Part 3: Elliott waves · Chapter 10 of 14 · 8 min
Ratios between waves: Fibonacci inside the impulse
What ratios the waves of a valid impulse really have: wave 2 to wave 1 0.66, wave 3 to wave 1 1.90, wave 4 to wave 3 0.40, wave 5 to wave 1 0.95 (medians over 14 impulses). The classic ratios (wave 3 = 161.8% of wave 1) fall 1 time in 14, and those that seem to hold change with the swing threshold.
The expected ratios
The theory associates Fibonacci ratios with the waves of an impulse:
| Wave | Expected ratio |
|---|---|
| Wave 2 | 50% or 61.8% of wave 1 |
| Wave 3 | 161.8% of wave 1 (often more, if extended) |
| Wave 4 | 38.2% of wave 3 |
| Wave 5 | equal to wave 1, or 61.8% of wave 1 |
The book measures them on the valid impulses of chapter 7: windows of six H4 swings (legs of at least 2 ATR) that meet the three rules. There are 14, in gold from December 2025 to October 2026.
What is measured on the 14 impulses
| Ratio | Median | 25th to 75th percentile |
|---|---|---|
| Wave 2 / wave 1 | 0.66 | 0.42 - 0.79 |
| Wave 3 / wave 1 | 1.90 | 1.62 - 2.24 |
| Wave 4 / wave 3 | 0.40 | 0.24 - 0.54 |
| Wave 5 / wave 1 | 0.95 | 0.73 - 1.03 |
Wave 3 is on median almost double wave 1 (1.90), and 161.8% sits at the low limit of the central half: the classic target is exceeded by most impulses. Wave 5 is on median 0.95 of wave 1, and wave 4 is 40% of wave 3. These are numbers close to the classic ratios (0.382, 1), with a wide spread.
How often the ratios fall on the levels
We count how often each ratio falls within a band of the classic value and compare with the valid impulses of random paths (same leg sizes, shuffled order, 2,000 trials). Three swing size thresholds, with 14, 19 and 11 impulses:
| Classic ratio | 2 ATR (14 impulses) | 1.5 ATR (19) | 3 ATR (11) |
|---|---|---|---|
| Wave 2 = 50% or 61.8% of 1 (±0.05) | 1 (7%); chance 25% (10-41) | 5 (26%); chance 25% (12-39) | 4 (36%); chance 25% (6-50) |
| Wave 3 = 161.8% of 1 (±0.1) | 1 (7%); chance 8% (0-19) | 1 (5%); chance 8% (0-18) | 1 (9%); chance 8% (0-25) |
| Wave 4 = 38.2% of 3 (±0.05) | 3 (21%); chance 21% (7-37) | 6 (32%); chance 20% (8-33) | 6 (55%); chance 21% (0-44) |
| Wave 5 = 100% or 61.8% of 1 (±0.1) | 7 (50%); chance 23% (9-40) | 3 (16%); chance 23% (10-38) | 3 (27%); chance 23% (0-46) |
At 2 ATR wave 5 equals wave 1 or 61.8% in 7 impulses out of 14 (50%), above the 95th percentile of chance (40%). But with 1.5 ATR swings it is 3 impulses out of 19 (16%, below chance) and with 3 ATR 3 out of 11 (27%, inside). Wave 4 at 38.2% goes from 21% to 32% to 55% (above the 95th percentile of chance, 44%, only at 3 ATR). With four ratios, three thresholds and samples of 11 to 19, some cell beats chance by luck, and which cell does changes with the threshold. The impulses of the three thresholds overlap (they are the same moves), so the three columns are not three independent tests.
Figure 1 shows a downward impulse with wave 5 almost equal to wave 1: from 12 May 2026 at 00:00 (4,776.83 $) to 20 May at 00:00 (4,457.27 $), waves of 134.65, 86.41, 247.26, 110.38 and 134.44 $. Wave 5 is 99.8% of wave 1, wave 3 184%, wave 2 64%, wave 4 45% of wave 3. Figure 2 shows one with wave 3 at 161.8%: from 22 September at 16:00 (4,373.53 $) to 2 October at 00:00 (4,139.17 $), waves of 124.42, 69.57, 202.08, 110.22 and 87.65 $: wave 3 is 162% of wave 1, wave 2 56%, wave 4 55% of wave 3, wave 5 70%. Figure 3 shows one that respects no classic ratio: upward from 30 June at 00:00 (3,948.18 $) to 6 July at 20:00 (4,173.93 $), waves of 118.02, 103.45, 241.66, 72.96 and 42.48 $: wave 2 is 88%, wave 3 205%, wave 4 30% of wave 3, wave 5 36%.
