How is a Fibonacci fan different from Fibonacci retracement?

A fan ray changes price with time because it slopes from an anchor. A horizontal retracement level keeps the same price until its anchors change. The two can use similar fractions but test different conditions.

Which swing should I use for a Fibonacci fan?

Use a completed swing selected by a rule you can state before reviewing later bars. Record whether you use wicks or closes and how many bars confirm the endpoint. There is no uniquely correct pair of anchors for every chart.

Why is my TradingView fan pointing the wrong way?

Check the Reverse option in the drawing settings. TradingView documents it as the control for changing the fan direction. Also verify the two coordinates and ratio labels before changing the selected swing.

Should I draw Fibonacci fans on a log chart?

Choose the calculation before testing the fan. Linear geometry measures price differences; ordinary log geometry measures ratios between positive prices. Inspect the actual drawing after a scale change because display mode alone does not establish the formula used.

Does touching a Fibonacci fan mean price will reverse?

No. A touch records a defined intersection. Price can continue through the ray, fail on the same bar or never return to it. Keep those outcomes in the sample instead of retaining only attractive rebounds.

TradingView workflows

Fibonacci fans on TradingView: anchor rules and scale effects

A Fibonacci fan projects sloping rays from a selected swing toward proportional divisions of the move to a second swing. Each ray has a different price at each later bar. TradingView calls its drawing Fib Speed Resistance Fan and includes price and time levels. Freeze the two anchors, direction and scale before evaluating a touch; changing those choices after a break changes the hypothesis.

See the decision guide

Set the two anchors before judging the fan

Select Fib Speed Resistance Fan from the drawing toolbar. Place A at the start of a chosen swing and B at its later endpoint. Use the Coordinates settings to verify both bar positions and prices. In Style, record the enabled Price/Time Levels and Reverse setting. TradingView allows custom ratios, so a screenshot without those settings can leave the construction ambiguous.

Choose wick extremes or closing prices in advance and use the same convention for both anchors. Keep the symbol, interval, session and data adjustments in the record.
A swing high is not confirmed simply because it is the highest visible bar. If your rule needs two later bars to confirm B, begin evaluating the fan only after those bars have closed.
Keep A earlier than B. If the fan faces the wrong direction, inspect Reverse before swapping anchors or selecting a different move.
Start with the price-side rays needed for the study. The time-side divisions in TradingView are a separate construction; do not treat every displayed line as the same price retracement.

Calculate a price-side ray from one fixed swing

Use hypothetical anchors A at bar 0, price 100, and B at bar 10, price 200. For this example, r means the fraction of the upward price change retained above A at the time of B. A linear price-side ray is F(t) = 100 + r × (200 - 100) × t / 10. It starts at A and extends through the selected fractional height at bar 10.

A retracement measured downward from B uses the complementary fraction: a 61.8% retracement leaves 38.2% of the rise above A. Check the tool label convention and Reverse state before matching a displayed 0.618 line to this formula.
The 0.500 level is a common midpoint convention; it is not a ratio derived from adjacent Fibonacci numbers.
All these rays start at A, but only r = 1 passes through B itself. The 0.500 ray reaches 200 at bar 20, which shows why a fan level cannot be treated as a fixed horizontal price.
For a falling swing, use the signed difference B price minus A price. The same linear formula produces descending rays. A screenshot angle alone is not a rate because chart dimensions and zoom change its appearance.

Retained fraction r	Price at bar 10	Price at bar 15
0.382	138.20	157.30
0.500	150.00	175.00
0.618	161.80	192.70

Compare linear and log constructions with identical anchors

A straight ray in ordinary log-price coordinates follows G(t) = 100 × (200 / 100)^(r × t / 10). This is a defined mathematical comparison, not a claim that every TradingView drawing automatically uses this formula after a scale switch. With r = 0.500, the anchors and fraction stay fixed while the intermediate prices differ.

At bar 15, a price of 172 is below the linear ray and above the log-coordinate ray. Choosing a scale after seeing this bar can reverse the apparent conclusion.
For a chart comparison, save the Regular view, switch only the price axis to Logarithmic, and inspect the same drawing at the same later bar. Record both observed ray prices or mark a value as unreadable; do not fill it with the theoretical table.
Keep the same date range and chart dimensions for screenshots. Distinguish a change in screen angle from a change in the ray price at a fixed bar.
The official fan documentation describes coordinates and ratios but does not specify a log calculation switch. Verify the actual drawing against known points before using its visible intersection as a numeric condition.

