule, and it is the single commonest way a planned loss becomes an unplanned one.

Stop-market: you get out, at an unknown price

A stop-market order is an instruction to the exchange: when the trigger price trades, send a market order. A market order takes whatever is resting in the book, so it fills - but the price it fills at is whatever the book offers at that instant.

Normally that is a few ticks worse than the trigger. The difference is slippage, and it is bigger exactly when you need the stop most, because that is when the book is thin and everyone else’s stops are firing into it too.

The same stop, filled 0.0060 worse

trigger           =  1.2000
actual fill       =  1.1940

planned loss      =  1.2500 - 1.2000   =  0.0500   (4.00%)
realised loss     =  1.2500 - 1.1940   =  0.0560   (4.48%)
slippage          =  0.0060            (0.48% of entry)

A gap makes it much larger. If the last trade before a cascade is 1.2010 and the next one is 1.1300, your stop-market triggers and fills near 1.1300: the loss is 0.1200 per unit, 9.60% of entry, on a stop you placed 4.00% away. Nothing malfunctioned. There was simply nobody bidding in between.

Stop-limit: you get a price, or you get nothing

A stop-limit order carries two prices: a trigger, and a limit that the resulting order will not trade worse than. It solves slippage by refusing it, which means it can fail in the one way a stop must never fail - by not filling.

A stop-limit that does not fill

trigger           =  1.2000
limit             =  1.1950

price trades 1.1999 and keeps going to 1.1700
with nothing resting at 1.1950 or better

result            =  a live sell order at 1.1950,
                     a position still open at 1.1700,
                     and a loss of 6.40% and counting

The position you wanted out of at 4.00% is now open, unprotected, and being managed by an order sitting above the market. In a fast move it may never fill, and you are left deciding manually in exactly the conditions you wrote the stop to avoid deciding in.

Where the stop goes, and what it must never be

The defensible way to choose the price is to ask what would have to happen for the reason you entered to be wrong - a level reclaimed, a structure broken, a move larger than this coin normally makes - and put the stop past that. Then size the position to it. That order of operations matters, and the next lesson is entirely about it.

The indefensible way is the common one: decide how much you want to trade, discover the loss would be uncomfortable, and move the stop closer until the number looks acceptable. That does not reduce risk, it only moves the stop into the range price travels through routinely. The loss becomes smaller and much more frequent, and the arithmetic of module four is indifferent to which of those killed the account.

A stop you widen while losing is not a stop. A stop you never placed because you planned to watch the screen is not a stop either - it is an intention, and intentions do not work at four in the morning. If you are reading a signal from us or from anyone else, the stop is part of what the signal is; see how a signal is built.

The terms, defined

Stop: A price decided before entry, plus the rule that says what event at that price takes you out and with what order type. Both halves, or it is not a stop.
Stop distance: Entry minus stop, usually expressed as a percent of entry. The one number position sizing is built from.
On touch / on close: Two rules for the same price: exit when any trade prints through the level, or exit only if a bar finishes beyond it.
Stop-market: A trigger that sends a market order. It fills; the fill price is whatever the book holds at that moment.
Stop-limit: A trigger that sends a limit order with a worst acceptable price. It controls the price and can therefore fail to fill at all.
Slippage: The difference between the trigger price and the price you actually got. Largest in exactly the conditions that trigger stops.
Gap: Two consecutive trades at prices far apart, with nothing in between. No order type can fill inside a gap, because there were no trades there.

What this means in practice

Write both halves down: the price, and the rule about the price. If you cannot say the rule out loud, you do not have a stop yet.
Convert the stop to a percent of entry immediately. That percent is the input to everything else in this module.
Decide on touch or on close before you enter, and know how long you are unprotected under the rule you picked.
Expect a stop-market fill slightly worse than the trigger, and treat any plan that assumes an exact fill as understating its losses.
If you use a stop-limit, decide in advance what you do when it does not fill - because sometimes it will not.
Place the stop where the idea is wrong, then size to it. Never move the stop to make a position size comfortable.

Nothing in this lesson is financial advice or an instruction to trade. See the risk disclaimer. We never recommend a leverage setting - that decision is yours.