quant-statistics

ADF unit-root stationarity testing checks whether your time-series is ready for modeling by detecting non-stationary trends. This Skill provides ready-to-use implementations that output test statistics and interpretation guidance for equity and crypto datasets.

Cointegration analysis identifies long-run equilibrium relationships between two or more non-stationary time-series. You need it for pairs trading and statistical arbitrage to confirm that asset price spreads are mean-reverting rather than randomly drifting.

GARCH family volatility modeling captures time-varying variance in financial returns for risk management. This Skill provides diagnostic checks and example implementations that output conditional variance forecasts and model fit metrics using the arch library.

Bootstrap inference resamples your historical returns or performance metrics to estimate confidence intervals without relying on strict distributional assumptions. This Skill applies bootstrap methods to evaluate strategy robustness and performance metric significance across time-series datasets.

Yes, you need pandas, numpy, statsmodels, and arch libraries to execute the ADF, cointegration, GARCH, and bootstrap methods. These Python data-science dependencies handle the time-series manipulation and statistical computations required to reproduce the provided results.

Regression diagnostics validate model assumptions by checking residuals for autocorrelation, heteroskedasticity, and normality. This Skill automates diagnostic reporting on equity and fixed income datasets, outputting clear test results and interpretation guidance for your statistical models.