The same breakout bought on candle one or candle three, and the reward to risk each time

[Fairman]

The same breakout bought on candle one or candle three, and the reward to risk each time

EURUSD on M5. A small range sits between 1.0796 and 1.0802, and the previous swing high at 1.0820 is the obvious level overhead. Candle one closes at 1.0803, breaking the range. Candle two closes at 1.0808. Candle three closes at 1.0812. Three green bodies, 16 pips off the bottom of the range, and by now it feels like you're missing it.

Look at three possible entries with the stop at 1.0795, just under the range, which is where the idea is wrong.

Buy at 1.0803 on the break: risk 8 pips, room to 1.0820 is 17 pips. That's 2.1R.
Buy at 1.0806 on a small pullback after candle two: risk 11, room 14. That's 1.3R.
Buy at 1.0812 after candle three: risk 17, room 8. That's 0.47R.

Ignoring costs, the break-even win rates for those are about 32%, 44% and 68%.

You can tighten the stop on the late entry to get the ratio back, say 1.0805 for 7 pips of risk, but that's now inside the noise of candles that each ranged 4 or 5 pips. It gets stopped by ordinary breathing.

Position size follows the stop. With $100 at risk, the first entry is 1.25 lots and the third is 0.59 lots. If price does reach 1.0820, that's about $212 for the first and $47 for the third, from the same $100 of risk.

[FXS Prop Desk]

Fairman, the three entries on the same breakout show neatly how the reward to risk shrinks as confirmation increases. What makes it interesting is that the win rate likely moves in the other direction. Recording all three entries on each breakout for a while, as hypothetical trades, would show which combination of win rate and R gives the best expectancy for the pair.

[Fairman]

There is one thing the late entry does better. The spread is a smaller share of a 17 pip risk than of an 8 pip risk. With a 1 pip spread, that's about 6% of the stop on the third entry and 12.5% on the first. It's a real advantage on paper, but it's small next to the change in room. Going from 2.1R to 0.47R costs far more than the 6.5 points of spread share you gain. If the spread is the only thing in favor of an entry, you're probably already late. The lower cost share also shrinks quickly if the late entry has to use a tightened stop.

[Fairman]

How do you decide a move has already used up its room? One crude test is to compare the distance from the breakout level to the current price with a typical thrust for that pair and session. Take the last 20 breakout thrusts on your chart at your hours, measure each in pips, and find the median. If the median is 12 pips and this one has already gone 16, the third candle is probably late. I don't have a clean threshold, and I'd expect it to differ by pair and by hour, so I'm curious what others use. Is it a fixed pip number, a multiple of ATR, or just the next level on the chart?

[Fairman]

A way to get the first entry without staring at the screen is a buy stop placed before the break, at 1.0803, one pip above the 1.0802 range high. The cost is that false breaks fill you too. One detail: a buy stop triggers when the ask reaches the level, and on a bid chart that's earlier than it looks. With a 1 pip spread, a bid chart high of 1.0802 means the ask already touched 1.0803, so the order fires on a touch that looks, on the chart, like it only reached the range high. It's the same thing that makes short stops fire above the chart high. If your platform shows the ask line, you can see it directly.

[Fairman]

Pull the last 20 entries and measure the room to the next obvious level at the moment of entry, in units of R. Sort them into under 1R, 1R to 2R and over 2R, then compare the average result of each bucket. Twenty trades is a small sample, so don't read too much into the exact figures, but if the under 1R bucket is the one with the losses, that tells you where to put a minimum. A floor of 1.5R room is a number that's easy to write on a sticky note. It will also cut the number of trades, which is a cost of its own, so check how many entries it would have removed before adopting it.