Skip to main content
Track Fibonacci projection
1 / 18
Library

2001issue C011-5

Constructing a Gartley from wave retracements

A gartley-construction is assembled in stages from an impulse-leg and a corrective-sequence. Each later swing is admitted only when its Fibonacci retracement or projection lands inside a stated price-range band. Editorial reading: treat a completed D as a hypothesis only when three independent ratio estimates form a d-cluster.

  • A gartley-construction does not exist until A reverses, and point-x is chosen so the impulse-leg XA can still produce a valid B.
  • Admit AB, BC, and CD only after each Fibonacci retracement or projection lands inside its stated price-range band.
  • D is the completion trigger when the d-cluster from the cd-band, the ab-equals-cd copy, and the xd-ideal comes together.
  • Only price-range ratios decide whether the wave remains valid. Chart angles are time artifacts, and a pullback beyond 78.6% of XA is treated as a failed construction.
Entries in this reading3 entries

Build the Gartley in stages

A gartley-construction is a five-point XA-AB-BC-CD assembly. It becomes a long or short completion only after each required ratio band is met. The elliott-wave labels are the working set for the impulse-plus-correction sequence, so each new pivot can be checked before the next ratio is applied.

The impulse-leg is the opening directional swing XA. Later Fibonacci retracement measures, and the overall XD measure, are taken against that range. The corrective-sequence is the alternating AB, BC, and CD reversals that complete the five-point wave after the impulse.

There is no pattern until A reverses

There is no pattern until A reverses. Point-x is a prior pivot chosen among earlier swing lows or highs so XA can later produce a valid B. B itself is not locked until price actually reverses there.

After B, continue if the AB/XA Fibonacci retracement falls near 57% to 67%, or at worst no deeper than 78.6% of XA. A pullback beyond 78.6% of XA is treated as a failed construction.

Admit later swings only inside the bands

The construction uses only four Fibonacci ratios: the successive-number limit 0.61804, its reciprocal 1.61804, and their square roots 0.786 and 1.27.

After the impulse-leg XA, AB is ideally 61.8% of XA. That first pullback is the ab-band: ideally near 0.618 of XA and still acceptable down to 0.786 of XA. BC is 61.8% to 78.6% of AB. That rebound is the bc-band, and it is measured in price, not line length. CD is a 127% to 161.8% Fibonacci projection of BC, which is the cd-band.

A Fibonacci retracement is a measured pullback of one completed price range. It is used to test AB versus XA, BC versus AB, and XD versus XA. A Fibonacci projection is an extension of a completed range. It is used to place the expected CD band from BC and the initial-target-dt from CD.

Treat D as a cluster of estimates

D is the completion trigger. The first DT objective is 61.8% of CD; that print is the initial-target-dt. Overall XD is ideally 78.6% of XA, which is the xd-ideal, the overall A-to-D pullback against range XA. CD is ideally equal to AB, which is the ab-equals-cd check.

The d-cluster is the group of independent D estimates from the CD projection band, the ab-equals-cd copy, and the 0.786 XA retracement.

A worked bullish completion

In the worked bullish example, XA is 60, AB is 37 (ratio 0.616), and BC is 28 (ratio 0.756 of AB). Both retracements stay inside their bands.

That example then clusters three D estimates at 67.84, 69, and 70.44. Actual D is 68.8, AD/XA is 0.77, CD/BC is 1.328, and the first T prints at 91.79.

Bearish inversion and simpler arithmetic

The same ratio stack is inverted for a bearish completion. That bearish-inversion applies the bands to a downward impulse and produces a short completion at D.

Arithmetic may be simplified with fifty-cent-rounding. The shortcut drops cents and replaces 61.8%, 78.6%, and 161.8% with 62%, 79%, and 162%. The archive also calls this the 50-cent rule.

Judge validity by price-range ratios

No single absolute retracement or fixed go/no-go tolerance is given. Chart angles are time artifacts, and only price-range ratios decide whether the wave remains valid.

Educational research material, not investment advice. Historical source context does not establish present-day performance.
1 of 18 in the Fibonacci projection track
20011-2 pp.Next on Fibonacci projectionConstructing butterfly-spreads from a weekly impulse to daily DConstruction starts by marking an experimental XA on a weekly chart, then switches to a daily chart so later Fibonacci-retracement and Fibonacci-projection measurements can be placed more precisely.
All readings on this track · 18 readings
  1. 2001Constructing a Gartley from wave retracements
  2. 2001Constructing butterfly-spreads from a weekly impulse to daily D
  3. 2002Projected Fibonacci targets from breakout levels
  4. 2004Constructing wave-ratio clusters with Fibonacci and Gann
  5. 2006Third-wave impulse counts wait for a dated cluster
  6. 2006Wave and ratio checkpoints versus a dominant dollar story
  7. 2007A 2007 EUR/JPY news shock as a Fibonacci retest drill
  8. 2008Constructing Fibonacci retracement, projection, and range maps
  9. 2010Swing projection targets and support-backed put sales
  10. 2010Fibonacci clusters as a tight support or resistance decision
  11. 2011Fibonacci projections and money-flow divergence on a Dollar Index downswing
  12. 2011Crude oil wave confluence as a decision map
  13. 2016Box range projections for cup breakout entries
  14. 2018Fibonacci pinball and the 2016 election-week scoring drill
  15. 2019Named line-break rules for a wave count and Fibonacci map
  16. 2019Objective Fibonacci grids for support, retracement, and projection
  17. 2020Gold Super Cycle nested waves and Fibonacci bands
  18. 2020Intra-swing Fibonacci fitting and completion targets
All 19 readings tagged Fibonacci projection
Also on Fibonacci projection5 readings