financial-data-validation 0.1.1

Statistical validation for synthetic financial time series

financial-data-validation

Statistical validation for synthetic financial time series.

Installation

pip install finanial-data-validation

Quick Start

from financial_data_validation import validate_paths
import numpy as np

# Your synthetic price paths (n_paths, n_timesteps)
paths = np.random.lognormal(0, 0.02, size=(1000, 252))

# Validate
report = validate_paths(paths)

print(f"Quality Score: {report.quality_score:.1f}/100")
print(f"Passed: {'✓' if report.passed else '✗'}")

What It Tests

Test	Validates	Pass Criteria
Ljung-Box	No spurious autocorrelation in returns	p > 0.05
ARCH	Volatility clustering present	p < 0.05
Jarque-Bera	Returns approximately normal	p > 0.01
Kolmogorov-Smirnov	Distribution shape matches expectation	D < 0.05

Quality Scores

• 90-100: Excellent — indistinguishable from real markets
• 80-89: Good — suitable for most applications
• 70-79: Acceptable — passes minimum requirements
• < 70: Poor — may produce unreliable results

Why This Exists

Synthetic market data is only useful if it's statistically realistic. This package validates whether generated price paths exhibit the properties of real financial markets: volatility clustering, fat tails, and proper correlation structure.

Built by QPaths — we use this to validate every dataset we generate.

Example Output

ValidationReport(
    ljung_box_score=0.87,
    arch_score=0.92,
    jarque_bera_score=0.81,
    ks_score=0.85,
    quality_score=86.4,
    passed=True
)

Use Cases

• Validate synthetic data before backtesting trading strategies
• Quality-check Monte Carlo simulations
• Verify stochastic model implementations
• Test financial data generation pipelines

Documentation

Full documentation: github.com/qpaths/financial-data-validation

License

MIT

---

Part of the QPaths ecosystem — Generate validated synthetic market data at qpaths.io