Bar t	Linear F(t)	Log-coordinate G(t)
5	125.0000	118.9207
10	150.0000	141.4214
15	175.0000	168.1793
20	200.0000	200.0000

Record a touch separately from failed support

For an illustrative rising-fan study, evaluate each completed bar against the selected ray value at that bar. Define a range touch as low ≤ ray ≤ high. Define failed support as close < ray - 1 price unit, then retire that hypothesis at the first failure. The one-unit buffer is an example study choice, not a recommended stop. At bar 15, the two candidate rays are 175 and approximately 168.1793.

These are separate hypothetical cases, not a historical sequence or a measured success rate. A bar can fail without touching the ray if its entire range is below it.
A bar range containing the sampled ray price is an OHLC classification. It does not prove when an intrabar crossing occurred or that an order could have filled at that price.
For a descending resistance study, specify the mirrored rule: a completed close strictly above the ray plus the buffer. Do not reuse the rising-support inequality.
Log every activated fan, including immediate failures and fans with no touch before a predefined observation window ends. A later rebound does not erase an earlier recorded failure.

Independent bar-15 case	Linear ray: 175	Log ray: 168.1793
Low 174, high 178, close 176	Touch; no failure	No touch; no failure
Low 169, high 171, close 169	No touch; failed support	No touch; no failure
Low 166, high 170, close 166	No touch; failed support	Touch; failed support
Low 173, high 176, close 174	Touch; no failure at the exact 174 threshold	No touch; no failure

Reject unusable geometry and hindsight

The generic linear formula is A price + r × (B price - A price) × (t - A bar) / (B bar - A bar). Require a positive bar span. The log version replaces the signed price difference with ln(B price / A price), then transforms the result back to price. Both anchor prices must be strictly positive for that ordinary log model.

Identical anchor bars cause division by zero. Equal anchor prices collapse all price-side rays onto the same horizontal line; this does not describe a directional swing.
Reject zero or negative anchor prices for the log model. A platform display around zero does not make ln(price) valid for those values.
Missing bars, changed sessions or a different interval change the horizontal measurement. These examples count bars, not elapsed calendar days; weekends do not add bars.
A new extreme or a changed scale can justify a new study, but give it a new record. Keep the original anchors and failure visible in the old record.
TradingView warns that drawing alerts on logarithmic scales may not match the drawing position. Do not assume a visible fan offers a usable alert condition, or that its alert behavior matches a custom script formula.

Turn the frozen rule into a chart aid and review log

Give Pineify Pine Script AI Coding Agent the two anchor coordinates, confirmation time, ratio convention, calculation model and completed-bar failure rule. Ask it to display the computed ray price and the touch and failure flags. Check the generated source against the bar-15 cases before creating an alert; a script needs explicit anchor inputs or a defined selection rule.

Keep signal availability separate from the anchor timestamp. A label drawn back at B must not imply that B was already known on that bar.
In Trading Journal, save the hypothesis in a plan and add later observations in diary notes. Save screenshots separately and include their share links or file references in those notes.
Use one record per frozen fan: symbol, interval, session, scale, A/B coordinates, confirmation time, enabled ratios, Reverse state, buffer, expiry bar, first touch and first failure. Add both scale screenshots without replacing the original.
Compare the total number of activated fans with touches, failures and expired untouched cases. Assess trading performance only after specifying entries, exits, position sizes and costs; geometry alone supplies none of them.

Where Pineify fits

Pine Script AI Coding Agent

Draft an editable chart aid from explicit fan coordinates and event rules, then compare its output with known prices before using alerts.

Also useful: Trading Journal. Keep the original fan hypothesis, screenshot references and review notes, including invalidated and untouched cases.

Draft a Fibonacci fan chart aid

Related resources

TradingView workflows Log versus linear charts Record indicator settings Trading Journal

Frequently asked questions

Educational information, not investment advice. Prices, ray formulas and event rules are hypothetical illustrations, not measured TradingView drawing results, live signals or evidence of profitable trading.

Sources and verification

Last verified: 2026-10-17

Published by Pineify. Pineify is an independent product and is not affiliated with TradingView. Sources support the technical explanations; examples do not establish trading performance.