Auto-generated. This article is rebuilt from app/signals/config/signal_definitions.json by scripts/build_signals_kb.py. Edit the registry entry and re-run the script — do not edit this file directly.

GEX

What it is

GEX — registry key gex.

Classifier metric. See the bands table below for the band-by-band reading.

Data-pyramid tier

T1 — Signal. This metric sits at the signal layer of the platform’s six-tier data pyramid (T0 raw inputs → T5 actionable read). The tier reflects how far the value is from a raw measurement — not how strongly it is validated. Abstraction and validation are separate axes: a higher tier is not a claim of stronger evidence.

Horizon & validation

Horizon: 10-20d — the window this metric’s measured skill speaks to. A horizon stamp is an evidence claim, not a trading signal.

Evidence: Drawdown-lens gamma leg (via net_gex_rank 252d percentile): 15y re-validation (U2 2026-07-13) non-OL rank-IC +0.10/+0.11 vs forward 10-20d max-drawdown (halved from the two-episode discovery +0.25); quintile tail-lift ~5-6x monotone, sign transfers to 2011-2020; leg weight re-sized 2x -> 1.5x. INDEX LANE only - all25 ~1.0x (R7).

Source

Source module: darkpool
Data source: back_fillable

Fetched directly from app/sources/darkpool.py; see source code for the upstream API call and any provider-specific handling.

How it’s computed

Raw aggregate dealer gamma exposure from the dealer-gamma analytics daily CSV (same pull as dix), normalised to $B notional per 1% SPX move. Positive = dealers are net long gamma (dampening realised vol); negative = net short (amplifying moves). The fetch divides the raw column by 1e9 when the value exceeds 1e6 (app/sources/darkpool.py:31). Daily cadence. See app/sources/darkpool.py:28 (_fetch_dix_gex). The classifier bands (Strong suppression / Vol dampening / Mild amplification / Vol amplification / Severe amplification) are applied downstream via signal_defs.band_for('gex', value).

Where it surfaces

• API field: signals.gex on GET /api/v1/signals/latest
• Surface: daily_signals row (daily-cadence persistence)
• Surface: intraday_signals row (per-cycle snapshot)
• Surface: /signals/sparklines (UI sparklines)

Bands / thresholds

Classifier direction: higher_is_better.

Health-score / alignment role

• Health score component: gex (weight + math in app/signals/config/score_weights.json and app/signals/score.py).
• Alignment category: gamma, default weight 1.5 (see /kb/alignment for the 11-category framework).
• Contributes to: health_score.GEX, implication.gamma, alignment.gamma, transition.gex_sign

Release cadence

• Publishes: daily
• Approximate time (ET): 16:30
• Lag days: 0

_Shares the dealer-gamma analytics EOD CSV publish with DIX; during the live session the most recent row is the prior close. Mapped to freshness_class=prior_close by routes.PUBLISH_TO_FRESHNESS_CLASS.

See also

• /kb/gex — narrative-style article on this metric.
• /kb/alignment — Alignment framework — 11-category implication vs. price.
• /kb/health-score — Health score — 0-100 composite.
• /kb/api/get-signals-latest — API endpoint that serves this field.