Institutional Benchmark • Transparency-First • Audit-Friendly

GTIXT — Integrity Beacon for Global Prop Trading

The first public, versioned benchmark that converts messy prop firm reality into an institutional-grade signal: transparency, payout reliability, risk model integrity, and compliance.

📊 Explorer l'Index🔒 Vérification📖 Méthodologie
A • TransparencyB • Payout ReliabilityC • Risk ModelD • Legal & ComplianceE • Reputation & Support

Index Overview

Real-time benchmark metrics

Total Firms
Avg Score
Data Coverage
Pass Rate
NA Rate

Data coverage highlights how many core fields are verified in the latest snapshot.

Snapshot:
SHA-256:
Updated:
🔐 Verify SHA-256📈 View Ranking

Design Principles

GTIXT is built on four foundational principles that ensure institutional credibility and scientific rigor.

⚙️

Deterministic

Same inputs → same outputs. No discretionary overrides. Scoring is spec-driven, reproducible, and immune to manual intervention.

📋

Audit Trail

Every score is traceable to captured public evidence with timestamps. Inputs are stored immutably—enabling full historical reconstruction.

⚖️

NA-Neutral

Missing datapoints are scored at a neutral baseline (0.5) to avoid rewarding opacity. Transparency is incentivized, not penalized.

🔄

Versioned

Methodology evolves through explicit versions (v1.0, v1.1, etc.). Past scores remain reproducible under their original version.

Scoring Framework

GTIXT v1.0 measures operational quality across five pillars. Each pillar is normalized (0–1), then combined via weighted sum to produce a final score (0–100).

Master Formula
score = 100 × Σ (weight × pillar_score)

Where each pillar_score is normalized to [0, 1] and weights sum to 1.0

Normalization

Each metric is mapped to [0, 1] via score maps, bins, or boolean mappings. NA values default to 0.5 (neutral).

Aggregation

Pillar scores are computed as weighted averages of their constituent metrics. Sub-pillar weights are defined in the spec.

Fallback Chain

When primary inputs are missing, fallback hierarchy is applied. If all inputs are NA, pillar defaults to 0.5.

Data Collection & Scoring Agents

GTIXT v1.0 operates 8 specialized agents that collect evidence and compute metrics across the 5 pillars. Each agent is deterministic, auditable, and produces specific evidence types.

Agent A - Crawler

Fetches firm website HTML with Playwright/Puppeteer. Timestamps all evidence and stores raw snapshots in MinIO.

HTML snapshots: Firm URLs

Agent B - Extractor

Parses HTML to extract structured rule data, pricing, payout terms. Detects changes vs previous versions.

Extracted rulesPillars A, B, C

RVI - Registry Verification

License verification (FCA, CySEC, ASIC). Confirms regulatory registration status and permissions.

License verificationPillar D (Legal)

SSS - Sanctions Screening

Watchlist screening (OFAC, UN, EU, PEP lists). Detects sanctioned entities and compliance risks.

Watchlist matchPillar D (Legal)

REM - Regulatory Events

Monitors SEC, FCA, CySEC regulatory events. Tracks enforcement actions and compliance changes.

Regulatory eventsPillar D (Legal)

IRS - Review System

Processes manual submissions (complaints, disclosures). Validates evidence quality and relevance.

Submission verificationAll pillars

FRP - Reputation & Payout

Analyzes TrustPilot reviews, payout reliability, sentiment trends. Aggregates reputation signals.

3 types (reputation, payout, sentiment)Pillars B, E

MIS - Investigation System

WHOIS verification, company registration checks, news analysis, fraud pattern detection.

4 types (domain, company, news, patterns)Pillars D, E

Agent C - Integrity Gate (Q2 2026)

Validates data quality before publication. Enforces NA-rate controls, detects anomalies, and determines universe inclusion. Currently in development—will complete v1.0 release.

The 5 Pillars

Each pillar measures a distinct dimension of operational quality. Weights reflect institutional priorities.

Transparency

27%

Clarity and accessibility of rules, pricing, fees, and terms. Measures whether a firm provides sufficient public information for evaluation.

rules.clarityrules.accessiblefees.disclosedpricing.clear

Payout Reliability

27%

Structural consistency of payout logic and conditions. Evaluates whether payout rules are well-defined, explicit, and verifiable.

payout.structurepayout.conditionspayout.logic

Risk Model

21%

Quality of loss limits, drawdown logic, and risk model type. Measures whether risk parameters are clearly defined and institutionally sound.

risk.model.typeloss.limitsdrawdown.logic

Legal Compliance

15%

Jurisdiction risk tier, disclaimer quality, and regulatory exposure. Evaluates whether the firm operates in a credible legal framework.

jurisdiction.tierdisclaimer.qualityexposure.risk

Reputation & Support

10%

External signals and support quality proxies. Captures Trustpilot ratings, domain age, and operational longevity as secondary indicators.

trustpilot.scoredomain.agesupport.signals

Open Specification

The complete scoring specification is published as structured JSON. Every metric, weight, mapping, and fallback is machine-readable and auditable.

gpti.score.v1

📥 Download JSON
{
  "version": "gpti.score.v1",
  "description": "GTIXT scoring specification v1.0",
  "pillars": {
    "transparency": {
      "weight": 0.25,
      "metrics": {
        "rules.clarity": {
          "type": "score_map",
          "weight": 0.35,
          "mapping": {
            "0": 0.1, "25": 0.4, "50": 0.6,
            "75": 0.85, "100": 1.0
          }
        },
        "rules.accessible": {
          "type": "boolean",
          "weight": 0.30,
          "mapping": { "true": 1.0, "false": 0.0 }
        },
        "fees.disclosed": {
          "type": "boolean",
          "weight": 0.20,
          "mapping": { "true": 1.0, "false": 0.0 }
        }
      }
    },
    "payout_reliability": { "weight": 0.25 },
    "risk_model": { "weight": 0.20 },
    "legal_compliance": { "weight": 0.20 },
    "reputation_support": { "weight": 0.10 }
  },
  "na_policy": {
    "default_score": 0.5,
    "description": "Missing inputs default to neutral"
  }
}

Full spec includes all metrics, sub-weights, score maps, bins, fallback chains, and NA handling rules. Download the JSON for programmatic integration.

