Part 3: Trading it · Chapter 12 of 14 · 10 min
When an order block fails: lifespan, the break and streaks of stops
How long an order block lives, when it gives way, what happens after the break (one in three comes back inside) and how long a streak of stops can last, with three real gold examples and measurements on the engine's 149 zones: zones are nearly all disposable.
An order block is a zone that lasts a few days
Of the engine's 149 order blocks, 124 (83%) gave way before the end of the data: an H4 candle closed beyond the far side. Only 25 were still alive. The time from displacement to break has a median of 21 candles (3 and a half days). A quarter give way within 6 candles (a day), a quarter after 77 (about 13 days).
| After | Still alive |
|---|---|
| 5 candles (20 hours) | 81% |
| 10 candles (40 hours) | 69% |
| 30 candles (5 days) | 51% |
| 60 candles (10 days) | 39% |
The visit does not save the zone. Of the 114 visited zones, 104 gave way sooner or later: 17 in the first visit's candle, 44 within the next 10 candles, 43 later. Only 10 had not given way by the end of the data. Of the 35 zones not visited within 40 candles, 20 still gave way later.
Figure 1 shows a zone that holds and then gives way, on H4, from 11 to 17 January 2026. It is the chapter 1 order block (from $4,578.32 to 4,596.98). On 12 January at 20:00 price comes back (low 4,581.00, high 4,609.40) and closes at 4,600.86: the zone holds. On 16 January at 12:00 a candle opens at 4,614.18, drops to 4,540.03 and closes at 4,576.62, 1.70 dollars below the low edge (0.06 ATR, the ATR is 30.2 dollars): the zone has failed, four days after the visit. The next candle closes at 4,592.73: back inside the zone.
Read the chart in words
H4 gold chart from 11 to 17 January 2026. A blue band between 4,578.32 and 4,596.98 dollars is a demand order block. Price returns to the band on 12 January and climbs. On 16 January a candle with a long wick closes just below the low edge, marked as the zone failing, and the next candle closes inside again.
| When | What | Price |
|---|---|---|
| 12 Jan 2026 08:00 UTC | Demand order block | 4,578.32 – 4,596.98 |
| 12 Jan 2026 08:00 UTC | origin | 4,578.32 |
| 12 Jan 2026 12:00 UTC | displacement | 4,624.99 |
| 12 Jan 2026 20:00 UTC | first visit | 4,581.00 |
| 16 Jan 2026 12:00 UTC | closes below: zone fails | 4,540.03 |
Times in UTC.
After the break: does it continue or come back?
When a candle closes beyond the far side, we look at the next 10 candles: price continues by at least 1.5 ATR beyond the far side before coming back, or a close returns inside the zone.
- 79 (64%) continue, 37 (30%) come back with a close inside the zone, 7 do neither.
It depends on how wide the break is:
| The close goes beyond the far side by | Cases | Continue | Come back |
|---|---|---|---|
| Less than 0.25 ATR | 42 | 20 (48%) | 17 (40%) |
| 0.25 to 1 ATR | 50 | 28 (56%) | 20 (40%) |
| 1 ATR or more | 31 | 31 (100%) | 0 |
The last group is nearly a given: if the candle already closes 1 ATR beyond, reaching 1.5 ATR is easy and coming back is hard. The figure that counts is the first: four marginal breaks out of ten (under 0.25 ATR) come back. The 16 January break (0.06 ATR) was of that kind.
Figure 2 shows a clearer case, on H4, from 5 to 9 September 2026. The supply zone runs from $4,431.00 to 4,439.50 (8.50 dollars). On 7 September at 20:00 price brushes it (high 4,431.71) and closes at 4,427.19. On 8 September at 00:00 a candle climbs to 4,444.09 and closes at 4,441.75, 2.25 dollars above the high edge (0.12 ATR, the ATR is 18.5 dollars): the zone has failed. The next candle closes at 4,396.75, inside and below the zone, and five candles later the close is 4,353.77.
