Part 3: From theory to the account · Chapter 8 of 14 · 11 min

The stop and the ounces

The ounces to buy are the dollar risk divided by the stop distance: with 10,000 $ and 1%, a stop of 12.2 $ is worth 8.2 ounces and one of 43.3 $ is worth 2.3, and the risk stays 100 $. In our series the stop goes from 5 to 175 $ (median 21.7 $), so the quantity changes by almost four times from one trade to another. The stop, though, is an order, not a guarantee: in LA, in 2 stopped trades out of 276, the candle opened beyond the stop and the loss would have been about −2.6 R instead of −1 R.

The formula, with numbers

In chapter 5 we wrote the rule: ounces = capital × risk percentage ÷ stop distance. A trade on gold gains or loses, for every ounce, as much as the price moves in dollars: with 5 ounces and a 10 $ move the result is 50 $. So the dollar risk of a position is:

> risk = ounces × stop distance (in $ per ounce)

and the ounces follow from it. With a 10,000 $ account and a 1% risk (100 $):

Stop distanceOuncesLots of 100 ounces
10 $100.10
15 $6.70.067
20 $50.05
30 $3.30.033
50 $20.02
100 $10.01

At most CFD brokers a standard lot of gold is 100 ounces, but specifications vary (some brokers have lots of 10 ounces, others contracts of 1 ounce): check your platform before using the table. The principle does not change: halving the stop doubles the ounces.

The real stops of our series

The stop distance (the "risk" of each trade, in $ per ounce) in the three real series:

SeriesMinimum10th percentileMedian90th percentileMaximum
LA (552 trades)5.4 $11.4 $21.7 $43.9 $174.5 $
LD (367)5.1 $12.5 $20.5 $40.7 $154.6 $
SD (55)56.4 $61.3 $84.1 $142.3 $397.0 $

In LA the median stop (21.7 $) is 0.48% of the median price (4,495 $) and is worth 1.10 times the hourly ATR at entry. With 10,000 $ and 1% the corresponding ounces are:

SeriesStop at 10th percentileMedian stopStop at 90th percentile
LA8.8 ounces4.6 ounces2.3 ounces
LD8.0 ounces4.9 ounces2.5 ounces
SD1.6 ounces1.2 ounces0.7 ounces

Within the same series the quantity changes by almost four times (from 8.8 to 2.3 ounces in LA) and the dollar risk stays the same. This is the point of sizing: the quantity adapts to the stop, the risk stays fixed. Two real examples from LA, both closed in profit, with 10,000 $ and 1%:

Figure 1. Narrow stop: long entry at 4,534.18 $ (21 May 2026, UTC hours), stop at 4,521.95 $ (12.2 $), target at 4,546.41 $ reached at 22:00. With 100 $ of risk the ounces are 8.2; the result is +0.98 R, that is about +98 $.Data: gold, XAUUSDT perpetual future (real candles) · 1 hour (H1) · from 21 May 2026 06:00 UTC to 22 May 2026 17:00 UTC.
Read the chart in words

Hourly gold chart of 21-22 May 2026 with the entry, the close stop and the target of a long trade.

WhenWhatPrice
21 May 2026 18:00 UTCentry4,534.18
21 May 2026 18:00 UTCstop: −1 R (12.2 $)4,521.95
21 May 2026 18:00 UTCtarget: +1 R4,546.41

Times in UTC.

Figure 2. Wide stop: short entry at 5,086.55 $ (4 February 2026, UTC hours), stop at 5,129.81 $ (43.3 $), target at 5,043.29 $ reached at 11:00. With 100 $ of risk the ounces are 2.3; the result is +0.99 R, that is about +99 $.Data: gold, XAUUSDT perpetual future (real candles) · 1 hour (H1) · from 3 Feb 2026 19:00 UTC to 5 Feb 2026 06:00 UTC.
Read the chart in words

Hourly gold chart of 3-5 February 2026 with the entry, the far stop and the target of a short trade.

WhenWhatPrice
4 Feb 2026 07:00 UTCentry5,086.55
4 Feb 2026 07:00 UTCstop: −1 R (43.3 $)5,129.81
4 Feb 2026 07:00 UTCtarget: +1 R5,043.29

Times in UTC.

The narrower stop required a position of 8.2 ounces, the wider one 2.3. The risk is the same (100 $) and so is the gain (about 98 and 99 $): the stop distance changes the size, not the result in R. What changes is the weight of the cost: the 0.3 $ cost (chapter 10) is 0.025 R with the stop at 12.2 $ and 0.007 R with the one at 43.3 $.

The stop is chosen first

A typical mistake is to shorten the stop to be able to buy more ounces, or to lengthen it so as "not to be thrown out". The stop goes where the idea is wrong (chapter 12 of book 10 shows it with data), and the size is computed afterwards. If you feel like changing the stop to make the quantity work, the problem is the size, not the stop.

