Part 3: Triangles, channels, wedges, flags and the rest · Chapter 8 of 14 · 6 min
Broadening formations: the pattern with no usable stop
The broadening formation (megaphone) has ever higher highs and ever lower lows: on H1 gold it shows up 11 times in 9.7 months, as many as in a random series, and only 2 can be tested with a stop within 6 ATR: the stop on the opposite line, already widened, is 3.6 ATR or more away.
The rule
A broadening formation (or megaphone) has five swings with the highs rising and the lows falling: the two lines diverge. The book asks for it like this (H1): highs rising by more than 0.5 ATR and lows falling by more than 0.5 ATR over the whole duration, with the final width above 125% of the initial one. The theory considers it a reversal pattern: the expected break is in the direction opposite to the move that preceded. Breakout: first close beyond one of the two lines (already widened) within 30 candles of the last swing. Trade: entry at the close, stop beyond the opposite line plus 0.25 ATR (risk 0.5-6 ATR), target equal to the initial width, exit after 48 candles, cost 0.30 $.
How many show up
On H1 gold the broadening formations are 11 (3 on H4, 0 on the daily chart), as many as in a random series (median 11, from 5 to 15). They last a median of 30 candles and the breakout comes 14 candles after the last swing: much later than the other patterns, because to break, price has to leave a range that has already widened. They break 4 times upward and 7 downward; in the expected direction (opposite to the previous move) it happens 4 times out of 11 (36%), as in random series (36%). After the break price travels 1.5 ATR in the break direction first in 5 cases out of 11 (50%, ±31 points) and in the opposite direction in 5.
The trade that is not there
The initial width is small (a median of 18.0 $, 0.9 ATR), but at the moment of the breakout the lines have widened, and the stop on the opposite line is far: 9 broadening formations out of 11 have a risk beyond 6 ATR and are skipped. Two trades remain, both at the target (+0.48 and +0.33 R, mean +0.41 ± 0.15 R), with a target worth a median of 0.48 times the risk: to break even 67% of targets would be needed. With 2 cases and 9 skipped there is no result to report: the pattern, with an operating rule with a stop on the opposite line, does not lend itself to a test.
Figure 1 shows one of the two trades: swings at 4,996.18, 4,880.00, 5,094.80, 4,855.01 and 5,027.80 $ (3-4 February 2026), downward breakout on 5 February at 02:00 below 4,843.11 $, entry at 4,814.40 $, stop at 5,047.03 $ (3.6 ATR, 232.63 $), target at 4,701.79 $ (width 112.61 $) reached at 23:00: +0.48 R. The risk is over twice the target: with the stop beyond the extreme, +0.48 R is the most the trade could give. Figure 2 shows one of the skipped broadening formations: swings at 4,315.60, 4,353.98, 4,324.16, 4,366.58 and 4,175.27 $ (8-10 June 2026), downward breakout on 10 June at 18:00; the stop on the opposite line was more than 6 ATR from the entry.
Read the chart in words
H1 gold chart from 3 to 6 February 2026 with the five swings numbered, the two diverging lines, the breakout and the entry, stop and target lines.
| When | What | Price |
|---|---|---|
| 3 Feb 2026 16:00 UTC | Point 1 | 4,996.18 |
| 3 Feb 2026 19:00 UTC | Point 2 | 4,880.00 |
| 4 Feb 2026 08:00 UTC | Point 3 | 5,094.80 |
| 4 Feb 2026 16:00 UTC | Point 4 | 4,855.01 |
| 4 Feb 2026 23:00 UTC | Point 5 | 5,027.80 |
| 5 Feb 2026 02:00 UTC | breakout down | 4,810.50 |
| 5 Feb 2026 02:00 UTC | Entry | 4,814.40 |
| 5 Feb 2026 02:00 UTC | Stop | 5,047.03 |
| 5 Feb 2026 02:00 UTC | Target | 4,701.79 |
| 5 Feb 2026 23:00 UTC | target reached | 4,700.00 |
Times in UTC.
Read the chart in words
H1 gold chart from 8 to 11 June 2026 with the five swings numbered, the two diverging lines and the breakout.
| When | What | Price |
|---|---|---|
| 8 Jun 2026 23:00 UTC | Point 1 | 4,315.60 |
| 9 Jun 2026 05:00 UTC | Point 2 | 4,353.98 |
| 9 Jun 2026 08:00 UTC | Point 3 | 4,324.16 |
| 9 Jun 2026 13:00 UTC | Point 4 | 4,366.58 |
| 10 Jun 2026 04:00 UTC | Point 5 | 4,175.27 |
| 10 Jun 2026 18:00 UTC | breakout down | 4,103.04 |
Times in UTC.
Procedure in four steps
- Write the rule: highs rising and lows falling over the whole duration, final width above 125%.
- Define stop on the opposite line and target, and check the ratio before counting.
- Count how many patterns are skipped for risk: if they are most of them, the trade does not exist.
- Do not report a mean R on 2 cases.
Exercises
Chapter exercises
1. On the chart the five swings of a broadening formation are marked with the letters A-E. Which letter is the highest high (the swing that widens the upper line)?
Read the chart in words
| When | What | Price |
|---|---|---|
| 3 Feb 2026 16:00 UTC | Point A | 4,996.18 |
| 3 Feb 2026 19:00 UTC | Point B | 4,880.00 |
| 4 Feb 2026 08:00 UTC | Point C | 5,094.80 |
| 4 Feb 2026 16:00 UTC | Point D | 4,855.01 |
| 4 Feb 2026 23:00 UTC | Point E | 5,027.80 |
| 5 Feb 2026 02:00 UTC | breakout down | 4,810.50 |
Times in UTC.
2. A broadening formation has an initial width of 20 (target) and the stop, on the opposite already-widened line, sits 100 from the entry (hypothetical numbers). What is the target worth in R?
3. 9 broadening formations out of 11 have the stop beyond 6 ATR and are skipped. What does it mean?
4. The 2 valid trades give +0.41 R on average. What should be concluded?
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.