# Fib Clusters

A Support/Resistance & Levels concept (Fibonacci suite) in the LuxAlgo Library, with 1 indicator implementation.

## What are Fib Clusters?

Fib clusters are price zones where several independent Fibonacci measurements land on top of each other: a [retracement](https://www.luxalgo.com/library/concept/fib-retracement/) of one swing, an [extension](https://www.luxalgo.com/library/concept/fib-extension/) of another, a projection from a third, often taken from different pivots or different timeframes. Any single Fibonacci draw depends entirely on its anchors, which makes isolated levels easy to overfit. The confluence logic is a defense against that arbitrariness: when measurements from unrelated anchors agree on a narrow band, the level is less likely to be an artifact of one person's swing selection.

Confluence-based Fibonacci work has been taught for decades: Robert Fischer's 1990s books formalized multi-swing measurement grids, and Carolyn Boroden's trading work built a whole methodology on price zones where three or more relationships stack, popularizing the vocabulary most retail platforms now echo. The tooling followed, with scanners and toolkits automating what was originally a patient exercise in drawing every meaningful swing by hand.

The idea is baked into other frameworks. Harmonic traders' [potential reversal zone](https://www.luxalgo.com/library/concept/potential-reversal-zone/) is a Fib confluence zone by construction, the overlap of ratios measured from each leg of the pattern. The tighter the band and the more distinct the contributing draws, the more the zone is trusted; a loose scatter of levels is just noise with extra lines.

Computed versions make the definition explicit. Automated tools collect candidate levels from every qualifying swing, bin them within a tolerance, and score each bin by count, draw diversity, and timeframe weight, the Fibonacci-specific case of general [level clustering algorithms](https://www.luxalgo.com/library/concept/level-clustering-algorithms/). The same scoring logic extends to mixed confluence, where a Fib band gains further standing by coinciding with independent references such as [prior period levels](https://www.luxalgo.com/library/concept/prior-period-levels/), [floor pivots](https://www.luxalgo.com/library/concept/floor-pivots/), or a well-tested [S/R zone](https://www.luxalgo.com/library/concept/s-r-zone/).

## How to identify a Fib cluster on a chart

The exercise is measuring several real swings and looking for agreement, not decorating one swing with more ratios.

1. Select the meaningful swings: the last two or three significant legs on the trading timeframe, plus the dominant leg one timeframe up.
2. Measure each independently: retracements of the corrective legs, extensions and projections of the impulse legs, each from its own genuine pivots.
3. Mark where levels from different draws land within a tight tolerance of each other; two agreeing draws is minimum confluence, three or more from distinct anchors is proper confluence.
4. Respect width discipline: the zone spans the overlapping levels only, and a band wide enough to catch everything proves nothing.
5. Upgrade zones that also coincide with non-Fibonacci references, such as a [period open](https://www.luxalgo.com/library/concept/period-opens/), a pivot formula level, or a tested [support](https://www.luxalgo.com/library/concept/support-level/) or [resistance level](https://www.luxalgo.com/library/concept/resistance-level/).
6. Plan around the zone's edges: reaction entries, stops beyond the far edge, and invalidation on acceptance through it, per standard [level interaction rules](https://www.luxalgo.com/library/concept/level-interaction-rules/).

## How it's calculated

Price zones where Fibonacci levels drawn from several different swings coincide, read as stronger support or resistance than any single level.

```
1. Select M significant swings on the chart, each defined by a start price A_j and an end price B_j, for j = 1..M.
2. For each swing j, compute the retracement levels R_j(r) = B_j - r × (B_j - A_j) for each ratio r in {0.236, 0.382, 0.5, 0.618, 0.786}.
3. Optionally add extension and projection levels from the same swings to the pool, using ratios above 1 or a third anchor.
4. Pool every level from every swing into one sorted list.
5. Group levels lying within a tolerance w of each other; a group qualifies as a confluence zone when it holds at least k levels from at least two different swings.
6. Mark each zone as the band from its lowest to its highest grouped level; more levels in a tighter band rank the zone stronger.

  M: number of swings analyzed (user choice)
  j: swing index, 1..M
  A_j: start price of swing j
  B_j: end price of swing j
  r: retracement ratio (0.236, 0.382, 0.5, 0.618, 0.786)
  R_j(r): retracement level of swing j at ratio r
  w: grouping tolerance, in ticks or percent of price (platform specific)
  k: minimum coincident levels to form a zone (commonly 3)
```

Carolyn Boroden's confluence method looks for at least three coincident Fibonacci price relationships within a tight range.

There is no universal standard for swing selection, ratio sets, or grouping tolerance, so zones vary across implementations.

The retracement formula handles both swing directions because B_j - A_j keeps its sign.

