Home/Documentation/Quantitative Methodology
Mathematical Rigor · Elvaris Capital Quant Lab

Quantitative Research & Backtesting Methodology

Every trading strategy listed on StratMarket by Elvaris Capital (elvariscapital.in) is subject to strict institutional mathematical protocols. We reject over-optimized curve fitting and enforce zero look-ahead bias, realistic broker execution drag, and multi-regime stress testing.

The Four-Stage Quant Pipeline

1

Generate: AI-Assisted Hypothesis Codification

Strategies start from an economic or statistical thesis (e.g. Asian session liquidity compression followed by London volume breakout, or volatility exhaustion reversion on Gold). Code is written in strict, deterministic rules with zero discretionary human intervention.

2

Backtest: High-Fidelity Tick Historical Simulation

Models run against institutional tick data spanning 5,000 to over 100,000 bars. Fixed transaction fees ($2.50 per standard lot) and 1.0 to 1.5 pip slippage drag are applied to every single entry and stop-loss trigger.

3

Improve: Walk-Forward Optimization & Versioning

Parameters are tested across rolling historical windows (Walk-Forward Optimization). Any strategy that demonstrates performance cliffs under minor parameter shifts is discarded. Approved models receive immutable semantic version numbers (e.g. v1.2.0).

4

Validate: Blind Out-of-Sample & Monte Carlo Stress Testing

The final 30% of data is evaluated blindly as Out-of-Sample (OOS). Furthermore, 1,000 Monte Carlo simulations randomly permute trade orders and test equity curves against catastrophic 99% Value-at-Risk (VaR) scenarios.

Core Mathematical Metrics

RISK-ADJUSTED RETURN

Sharpe Ratio

SR = (E[Rp] - Rf) / σp

Measures annualized excess return per unit of total portfolio volatility. StratMarket strategies range from 1.65 to 10.71.

DRAWDOWN RECOVERY

Calmar Ratio

CR = Annualized Return / |Max Drawdown|

Quantifies return relative to the single deepest peak-to-trough decline. A Calmar ratio above 5.0 represents elite institutional risk containment.

GROSS EXPECTANCY

Profit Factor

PF = Gross Profits / Gross Losses

Gross dollar winnings divided by gross dollar losses. PS-02 Apex achieves a 9.31 Profit Factor, and PY-05 achieves 2.28 over 5 years.

WORST-CASE DROP

Maximum Drawdown

MDD = (Peak - Trough) / Peak

All 10 Python ensembles are engineered with hard single-digit drawdowns (< 10%), and Pine Script strategies under 3%.