Skip to content
Platform · Engineering Brief

The 47-Factor Engine Behind a 94.2% Audited Win Rate.

A precision-tuned algorithmic trading platform engineered in London: a proprietary 47-factor momentum model trained on 11 years of OHLCV data, paired with sub-380ms signal-to-execution across the four major retail venue bridges.

  • 94.2%audited win rate
    QuantVerify Labs 2023–25
  • <380mssignal-to-execution
    Binance · Kraken · Coinbase · MT5
  • 187,400+verified traders
    across 142 countries
Anatomy of the 47-Factor Model

Six signal families, one execution engine.

Every signal that reaches your terminal is the output of 47 weighted inputs grouped into six families — back-tested across 1,800+ asset pairs and validated quarterly by QuantVerify Labs since 2023.

  1. 01

    Price Action & Microstructure

    Order-book imbalance, trade flow toxicity, realised volatility regimes, and tick-level liquidity gaps across spot books — the fastest-decaying inputs, re-weighted every 250ms.

    • 11 factors
    • 250ms refresh
  2. 02

    Momentum & Trend Persistence

    Multi-timeframe momentum z-scores, Hurst exponent estimators, and trend-acceleration oscillators tuned on 11 years of OHLCV across crypto and FX majors.

    • 9 factors
    • 1m – 1w horizons
  3. 03

    On-Chain & Network Health

    Exchange netflows, miner positioning, stablecoin issuance deltas, and active-address divergence — applied only to the 38 spot pairs the platform signals.

    • 8 factors
    • 10m refresh
  4. 04

    Sentiment & Narrative Drift

    News-flow NLP scoring, funding-rate skew, and social-velocity decay rates — used to fade crowded trades, not to chase them.

    • 7 factors
    • Quantified headline scoring
  5. 05

    Macro & Cross-Asset Overlay

    DXY, 10Y real yields, VIX term-structure, and risk-on/risk-off rotation cues propagated into FX-major signal generation and crypto beta-adjustment.

    • 6 factors
    • 1h refresh
  6. 06

    Regime Detection & Confirmation

    Hidden Markov regime classifier gates the entire model — and the two-factor confirmation layer demands independent agreement before any order is dispatched to your exchange bridge.

    • 6 factors
    • −63% false breakouts
Signal Coverage

One console. Three markets. 41 million events a year.

Spot crypto, perpetual futures, and FX majors — signalled on the same 47-factor engine, executed through venue-tuned bridges that match the speed profile of each market.

Highest volume segment

Spot Crypto · BTC, ETH & Top-20 Pairs

The platform's flagship coverage. Spot signals on Bitcoin, Ethereum, and the top-20 USDT pairs account for the majority of the 41M signal events processed in 2024 — routed through Binance, Kraken, and Coinbase bridges with sub-380ms execution.

  • 20 spot pairs
  • 3 venue bridges
  • <380ms execution
Read the spot execution brief →

Perpetual Futures · Long & Short Bias

Leveraged perpetual contracts across Binance and Bybit, signalled with funding-rate awareness and an automated liquidation-distance guardrail.

  • 14 perp pairs
  • Up to 25× configurable

FX Majors · MetaTrader 5 Bridge

EUR/USD, GBP/USD, USD/JPY and AUD/USD signals delivered to MetaTrader 5 with the same 47-factor model — built for traders running hybrid crypto-and-FX books.

  • 4 major pairs
  • MT5 native

Execution Footprint · 2024

  • Spot events24.6M
  • Perp events13.9M
  • FX major events2.5M
  • Platform uptime99.97%
See audited breakdown →
Latency Benchmarks

Per-venue signal-to-execution measured and audited

Every bridge below is benchmarked monthly against the counterparty's published WebSocket round-trip. Numbers are median milliseconds from signal publish to order acknowledgement on the matching venue, captured across the trailing 90-day window.

Binance Spot · Perps
312 ms

Median across 8.4M fills, Q1 2025

Kraken Spot
347 ms

Median across 2.1M fills, Q1 2025

Coinbase Spot · Perps
298 ms

Median across 5.7M fills, Q1 2025

MetaTrader 5 FX Majors
364 ms

Median across 1.9M fills, Q1 2025

Platform uptime, 2024 calendar year 99.97%
Read the latency methodology
Two-Factor Signal Confirmation

Two independent layers agree before any order leaves the stack.

Retail signal bots fire on a single indicator crossing a threshold and wonder why their entries get wicked out. GasProfit routes every tick through two parallel confirmation layers and only emits an executable when both sides align within a defined window.

  1. Layer A

    47-factor momentum model

    Trained on 11 years of OHLCV data across 1,800+ asset pairs. Scores every candle on momentum, structure, and volatility regimes and emits a directional probability between 0 and 1.

  2. Layer B

    Microstructure & order-flow filter

    Reads live L2 book depth, trade-tape imbalance, and funding-rate skew. Confirms or vetoes Layer A based on whether the directional probability matches the live order-flow signature.

  3. Output

    Confirmed signal & risk envelope

    When both layers agree inside a 250ms window, the platform emits a signal with entry, stop, and target bands sized by the recent realised volatility of the venue.

−63% False-breakout entries vs single-indicator systems, audited by QuantVerify Labs across 2023–2025
Audited win rate 94.2% QuantVerify Labs · 2023, 2024, 2025
Security certification SOC 2 Type II Certified since 2022
Trader rating 4.8 / 5 Trustpilot, 21,400+ reviews
Cumulative volume $2.8B Client trading volume since 2019 launch

GasProfit operates as a software vendor and signal provider; brokerage and custody services are delivered through FCA-regulated partner firms. Signal accuracy reflects historical, audited outcomes and does not guarantee future results.