Read the chart in words
H4 gold chart from 5 to 9 September 2026. A grey band between 4,431.00 and 4,439.50 dollars is a supply order block. A candle closes just above the high edge, marked as closes 2.25 dollars above. The next candle closes inside the zone, marked as back inside, and price falls.
| When | What | Price |
|---|---|---|
| 6 Sept 2026 08:00 UTC | Supply order block | 4,431.00 – 4,439.50 |
| 6 Sept 2026 08:00 UTC | origin | 4,439.50 |
| 8 Sept 2026 00:00 UTC | closes 2.25 dollars above | 4,444.09 |
| 8 Sept 2026 04:00 UTC | back inside | 4,394.51 |
Times in UTC.
Figure 3 shows a break that continues, on H4, from 30 April to 5 May 2026. The demand zone (merged with an FVG) runs from $4,586.86 to 4,612.08. On 1 May at 16:00 price barely touches it (low 4,612.08) and closes at 4,614.00. On 4 May at 04:00 a candle from 4,617.86 to 4,581.64 has a body of 3.75 ATR (the ATR is 9.7 dollars) and closes 5.22 dollars below the low edge. The closes of the next five candles are 4,563.15, 4,524.47, 4,520.95, 4,521.66 and 4,536.90, and over the ten candles the low is 4,505.00.
Read the chart in words
H4 gold chart from 30 April to 5 May 2026. A blue band between 4,586.86 and 4,612.08 dollars is a demand order block. Price brushes it and stays above for a few candles. Then a very long bearish candle closes below the band, marked as closes below, and price falls.
| When | What | Price |
|---|---|---|
| 1 May 2026 04:00 UTC | Demand order block | 4,586.86 – 4,612.08 |
| 1 May 2026 04:00 UTC | origin | 4,580.65 |
| 1 May 2026 12:00 UTC | displacement | 4,661.28 |
| 1 May 2026 16:00 UTC | first visit | 4,612.08 |
| 4 May 2026 04:00 UTC | closes below: zone fails | 4,580.80 |
Times in UTC.
Signals on arrival
Two signals can be looked at when price reaches the zone.
- A strong candle against the zone: the visit candle has a body of at least 1 ATR in the direction against the zone (bearish for a demand). Of 51 such visits, 24 hold (47%) and 27 give way (53%). Of the other 63, 38 hold (60%) and 17 give way (27%). Careful: the figure includes the visit candle's own close, which with a big body is more likely to close already beyond the far side, and it is known only after the candle closes: it is of no use for a limit order already placed.
- An approach in a row: three or more consecutive candles against the zone before the visit. That is 10 visits: 4 hold (40%), 6 give way (60%), against 56% and 37% for the other 104. Too few to say anything.
The streak of stops
The 87 trades of chapter 9 (order at the near edge, 2R), put in order of entry, have 63 stops. The longest streak of consecutive stops is 9. There are 10 streaks of at least 3 stops and 5 of at least 5. With 1% risk per trade, 9 stops in a row take 8.6% off the account (0.99⁹ = 0.914). With 2% they take 16.6%. It is the same lesson as book 1 (chapter 13): set the risk per trade from the streak you can bear, not from the one you hope for.
A four-step routine
- Write in advance how a zone is declared failed: a close beyond the far side, with or without a margin (for example 0.25 ATR), and what you do if the next candle comes back.
- Give the order an expiry: after 30 candles (five days) the zone is still alive in 51% of cases. Reassess it or cancel the order.
- Tell the stop (a price, hit by a wick too) from the invalidation (a close): they are two different things and produce two different outcomes.
- Work out the streak of stops you can bear and choose the risk per trade accordingly: with 9 stops in a row, 1% costs 8.6% of the account.
Exercises
Chapter exercises
1. On the chart the demand zone held at the first visit (B). Which letter is the candle that declares it failed, meaning the first to close below the low edge?
Read the chart in words
| When | What | Price |
|---|---|---|
| 12 Jan 2026 08:00 UTC | Point A | 4,596.98 |
| 12 Jan 2026 20:00 UTC | Point B | 4,609.40 |
| 16 Jan 2026 12:00 UTC | Point C | 4,620.00 |
| 16 Jan 2026 20:00 UTC | Point D | 4,599.39 |
Times in UTC.
2. After a close beyond the far side by less than 0.25 ATR, 17 zones out of 42 came back with a close inside the zone in the next 10 candles. How do you use it?
3. After 60 candles only 39% of order blocks are still alive. What do you do with a limit order placed on a zone from 60 candles ago?
4. You have 9 stops in a row risking 2% per trade. About how much does the account lose?
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.