Versioning & Evolution

GTIXT methodology evolves through explicit semantic versions. Each version is immutable and backward-compatible for historical reproducibility.

v1.0

Initial Release

Released: December 2025

Foundational 5-pillar framework with 30+ metrics. Covers transparency, payout reliability, risk models, legal compliance, and reputation.

View Whitepaper →
v1.1 CURRENT

Institutional-Grade Verification

Released: February 24, 2026

Multi-level cryptographic hashing (5 levels: evidence → firm → pillar → dataset → ECDSA-secp256k1 signatures). Four institutional provenance endpoints for complete audit trails (/api/provenance/trace, graph, evidence, verify). Governance framework with advisory board. Fully backward compatible with v1.0.

🔐 View Institutional Features →
v1.2

Data Access & Dashboard

Planned: Q2-Q3 2026

Public snapshot browser with historical queries. Institutional dashboard for data exploration. Batch operations and advanced filtering. Webhooks and real-time streaming capabilities.

v2.0

Planned

Target: Q3 2026 (Q3 2026 - Planned)

Planned: SEC EDGAR integration, IRS agent enhancements, expanded legal compliance signals, and multi-jurisdiction risk modeling.

Versioning Policy

  • Major versions (v1 → v2): Breaking changes to pillar weights, metric definitions, or scoring logic.
  • Minor versions (v1.0 → v1.1): Refinements, new metrics, or enhanced data sources—backward compatible.
  • Historical reproducibility: All past snapshots include version metadata. Scores can be recomputed using original spec.
  • v1.1 Institutional Features: Multi-level hashing and ECDSA-secp256k1 signatures provide cryptographic non-repudiation without changing underlying scoring logic or pillar definitions. All v1.0 scores remain reproducible and verifiable.

🔐 Institutional Cryptographic Verification (v1.1)

Released February 24, 2026 — Multi-level hashing and ECDSA signatures for non-repudiation

⛓️

Multi-Level Hashing

SHA-256 cryptographic hashing at 5 levels: evidence (snapshot excerpts), firm (aggregated per firm), pillar (across metrics), dataset (complete snapshot), and ECDSA-secp256k1 signature at the top level. Creates complete hash chain for audit trail verification.

🔗

Provenance Endpoints

Four institutional endpoints for complete data lineage: /api/provenance/trace (hash chain), /api/provenance/graph (firm history), /api/provenance/evidence (with proofs), and /api/provenance/verify (ECDSA validation). Full backward compatibility with v1.0 API.

✍️

ECDSA-secp256k1 Signatures

Non-repudiation guarantee with ECDSA-secp256k1 elliptic curve cryptography. Detects any tampering with published snapshots. Signer identity and timestamp embedded in signature verification response. Institutional standard for regulatory compliance.

📋

Auditability

Complete transformation chain traceable. Evidence collection, validation, scoring, and snapshot generation steps all documented and verifiable. Supports multi-level audit trails for institutional compliance frameworks (SOC 2, ISO 27001, regulatory reporting).

Evidence Model

Every datapoint in GTIXT is backed by captured public evidence. Agents extract, validate, and link each input to its authoritative source.

🔍

Evidence Excerpts

Agents capture machine-readable excerpts from official sources (rules PDFs, terms pages, pricing disclosures). Each excerpt includes URI, timestamp, and extraction method.

🔗

URI Provenance

Every input links to its authoritative source URL. Auditors can verify that extracted data matches the official publication at the time of capture.

⛓️

Cryptographic Chain

Source → Excerpt → Snapshot → SHA-256 → Pointer. This chain ensures that every score can be traced back to its regulatory and public origin.

Evidence Validation Flow

1
Capture: Agents crawl firm websites and extract structured data from rules, terms, and pricing pages.
2
Validate: Agents check for missing fields, inconsistencies, and violations. Evidence is flagged if incomplete.
3
Score: Validated inputs are mapped through the scoring spec. NA values default to 0.5.
4
Snapshot: All inputs, scores, and evidence URIs are written to immutable snapshot JSON with SHA-256 hash.

Index Scope & Universe

GTIXT covers proprietary trading firms offering funded programs and challenge-style evaluations. Universe inclusion is rule-based and deterministic.

Universe States

  • candidate: Eligible to crawl & score. Meets minimum data requirements.
  • watchlist: Monitored, partial signals. Not yet candidate.
  • excluded: Fails exclusion rules or insufficient public footprint.

Data Requirements

  • Rules & Terms (or equivalent public disclosure)
  • Pricing / Fees information
  • Basic operational footprint (active website + identifiable entity)
  • Evidence URLs + capture timestamps

Integrity Gate (Agent C)

Only firms passing the Integrity Gate are eligible for public publication. The gate is deterministic, spec-driven, and explains its reasons. GTIXT is resilient against "silent" low-quality data and prevents publishing unreliable rankings.

Disclaimer: GTIXT does not provide financial advice, predict returns, or evaluate trader performance. It measures structural quality using publicly observable signals. Scores are not endorsements. Users should conduct independent due diligence.