Signal Validation Results & Analytics Transparency

Explore backtest results, independent validations, performance statistics, and the methodology that ensures Axiom's analytics meet the highest institutional standards.

All Results Are Hypothetical & Simulated

All metrics below are derived from out-of-sample backtests conducted on historical data from 2018–2024, using walk-forward cross-validation and strict adherence to no-look-ahead bias protocols.

Sharpe Ratio

1.2–1.8range

Risk-adjusted returns across market regimes

Simulated, out-of-sample, 2018-2024

Maximum Drawdown

15–25%range

Peak-to-trough decline in simulated equity

Hypothetical, excludes slippage/costs

Win Rate

52–58%percentage

Proportion of profitable signals

Cross-validated, multiple regime types

Average Annual Return

8–18%simulated

Hypothetical annualized performance

Not indicative of future results

Important Performance Disclosure

All performance statistics are hypothetical, based on simulated backtests, and not derived from live trading accounts. Results do not account for real-world factors including transaction costs, slippage, market impact, or execution delays. Past performance is not indicative of future results. Axiom provides analytics only—not investment advice or portfolio management.

Axiom's AI models are stress-tested against historical market shocks, regime changes, and extreme volatility events to ensure robustness across diverse conditions.

COVID-19 Market Shock (Mar 2020)

Pass

Signals reduced exposure 14 days prior; max simulated drawdown 12%

2022 Rate Hike Regime (Fed tightening)

Pass

Regime detection triggered defensive positioning; simulated volatility-adjusted return +6%

High Volatility Period (VIX > 30)

Pass

Signal confidence intervals widened appropriately; win rate maintained at 54%

Low Liquidity Environment (Asian session)

Pass

Reduced signal frequency by 40%; avoided false positives in thin markets

Stress Test Methodology

All scenarios are evaluated using historical data replay with no look-ahead bias. Models are frozen at the point in time before each event, and signals are generated using only information available at that moment. Results reflect simulated performance and do not guarantee future behavior.

Axiom employs institutional-grade validation protocols to ensure signal quality, prevent overfitting, and maintain model integrity across market conditions.

Walk-Forward Cross-Validation

Models are trained on 70% of historical data and tested on the remaining 30% in a rolling time-series fashion. This prevents look-ahead bias and ensures out-of-sample performance evaluation.

K-Fold Cross-Validation

5-fold cross-validation used for hyperparameter tuning. Each fold represents a distinct market regime to prevent regime-specific overfitting.

No Look-Ahead Bias Protection

Strict temporal ordering enforced in all training and testing. Future data is never used to generate historical signals. All feature engineering respects information available at signal generation time.

Out-of-Sample Performance Dashboard

All publicly reported metrics are derived exclusively from out-of-sample test sets. In-sample (training) performance is monitored internally but never used for marketing or client communication.

Critical Performance Disclaimer

All reported results are based on simulated backtests using historical data and are not derived from live trading accounts. Past performance is not indicative of future results. Actual trading may result in losses, and no guarantee of profitability is made or implied.

  • Hypothetical & Simulated Performance

    All metrics reflect simulated performance based on backtests. No real capital was deployed. Results do not account for the psychological, operational, or technical challenges of live trading.

  • Transaction Costs & Slippage Not Fully Modeled

    Backtests assume conservative transaction costs (spreads, commissions) but do not fully account for slippage, market impact, or liquidity constraints that occur in live execution.

  • Market Conditions May Change

    Historical market structure, volatility, and liquidity patterns used in backtests may not persist in future markets. Regime shifts, regulatory changes, or structural breaks can invalidate historical patterns.

  • No Guarantee of Signal Accuracy

    Axiom's AI models generate probabilistic signals based on historical patterns. There is no guarantee that any signal will be accurate, profitable, or suitable for any particular investor or strategy.

  • User Responsibility for Risk Management

    Axiom provides analytics only—not investment advice, portfolio management, or risk oversight. Users are solely responsible for their own risk management, position sizing, and compliance with applicable laws.

Independent Audit & Oversight

Axiom's methodology and performance results are subject to regular independent review to ensure transparency, accuracy, and compliance with institutional standards.

  • Annual review and verification by independent quantitative audit firms
  • Peer review with internal and external quantitative advisors
  • Compliance oversight by Chief Compliance Officer and legal counsel
  • All methodology changes logged and reviewed before deployment
  • Institutional clients may request summary audit reports or detailed methodology documentation

Request Detailed Methodology & Audit Reports

For institutional due diligence, compliance requests, or detailed methodology documentation, contact our compliance team. Full audit summaries, raw data, and custom reports available for qualified institutional clients.

Download Full MethodologyView Security & Risk

Legal & Compliance Notice

All reported metrics are based on simulated backtests and must not be interpreted as guarantees of future performance or investment advice. Axiom is not a registered investment advisor, broker-dealer, or fiduciary. Axiom provides analytics-only services—all trading execution, custody, and risk management remain with user-selected regulated brokers. Detailed methodology, assumptions, and raw data available on request for institutional due diligence.

Terms of ServicePrivacy PolicySecurity & MethodologyContact Compliance