QUANTITATIVE EDGE
IN CURRENCY MARKETS
AI-POWERED FX ENSEMBLE SYSTEM BUILT ON STATISTICAL PHYSICS
0
SHARPE RATIO (OOS)
0
VALIDATED TRACK RECORD
ZERO
LOSING CALENDAR YEARS
0
DATA POINTS ANALYZED
Neural Predictiva is a fully automated quantitative FX trading system operating across five currency pairs with a fleet of live daemons. The mathematical edge comes from a stochastic equilibrium process rooted in statistical physics. Five diverse strategies vote through a weighted ensemble, filtered by a 12-layer risk gate before any trade executes. Every metric is net of conservative transaction costs.
THE OPPORTUNITY
FX pairs exhibit measurable equilibrium convergence on hourly timeframes, driven by central bank policy anchoring, carry trade dynamics, and institutional rebalancing. This structural behavior has been studied in academic literature for decades. Neural Predictiva captures it through rigorous mathematical modeling.
$7.5T
DAILY FX VOLUME
The most liquid market in the world. 24/5 operation with no single point of failure.
0.0000
PBO SCORE
Zero probability of backtest overfitting. The strongest possible anti-overfitting evidence.
310+
RESEARCH SESSIONS
Two months of intensive, hypothesis-driven research. Over 30% documented as dead ends, evidence of rigorous filtering.
WHY IT WORKS
01
Structural Edge
The core process model comes from statistical physics. The edge comes from the mathematical structure of equilibrium-seeking dynamics, not from historical patterns of any specific pair.
02
Path-Dependent Gating
MAE and MFE are nearly orthogonal (near-zero correlation). Predicting them separately decomposes PnL into independent risk and reward dimensions. 2.7x better than direct PnL gating.
03
12-Layer Filtering
Raw signals pass through multi-dimensional filtering across risk, reward, and regime dimensions. Defense in depth.