Two practical rules:

  1. Always round down. With 4.61 ounces computed, you buy 4.6: the risk is 99.8 $. With 4.7 the risk is 102 $, above the limit.
  2. Check the minimum size. Your broker's minimum lot (often 0.01 lots, that is 1 ounce on a 100-ounce lot) imposes a minimum account: account = minimum size × stop ÷ risk percentage.
StopRiskMinimum account with 1 ounceMinimum account with 0.1 ounces
15 $1%1,500 $150 $
22 $1%2,200 $220 $
30 $1%3,000 $300 $
22 $2%1,100 $110 $
22 $0.5%4,400 $440 $

With a 1,000 $ account, a 22 $ stop and a minimum size of 1 ounce, the minimum possible risk is 22 $, 2.2%: more than double the 1% you wanted. In that case the choice is another: a smaller minimum size (micro lots), a narrower stop if the idea allows it, or giving up the trade.

When the stop is not enough

The stop is an order. When the price reaches it, the order becomes a market order and is executed at the price available, which can be worse. In our series losses are worth between −1.01 and −1.06 R because we assumed execution at the stop level plus the 0.3 $ cost; reality can be worse, and we can partly measure it with the data.

  1. The hourly candle of gold can be wider than a whole stop. Range (high minus low) of the H1 candle in the data: median 12.8 $, 90th percentile 33.0 $, 99th 89.9 $. 24.0% of candles are wider than 21.7 $ (the median stop of LA) and 6.7% wider than 40 $.
  2. The extreme case. On 29 January 2026 at 15:00 UTC the hourly candle opened at 5,503.79 $, touched 5,504.06 $ and 5,112.55 $ and closed at 5,207.36 $: 391.5 $ in an hour, 18 times the median stop of LA.
Figure 3. The widest hourly candle in the data: 29 January 2026, 15:00 UTC, from 5,504.06 $ to 5,112.55 $ (391.5 $ in an hour). A stop of 21.7 $ would have been crossed many times; a position of 4.6 ounces would have lost, between high and low, about 1,800 $.Data: gold, XAUUSDT perpetual future (real candles) · 1 hour (H1) · from 29 Jan 2026 01:00 UTC to 30 Jan 2026 06:00 UTC.
Read the chart in words

Hourly gold chart of 29-30 January 2026 with a very long candle at 15:00 UTC and the values of high and low.

WhenWhatPrice
29 Jan 2026 15:00 UTChigh 5504.065,504.06
29 Jan 2026 15:00 UTClow 5112.555,112.55

Times in UTC.

  1. Candles that open beyond the stop. In LA there are 276 trades closed by the stop. In 2 (0.7%) the candle in which the stop is hit opens already beyond the level. If in those cases you execute at the open, the loss is on average −2.58 R (the worst −2.96 R) instead of about −1.02 R. On the series the sum goes from −18.2 R to −21.4 R: two events out of 276 cost 3.2 R.

A limit of the data: they are a perpetual future that trades continuously, without weekend closures, and there are no jumps between one close and the next open above 27 $. In gold CFDs, weekends and some news produce larger jumps, and the spread widens: the real frequency of losses beyond 1 R can be higher than measured. Some brokers offer a guaranteed stop for a fee.

What to make of it: the risk per trade you choose is the risk if all goes well. If in one case out of a hundred you lose 2-3 R, the true average risk is a little above the one chosen, and the worst drawdown can become deeper than the one estimated in chapter 4. It is one more reason to keep f small.

What to do with it

  1. Put the stop where the idea is wrong, then compute the size. Never the reverse.
  2. Compute the ounces = capital × risk ÷ stop distance, and round down.
  3. Check your broker's minimum size and the minimum account that follows; if it is not enough, give up, or narrow the stop only if the idea allows it.
  4. Do not use the same number of ounces for different stops: the dollar risk would change.
  5. Allow for the real loss exceeding 1 R, especially at night, on weekends and on news days.

Procedure in four steps

  1. Set the stop where the idea is wrong, and measure the distance in dollars per ounce.
  2. Compute the dollar risk (capital × percentage) and divide it by the distance: those are the ounces.
  3. Round down and check the broker's minimum size.
  4. Check that, with that size, the true risk is within the limit (and allow for the worst: 2-3 R in some cases).

Exercises

Chapter exercises

1. On the chart, a short trade enters at 4,931.83 $ with the stop at 4,956.76 $ and the target at 4,906.90 $ (18 February 2026, UTC hours). Which letter is the candle in which the stop is hit?

Chart exercise. Data: gold, XAUUSDT perpetual future (real candles) · 1 hour (H1) · from 17 Feb 2026 19:00 UTC to 19 Feb 2026 06:00 UTC.
Read the chart in words
WhenWhatPrice
18 Feb 2026 07:00 UTCentry4,931.83
18 Feb 2026 07:00 UTCstop: −1 R (24.9 $)4,956.76
18 Feb 2026 07:00 UTCtarget: +1 R4,906.90
18 Feb 2026 03:00 UTCPoint A4,922.39
18 Feb 2026 07:00 UTCPoint B4,942.50
18 Feb 2026 12:00 UTCPoint C4,948.40
18 Feb 2026 13:00 UTCPoint D4,970.95

Times in UTC.

2. A 10,000 $ account, 1% risk, stop at 24.9 $ per ounce (the trade on the chart). How many ounces do you sell?

3. The calculation gives 4.61 ounces. How many do you buy?

4. A 1,000 $ account, minimum size 1 ounce, stop at 22 $. What is the minimum possible risk?

Have a question or want to share your exercise?

Post in the community, or join the free signals room on Telegram.

Educational content, not financial advice. Trading involves risk.