Read the chart in words
H4 gold chart from 11 to 20 May 2026 with the broken line of the six swings numbered 0 to 5 and, on each wave, the length or the ratio to the reference wave.
| When | What | Price |
|---|---|---|
| 12 May 2026 00:00 UTC | Point 0 | 4,776.83 |
| 12 May 2026 12:00 UTC | Point 1 | 4,642.18 |
| 13 May 2026 00:00 UTC | Point 2 | 4,728.59 |
| 18 May 2026 00:00 UTC | Point 3 | 4,481.33 |
| 19 May 2026 00:00 UTC | Point 4 | 4,591.71 |
| 20 May 2026 00:00 UTC | Point 5 | 4,457.27 |
| 12 May 2026 08:00 UTC | 1: 135 $ | 4,709.51 |
| 12 May 2026 20:00 UTC | 2 = 64% of 1 | 4,685.39 |
| 15 May 2026 12:00 UTC | 3 = 184% of 1 | 4,604.96 |
| 18 May 2026 12:00 UTC | 4 = 45% of 3 | 4,536.52 |
| 19 May 2026 12:00 UTC | 5 = 100% of 1 | 4,524.49 |
Times in UTC.
Read the chart in words
H4 gold chart from 22 September to 2 October 2026 with the broken line of the six swings numbered 0 to 5 and, on each wave, the length or the ratio to the reference wave.
| When | What | Price |
|---|---|---|
| 22 Sept 2026 16:00 UTC | Point 0 | 4,373.53 |
| 24 Sept 2026 08:00 UTC | Point 1 | 4,249.11 |
| 25 Sept 2026 08:00 UTC | Point 2 | 4,318.68 |
| 28 Sept 2026 12:00 UTC | Point 3 | 4,116.60 |
| 30 Sept 2026 12:00 UTC | Point 4 | 4,226.82 |
| 2 Oct 2026 00:00 UTC | Point 5 | 4,139.17 |
| 23 Sept 2026 12:00 UTC | 1: 124 $ | 4,311.32 |
| 24 Sept 2026 20:00 UTC | 2 = 56% of 1 | 4,283.90 |
| 27 Sept 2026 00:00 UTC | 3 = 162% of 1 | 4,217.64 |
| 29 Sept 2026 12:00 UTC | 4 = 55% of 3 | 4,171.71 |
| 1 Oct 2026 08:00 UTC | 5 = 70% of 1 | 4,183.00 |
Times in UTC.
Read the chart in words
H4 gold chart from 29 June to 7 July 2026 with the broken line of the six swings numbered 0 to 5 and, on each wave, the length or the ratio to the reference wave.
| When | What | Price |
|---|---|---|
| 30 Jun 2026 00:00 UTC | Point 0 | 3,948.18 |
| 30 Jun 2026 12:00 UTC | Point 1 | 4,066.20 |
| 1 Jul 2026 04:00 UTC | Point 2 | 3,962.75 |
| 6 Jul 2026 00:00 UTC | Point 3 | 4,204.41 |
| 6 Jul 2026 12:00 UTC | Point 4 | 4,131.45 |
| 6 Jul 2026 20:00 UTC | Point 5 | 4,173.93 |
| 30 Jun 2026 08:00 UTC | 1: 118 $ | 4,007.19 |
| 30 Jun 2026 20:00 UTC | 2 = 88% of 1 | 4,014.48 |
| 3 Jul 2026 16:00 UTC | 3 = 205% of 1 | 4,083.58 |
| 6 Jul 2026 08:00 UTC | 4 = 30% of 3 | 4,167.93 |
| 6 Jul 2026 16:00 UTC | 5 = 36% of 1 | 4,152.69 |
Times in UTC.
Procedure in four steps
- Declare the expected ratios and the bands before looking at the data.
- Measure all the ratios of the valid impulses and give the median and the percentiles, not only the hits.
- Compare with the valid impulses of random paths with the same sizes.
- Repeat with another swing threshold: a ratio that holds at one threshold only is not a result.
Exercises
Chapter exercises
1. Wave 1 measures 120 $ and wave 3 measures 240 $ (hypothetical numbers). What is wave 3 as a percentage of wave 1?
2. Wave 5 equals wave 1 or 61.8% in 7 impulses out of 14 at 2 ATR, in 3 out of 19 at 1.5 ATR and in 3 out of 11 at 3 ATR. What does it say?
3. Wave 3 falls on 161.8% of wave 1 in 1 impulse out of 14, while chance does it in 8% of valid windows. What do you conclude?
4. The median of wave 3 over wave 1 is 1.90 (25th-75th percentile 1.62-2.24). What does it mean for a target at 161.8%?
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.