## How traders use it

- To rank levels: traders sort candidate zones by how many independent draws agree within a tight band, trading the confluent zones and ignoring one-off levels.
- As target stacking: extensions and projections from several legs that converge at one area make a natural take-profit zone, since multiple measurement styles point to the same shelf.
- To frame risk around the zone rather than a line: entries, stops, and invalidation reference the zone's edges instead of any single ratio inside it.
- As an automation layer: confluence toolkits compute every qualifying draw and surface only the scored bands, which removes the anchor-shopping temptation that manual Fibonacci work invites.
- As cross-family confluence: a Fib band that overlaps a [Fibonacci pivot](https://www.luxalgo.com/library/concept/fibonacci-pivots/), a [Camarilla](https://www.luxalgo.com/library/concept/camarilla/) level, or a mapped [supply or demand zone](https://www.luxalgo.com/library/concept/supply-and-demand-zones/) is treated as a first-class level, two unrelated methods having voted for the same price.

## Fib Clusters vs related level constructions

- **Fib Retracement** (https://www.luxalgo.com/library/concept/fib-retracement/): A retracement is one swing's ratio grid, fully hostage to its two anchors. Clusters demand agreement across several independent draws, which is the difference between a hypothesis and a vote.
- **Level Clustering Algorithms** (https://www.luxalgo.com/library/concept/level-clustering-algorithms/): Clustering algorithms generalize the idea: bin and score levels from any source, Fibonacci or otherwise, into zones. Fib clusters are the special case where all the inputs are Fibonacci measurements.
- **Fibonacci Pivots** (https://www.luxalgo.com/library/concept/fibonacci-pivots/): Fibonacci pivots apply fixed ratios to the prior period's range through a formula, identical for everyone. Fib clusters emerge from measured swings someone chose, so they carry more information when honest and more bias when not.

## FAQ

### How many overlapping levels make a Fibonacci confluence zone?

There is no fixed rule. Two overlapping draws is the minimum for confluence; three or more from genuinely different anchors (different swings, ideally different timeframes) is where most practitioners start treating the zone as meaningful. Width matters as much as count: five levels scattered across a wide band say less than three packed into a tight one.

### Are Fibonacci clusters more reliable than a single retracement level?

The argument for them is sound: agreement across independent measurements reduces the chance that a level is an artifact of one arbitrary anchor choice. That makes clusters a better filter, not a guarantee. Price trades through confluent zones regularly, so clusters are best used to select which levels deserve attention, with entries still requiring confirmation from price behavior.

### Which ratios should be included when building clusters?

The standard working set: 38.2, 50, 61.8, and 78.6 percent retracements, with 127.2 and 161.8 percent for extensions and projections. Adding every exotic ratio guarantees more overlaps by pure arithmetic, which quietly destroys the method's meaning. Confluence is only informative if the contributing grid was disciplined enough that agreement is surprising.

### Do cross-timeframe clusters matter more?

Most practitioners weight them higher. A weekly retracement agreeing with a daily extension means two genuinely different populations of anchors point at one price, which is stronger evidence than two draws from neighboring swings on the same chart. The higher-timeframe contribution also tends to define the zone's structural importance, with the lower timeframe refining its edges.

### How wide should a confluence zone be?

Tight relative to how the instrument moves: a common discipline is a fraction of the recent average range, so the zone reads as one shelf rather than a region. If honest measurement produces only wide, loose agreement, the correct conclusion is that no confluence exists there. Widening the tolerance until levels agree is anchor-shopping by another name.

### How do Fib clusters relate to the harmonic potential reversal zone?

The PRZ is a Fib confluence zone with a pattern attached: harmonic setups define which legs are measured and which ratios must converge, so completing the pattern means arriving at a pre-defined confluence zone. Generic confluence work drops the pattern requirement and simply asks where independent measurements agree. The trading treatment, zone edges, confirmation, invalidation, is the same.

## Implementations in the Library

- Fibonacci Confluence Toolkit (LuxAlgo): https://www.luxalgo.com/library/indicator/fibonacci-confluence-toolkit/

## Related concepts

- Fib Retracement: https://www.luxalgo.com/library/concept/fib-retracement/
- Fib Extension: https://www.luxalgo.com/library/concept/fib-extension/
- Fib Geometry Tools: https://www.luxalgo.com/library/concept/fib-geometry-tools/
- Fib Projection: https://www.luxalgo.com/library/concept/fib-projection/
- Fib Time Tools: https://www.luxalgo.com/library/concept/fib-time-tools/
- Golden Pocket: https://www.luxalgo.com/library/concept/golden-pocket/

---

Source: https://www.luxalgo.com/library/concept/fib-clusters/ (LuxAlgo Library, the encyclopedia of trading & technical analysis). Free to use with attribution: https://www.luxalgo.com/library/license/