THE MATHEMATICAL EDGE

A structural advantage derived from statistical physics, not from data mining or curve fitting. The indicators come from established mathematical fields with decades of academic validation. The edge transfers across regimes because it emerges from the mathematical structure of equilibrium-seeking processes, not from the specific history of any currency pair.

THREE STRUCTURAL ADVANTAGES

ρ ≈ 0

Orthogonal Decomposition

MAE and MFE correlation near zero. Predicting risk and reward as separate, independent dimensions produces dramatically better signal filtering than direct PnL prediction.

Defense Layers

Every signal passes through 12 sequential gates. Each addresses a different risk dimension. No single layer bears the full burden.

7 / 34

Rigorous Selection

Seven indicators deployed from 34 researched. A 21% deployment rate. The 79% rejected demonstrate intellectual rigor, not wasted effort.

FOUNDATION

dX = [███████████]

STOCHASTIC EQUILIBRIUM PROCESS
[███] PARAMETERS . [███] ESTIMATION METHOD . [███] CONVERGENCE CRITERIA

CLASSIFIED

CORE EQUATION AND PARAMETERS AVAILABLE UNDER NDA

The system models currency prices as a stochastic process that tends toward a dynamic equilibrium. When price deviates significantly from that equilibrium, the system computes the exact first-passage probability: the likelihood that price will reach the take-profit level before the stop-loss. This is an analytical probability computation from stochastic calculus, not a heuristic or approximation.

Trades execute only when that probability exceeds strict thresholds. When the process parameters indicate that convergence probability is marginal, the system stays silent. The best trade is often no trade at all.

THEORY-DRIVEN INDICATORS

Every indicator comes from an established mathematical field with decades of academic validation. None were invented to trade FX. Each addresses a fundamentally different dimension of market behavior. This reduces the risk of discovering spurious patterns by testing thousands of indicators until something fits historical data.

INDICATOR CLASS A STOCHASTIC PROCESSES 3 VARIANTS DEPLOYED

INDICATOR CLASS B INFORMATION THEORY 1 VARIANT DEPLOYED

INDICATOR CLASS C STATE ESTIMATION 1 VARIANT DEPLOYED

INDICATOR CLASS D EXTREME VALUE THEORY 1 VARIANT DEPLOYED

INDICATOR CLASS E SPECTRAL ANALYSIS 1 VARIANT DEPLOYED

INDICATOR CLASS F FRACTAL GEOMETRY DIAGNOSTIC ONLY

CLASSIFIED

INDICATOR NAMES, FORMULATIONS, AND PARAMETERS AVAILABLE UNDER NDA

34 mathematical indicator families researched across 6+ established fields. Only 7 deployed live. 7 confirmed dead ends. 20 remain as research diagnostics. The ratio itself is evidence of rigorous filtering: 79% of investigated approaches were deliberately not deployed.

5-STRATEGY ENSEMBLE

Nature does not optimize for a single species. It builds ecosystems: diverse populations that collectively adapt to changing environments. A forest with ten tree species survives a blight that would kill a monoculture. The same principle drives the ensemble architecture.

These strategies are not redundant. Their signal correlations are low. They fire on different bars, in different regimes, for different structural reasons. When one strategy class is profitable, another might be flat. The ensemble is not the five best individual strategies. It is the five strategies that, together, produce the most robust portfolio.

STRATEGY S-1

Primary signal class. High-frequency signal generator. Redacted methodology.

W: ███

STRATEGY S-2

Secondary signal class. Rare, high-conviction trades. Redacted methodology.

W: ███

STRATEGY S-3

Tertiary signal class. Selective trend capture. Redacted methodology.

W: ███

STRATEGY S-4

Volume contributor. Workhorse signal generator. Redacted methodology.

W: ███

STRATEGY S-5

Diversity contributor. Novel detection approach. Redacted methodology.

W: ███

CLASSIFIED

STRATEGY NAMES, DESCRIPTIONS, AND WEIGHTS AVAILABLE UNDER NDA

DIVERSITY SELECTION

Strategies are selected not for individual performance, but for collective robustness. A strategy highly correlated with an already-selected one is penalized. Diversity outweighs individual fitness in the selection process.

GENETIC OPTIMIZATION

A niched evolutionary algorithm enforces population diversity across strategy types. Adaptive mutation and crossover prevent evolutionary convergence to a single dominant strategy, maintaining ecosystem balance.