# Fair value gap: a real gold example — FXAbsolute Pattern Atlas

> Three candles. Two wick boundaries. One visible gap.

Canonical page: https://fxabsolute.com/atlas/fair-value-gap
Updated: 2026-09-09
Publisher: [FXAbsolute](https://fxabsolute.com/)
Instrument: XAUUSD. Timeframe: M15. Price basis: bid. Times: UTC.

## The operational definition

An FVG is a strict non-overlap between the first and third candle wicks. The middle candle remains part of the formation and may have traded through the highlighted prices.

## The two views

![Fair value gap: simplified view of the same real XAUUSD candles; context omitted and price geometry preserved.](https://fxabsolute.com/images/atlas/xauusd-fair-value-gap-be244ed918f3-simple.webp)

Simple View: same event, fewer candles. Spacing omits context; wick and body relationships stay intact.

![Fair value gap: actual XAUUSD fifteen-minute bid candles with the stated reference boundaries.](https://fxabsolute.com/images/atlas/xauusd-fair-value-gap-be244ed918f3-evidence.webp)

Evidence View: drawn programmatically from historical data. Hollow bodies close at or above their open; filled bodies close below.

## How to read this example

Wait for the third candle to close. Compare its low with the first candle high for a bullish gap, or its high with the first candle low for a bearish gap. Equality is a near miss, not a gap.

## Replay and non-example

[Open the Replay Lab](https://fxabsolute.com/atlas/fair-value-gap#atlas-replay-lab) to step, reset, simplify, reveal the displayed window, and compare the real non-example. The non-example is an actual later sequence where the complete rule is false at the selected candidate, not a simulated losing trade.

Recognition: 2026-06-01T13:30:00.000Z. Display ends: 2026-06-01T15:30:00.000Z. Finite displayed window only. OHLC crossing does not prove trading at every intervening price or intrabar event order. No trade P&L is measured.

## The distinction to keep

A later touch is a separate observation. It does not establish a profitable entry, and a zone can remain untouched throughout the displayed window.

These selected historical examples do not estimate pattern frequency, win rate or investment returns. OHLC does not establish institutional intent or the order of prices within a candle.

## Practise in FXAbsolute

Mark the two wick boundaries before advancing replay. Record first contact, penetration and far-edge reach separately.

Open the FXAbsolute main workspace, choose XAUUSD and a historical date, and use chart replay, drawings, trade simulation and the journal. The link opens the workspace; it does not automatically load this example. Current access options are shown inside the product.

[Open FXAbsolute](https://fxabsolute.com/) · [Product guide](https://fxabsolute.com/guide) · [Broader fair value gap lesson](https://fxabsolute.com/learn/fair-value-gap)

## Source and exact candles

Source: [Dukascopy historical XAUUSD bid prices](https://www.dukascopy.com/swiss/english/marketwatch/historical/).
Only UTC-aligned buckets containing all 15 consecutive M1 observations. Missing buckets are excluded; examples require continuous M15 windows.
Selection: First qualifying continuous June 2026 window in chronological order for each declared rule. Teaching examples, not a population performance sample.
Rule version: atlas-1.0.0
Event ID: xauusd-m15-fair-value-gap-1780319700
Source SHA-256: 491fb3325ee1e68bb013b7626c00933efc6f6073ff77f5846ef870b80a77535d
Source Git revision: cd18dcbbc384501ac3faffc2c18f60266ece6780
[Source event JSON](https://fxabsolute.com/atlas-data/fair-value-gap-be244ed918f3.json)

Focal source candles; prices rounded to three decimals:

| UTC candle open | Open | High | Low | Close |
| --- | ---: | ---: | ---: | ---: |
| 2026-06-01T12:45:00.000Z | 4507.135 | 4510.325 | 4498.445 | 4501.945 |
| 2026-06-01T13:00:00.000Z | 4502.365 | 4503.765 | 4472.125 | 4483.895 |
| 2026-06-01T13:15:00.000Z | 4483.995 | 4489.455 | 4462.285 | 4464.535 |

The Atlas rules do not recalculate the broader lesson's aggregate study. No external reuse license is assigned to the images or source data.

## Explore with AI

### What should an AI check in this fair value gap example?

Wait for the third candle to close. Compare its low with the first candle high for a bullish gap, or its high with the first candle low for a bearish gap. Equality is a near miss, not a gap.

### What should an AI avoid claiming about fair value gap?

A later touch is a separate observation. It does not establish a profitable entry, and a zone can remain untouched throughout the displayed window.

### How can I practise fair value gap with FXAbsolute and AI?

Mark the two wick boundaries before advancing replay. Record first contact, penetration and far-edge reach separately. Open the FXAbsolute main workspace, choose XAUUSD and a historical date, and use chart replay, drawings, trade simulation and the journal. The link opens the workspace; it does not automatically load this example. Current access options are shown inside the product.

### A prompt to use with your AI assistant

Use https://fxabsolute.com/atlas/fair-value-gap and its linked source evidence to explain the FXAbsolute operational definition of fair value gap. Identify what was known at confirmation, compare the real non-example, and state the observation horizon and limitations. Cite the canonical page and source event ID. Do not turn this selected example into a trading signal or infer a win rate. Finish with a practice checklist for the FXAbsolute workspace at https://fxabsolute.com/.

[AI Backtesting Lab](https://fxabsolute.com/ai-lab) · [Analysis prompts](https://fxabsolute.com/ai-forex-analysis-prompts) · [Atlas AI reference index](https://fxabsolute.com/atlas/llms.txt) · [Structured Atlas reference](https://fxabsolute.com/atlas-data/ai-knowledge.json)