ZenWaveCapital

Anonymous AI-accelerated quant research lab specializing in long-horizon, Walk-Forward-validated USDJPY engines built with institutional-grade structure, strict risk controls, and fully automated execution.

ZenWaveCapital Quantitative Validation Report

Models: ZenWave A & ZenWave B
Market: USDJPY
Methodology: Systematic · AI-Accelerated · Fully Automated · Anonymity by Design

ZenWaveCapital operates as an anonymous AI-accelerated quant research lab, using a dual-model validation workflow that mirrors the governance practices of modern institutional quant teams.

  • The primary internal AI is used for procedural validation:
    WFA slicing, statistical auditing, data boundary checks, leakage prevention.
  • A secondary external public AI is used as an independent auditor to verify interpretation, methodological integrity, and bias mitigation.

This two-layer validation framework ensures high procedural correctness, consistent out-of-sample behavior, and statistical defensibility.

This report documents the formal validation results for ZenWave A and ZenWave B.

📐 Walk-Forward Analysis (WFA) Overview

ZenWave uses multi-slice Walk-Forward Analysis (WFA) as its primary robustness framework.
Every engine must demonstrate profitable, stable out-of-sample (OOS) behavior across multiple distinct volatility regimes.

Validation includes:

  • AI-assisted verification of slicing boundaries
  • Independent IS/OOS partitions with zero leakage
  • Equity shape & structural consistency checks
  • PF/DD stability and regime alignment
  • Parameter drift analysis
  • Dual-AI audit to reduce human bias and methodological blind spots

This creates institutional-grade validation rigor and minimizes the chance of overfitting or false robustness.

🧭 A-Series WFA Timeline (Long-Term Engine)

Slice Role Period
A1 IS 2014–2018
A1 OOS 2019–2020
A2 IS 2016–2020
A2 OOS 2021–2022
A3 IS 2018–2022
A3 OOS 2023–2024
A4 IS 2019–2023
A4 OOS 2024–2025

A4 uses an extended IS window for validation only.
Live ZenWave A uses the full-window 2014–2025 parameter set, confirmed stable by all A1–A4 slices.

🧭 B-Series WFA Timeline (Medium-Term Engine)

Slice Role Period
B1 IS 07/2018–2020
B1 OOS 2021
B2 IS 2019–2021
B2 OOS 2022
B3 IS 2020–2022
B3 OOS 2023
B4 IS 2021–2023
B4 OOS 2024
B5 IS 2022–2023
B5 OOS 2024–2025

B5 is selected for production due to exceptional IS/OOS alignment, low fragility, and strong performance during the 2022–2025 volatility regime.


🧠 Portfolio Structure (A/B Engines)

ZenWave Portfolio
       │
┌──────┴────────┬──────────────┐
│               │              │
│       ZenWave A              │
│   Long-Term Breakouts        │
│                              │
└──────┬────────┘      ┌───────┘
       │               │
   ZenWave B (Medium-Term Breakouts)
       │
       ▼
 Combined Output
  • ZenWave A = long-term structural breakout engine (2014–2025)
  • ZenWave B = medium-term adaptive breakout engine (2018–2025)
  • Combined = multi-horizon robustness with reduced decay risk and broader regime coverage

This structure mirrors institutional multi-horizon portfolio design.

🔐 Risk-Control Framework (A & B)

Position Sizing (1% risk)
           │
           ▼
     Stop-Loss Logic
           │
           ▼
    Take-Profit Logic
           │
           ▼
┌────────────────────────────────────────────┐
│  No Martingale                             │
│  No Grid                                   │
│  No Simultaneous Hedge Positions           │
│  Reversal entries are independent signals  │
└────────────────────────────────────────────┘
           │
           ▼
    Server-Side Execution

This mirrors institutional risk standards:

  • No recovery trading
  • No leverage escalation
  • No offsetting hedge stacks
  • All entries independent and rule-driven


📘 ZenWave A — Quant Validation Summary

Strategy Type

Long-term USDJPY structural breakout engine (2014–2025).
Fully automated, deterministic rules, fixed SL/TP.

