Gap-Fill Confluence:
Turning the Morning Gap Into a Scored Setup

The gap-fill is one of the oldest patterns in intraday trading: price opens away from the prior close, then “fills the gap” by trading back toward it. It works often enough to be seductive and fails often enough to be dangerous. The problem was never the idea — it was the execution.

Most traders make one of three mistakes:

• They fade the gap immediately at the open, with no level and no confirmation — catching a falling knife.
• They wait for the “full fill” and enter as price closes the gap, precisely where momentum reverses on them.
• They ignore the regime entirely, fading a gap on a Trend Day that has no intention of coming back.

A gap is an overnight repricing. Between yesterday's close and today's open, futures moved on news, flows, or global sessions — and the cash index opens somewhere new. The engine measures that gap in two dimensions the moment the session begins:

• Direction — gap-up (open above prior close) or gap-down (open below).
• Size relative to ADR — the gap as a fraction of the average daily range, not just raw points. A 10-point gap means something very different on a 30-point day than on an 80-point day.

The engine only arms itself when the gap is meaningful — at least 10% of ADR. Below that, there is no gap-fill edge worth trading, and Gap-Fill Confluence stays in its resting no-gap state.

Note the direction carefully: the bias is with the gap, not against it. The retrace is the pullback you wait for; the trade is the resumption after price rejects the confluence pocket.

Gap-Fill Confluence is not a static indicator. It is a state machine that walks the session through five defined stages, and it will not hand you an actionable read until the sequence completes in order. This is what keeps it from firing prematurely.

A single support level is a coincidence. Five levels landing in the same narrow band is a decision zone where institutional order flow concentrates. The pocket is the heart of the engine.

SPXXL builds the pocket as a band — max(8% of ADR, 1.5 points) wide — centered on the 2-Day Anchored VWAP, then counts how many of the following independent levels fall inside it:

• 2-Day Anchored VWAP — the two-session volume-weighted fair value anchor.
• Session VWAP — today's developing fair value.
• Prior day close — the gap's origin and a heavily-watched reference.
• Initial Balance edge — the high or low of the first hour's range.
• Initial Balance midpoint — the rotational pivot of the opening range.
• Nearest magnet — the closest structural draw (round number, prior swing, key level).

The more levels that stack inside the band, the higher the level-stacking contribution to the score. A pocket holding four or five aligned levels is a genuinely rare, high-conviction location — and the engine rewards it accordingly.

Why anchor the pocket to a two-day VWAP specifically? Because it captures the volume-weighted consensus price across the gap itself. A single-session VWAP resets every morning and can be distorted by the opening imbalance. The 2-Day Anchored VWAP spans yesterday and today, so it “remembers” where real size traded on both sides of the gap.

The 2-Day Anchored VWAP and the 9EMA are both drawn live on the SPXXL chart, so you can see the pocket forming in real time rather than reconstructing it after the fact.

Arriving at the pocket is necessary but not sufficient. Price has to reject it — and the market has to participate in that rejection. The engine requires two confirmations before it moves to the active stage:

This dual gate is what separates a Gap-Fill Confluence read from a naked support bounce. A wick without volume is a guess; volume without a wick is noise. The engine waits for both to align in the pocket.

Every input above is fused into a single, fully transparent Confluence Score from 0 to 100. There is no black box: the score is a weighted sum of seven factors, and the engine shows you exactly how many points each one contributed and why.

The composite score maps to a letter grade so you can read conviction at a glance:

When a setup goes active, the engine projects a target ladder — a sequence of objective, structure-based levels to scale out against rather than a single arbitrary price:

Each rung is a level the market already respects — not a fixed R-multiple. That keeps the exit plan anchored to structure the same way the entry is.

A setup is only as disciplined as its exit-when-wrong. Every active Gap-Fill Confluence read publishes an explicit invalidation level: a price beyond the pocket and the rejecting wick where the thesis is simply no longer true. If price closes through it, the confluence has failed and the read is void.

The single biggest reason gap-fills fail is regime. A gap-fill is a mean-reversion trade, and mean reversion only pays in the right environment. This is why the engine folds SPXXL's session classification directly into the score as the regime-agreement factor.

On a Trend Day, the retrace you were waiting to fade is just a shallow pullback before continuation — the exact scenario that turns a gap-fade into a loss. By penalizing those regimes, the engine actively discourages you from fading a market that has no intention of reverting.

This is the line SPXXL will never cross. Gap-Fill Confluence is decision support and education — never a signal or financial advice. The engine does not tell you to buy or sell. It quantifies a setup, shows its full reasoning, and hands you the score, the levels, the targets, and the invalidation so you can make an informed decision.