Walk-Forward Findings (A1–A4)

ZenWave A demonstrates:

  • Fully profitable OOS behavior across all slices
  • No catastrophic regime failures
  • Stable profit factor
  • Controlled drawdowns
  • Natural parameter evolution
  • Structural consistency across 12 years

AI-Assisted Validation Findings

  • All slicing boundaries verified
  • Zero data leakage
  • IS/OOS behavior shows strong structural alignment
  • Robustness ranks as exceptional within ZenWave’s internal WFA benchmark library
  • External AI auditor confirms methodology as statistically disciplined and institutionally sound

Robustness Conclusion

ZenWave A passes with high institutional confidence.
Its full-window parameters form the long-term backbone of the ZenWave portfolio.

WFA Video: 🎥 Watch


📘 ZenWave B — Quant Validation Summary

Strategy Type

Medium-term USDJPY breakout engine (2018–2025).
Adaptive structural logic, deterministic execution.

Walk-Forward Findings (B1–B5)

All slices produce profitable OOS performance:

  • B1 — Moderate stability
  • B2 — Strong and balanced
  • B3 — Weakest slice but profitable
  • B4 — Excellent stability
  • B5 — Best alignment with modern volatility

Parameter Behavior

  • Controlled drift across slices
  • No chaotic jumps
  • Stable PF/DD characteristics

AI-Assisted Validation Findings

  • Verified slicing integrity
  • No overfitting detected
  • Behavior matches expectations for a medium-horizon structural engine

Robustness Conclusion

ZenWave B passes WFA with stable OOS results.
B5 is selected for live deployment due to regime alignment and low fragility.

WFA Video: 🎥 Watch

2025 Year-End Baseline (Post-Launch)

This section establishes the official post-launch baseline for ZenWave A and B as of 31 Dec 2025.

Engine Status (as of 31 Dec 2025)

Engine Status Production Parameters Notes
ZenWave A Live Full-window (2014–2025) Frozen
ZenWave B Live B5 slice Frozen

Governance: As of 31 Dec 2025, ZenWave A and B parameters are frozen.
No parameter, risk, or logic changes are permitted outside a formal retirement or replacement decision.

FPT Consistency Check

Backtest (Demo) vs Backtest (Live Account) vs Live Actual

To verify execution realism, a three-way forward-performance comparison was performed:

  • Backtest (Demo Account Assumptions) — optimistic execution envelope
  • Backtest (Live Account Assumptions) — conservative execution envelope
  • Live Actual — real execution

Result

Live execution remains within the conservative envelope of historical simulation.

Observed characteristics:

  • Comparable trade frequency and directionality
  • Divergence primarily attributable to spread variability and execution friction
  • No structural anomalies or tail losses beyond modeled risk

Execution Realism Note

Historical replay under demo and live account assumptions can produce different outcomes due to spread dynamics, commission modeling, and fill assumptions.

For ZenWave:

  • Demo backtest represents the upper bound
  • Live-account backtest represents the lower bound
  • Live actual represents reality

This relationship is expected and does not invalidate the underlying strategy logic.

Transition to Forward Performance Tracking

This Quantitative Validation Report documents what was built, how it was tested, and why it was approved for production.

It intentionally stops short of ongoing performance reporting.

From this point forward, live behavior is documented separately through Forward Performance Tracking (FPT) — a structured, periodic log that compares:

  • demo backtests (optimistic execution assumptions)
  • live-account backtests (conservative execution assumptions)
  • actual live trades

Each Forward Performance Tracking entry:

  • uses the same parameters frozen at year-end 2025
  • covers the same time windows
  • includes full video context explaining assumptions and observations

No changes are made retroactively.

Forward Performance Tracking

Live performance tracking for ZenWave A and B is published here:

👉 Forward Performance Tracking

The accompanying video explanations are available on the ZenWaveCapital YouTube channel.

This separation ensures that research, validation, and live observation remain clearly defined and auditable over time.