API Reference

This page contains a list of the essential end-user API functions and classes. The data sets that are included in the package are described in Data Sets, and the warnings and exceptions that the package raises are described in Warnings and Exceptions.

Volatility Modeling

High-level

arch_model(y[, x, mean, lags, vol, p, o, q, ...])

Initialize common ARCH model specifications.

Mean Specification

ConstantMean([y, hold_back, volatility, ...])

Constant mean model estimation and simulation.

ZeroMean([y, hold_back, volatility, ...])

Model with zero conditional mean estimation and simulation.

HARX([y, x, lags, constant, use_rotated, ...])

Heterogeneous Autoregression (HAR) model with optional exogenous regressors.

ARX([y, x, lags, constant, hold_back, ...])

Autoregressive model with optional exogenous regressors.

LS([y, x, constant, hold_back, volatility, ...])

Least squares model estimation and simulation.

MAMean([y, lags, constant, hold_back, ...])

Moving average (MA) mean model, with optional constant.

ARCHInMean([y, x, lags, constant, ...])

(G)ARCH-in-mean model and simulation.

Volatility Process Specification

ARCH([p])

ARCH process.

GARCH([p, o, q, power])

GARCH and related model estimation.

EGARCH([p, o, q])

EGARCH model estimation.

APARCH([p, o, q, delta, common_asym])

Asymmetric Power ARCH (APARCH) volatility process.

HARCH([lags])

Heterogeneous ARCH process.

FIGARCH([p, q, power, truncation])

FIGARCH model.

FIAPARCH([p, o, q, delta, truncation])

Fractionally Integrated Asymmetric Power ARCH (FIAPARCH) volatility process.

NGARCH([p, o, q])

Nonlinear asymmetric GARCH (NGARCH) volatility process.

MIDASHyperbolic([m, asym])

MIDAS Hyperbolic ARCH process.

EWMAVariance([lam])

Exponentially Weighted Moving-Average (RiskMetrics) Variance process.

RiskMetrics2006([tau0, tau1, kmax, rho])

RiskMetrics 2006 Variance process.

ConstantVariance()

Constant volatility process.

FixedVariance(variance[, unit_scale])

Fixed volatility process.

Shock Distributions

Normal(*[, seed])

Standard normal distribution for use with ARCH models.

StudentsT(*[, seed])

Standardized Student's distribution for use with ARCH models.

SkewStudent(*[, seed])

Standardized Skewed Student's distribution for use with ARCH models.

GeneralizedError(*[, seed])

Generalized Error distribution for use with ARCH models.

Residual Diagnostics

excess_kurtosis(std_resid)

Sample excess kurtosis of standardized residuals.

hill_estimator(std_resid[, k])

Hill tail-index estimator of standardized residuals.

var_ratio([p, distribution, parameters])

Ratio of Gaussian VaR to model-distribution VaR.

VaR Backtests

var_exceptions(returns, var)

Identify the exceptions of a VaR forecast.

kupiec_pof(returns, var[, p])

Kupiec proportion-of-failures test of the coverage of VaR.

christoffersen_independence(returns, var)

Christoffersen test of the independence of the exceptions of VaR.

christoffersen_conditional_coverage(returns, var)

Christoffersen test of the conditional coverage of VaR.

traffic_light(returns, var[, p, window])

Basel traffic light zone of the exceptions of VaR.

TrafficLightResult(zone, exceptions, ...)

Result of the Basel traffic light backtest of VaR.

Model Results and Forecasts

ARCHModelResult(params, param_cov, r2, ...)

Results from estimation of an ARCHModel model.

ARCHModelFixedResult(params, resid, ...)

Results for fixed parameters for an ARCHModel model.

ARCHModelForecast(index, start_index, mean, ...)

Container for forecasts from an ARCH Model.

ARCHModelForecastSimulation(index, values, ...)

Container for a simulation or bootstrap-based forecasts from an ARCH Model.

VarianceForecast(forecasts[, ...])

Container for variance forecasts.

WaldTestStatistic(stat, df, null, alternative)

Test statistic holder for Wald-type tests.

Extending the Model Components

ARCHModel([y, volatility, distribution, ...])

Abstract base class for mean models in ARCH processes.

VolatilityProcess()

Abstract base class for ARCH models.

VolatilityUpdater

Base class that all volatility updaters must inherit from.

Distribution(*[, seed])

Template for subclassing only.

Unit Root Testing

ADF(y[, lags, trend, max_lags, method, ...])

Augmented Dickey-Fuller unit root test.

DFGLS(y[, lags, trend, max_lags, method, ...])

Elliott, Rothenberg and Stock's ([ers]) GLS detrended Dickey-Fuller.

PhillipsPerron(y[, lags, trend, test_type])

Phillips-Perron unit root test.

ZivotAndrews(y[, lags, trend, trim, ...])

Zivot-Andrews structural-break unit-root test.

VarianceRatio(y[, lags, trend, debiased, ...])

Variance Ratio test of a random walk.

KPSS(y[, lags, trend])

Kwiatkowski, Phillips, Schmidt and Shin (KPSS) stationarity test.

auto_bandwidth(y[, kernel])

Automatic bandwidth selection of Andrews (1991) and Newey & West (1994).

Cointegration Testing

engle_granger(y, x[, trend, lags, max_lags, ...])

Test for cointegration within a set of time series.

phillips_ouliaris(y, x[, trend, test_type, ...])

Test for cointegration within a set of time series.

EngleGrangerTestResults(stat, pvalue, crit_vals)

Results class for Engle-Granger cointegration tests.

PhillipsOuliarisTestResults(stat, pvalue, ...)

Results class for Phillips-Ouliaris cointegration tests.

Cointegrating Relationship Estimation

CanonicalCointegratingReg(y, x[, trend, x_trend])

Canonical Cointegrating Regression cointegrating vector estimation.

DynamicOLS(y, x[, trend, lags, leads, ...])

Dynamic OLS (DOLS) cointegrating vector estimation.

FullyModifiedOLS(y, x[, trend, x_trend])

Fully Modified OLS cointegrating vector estimation.

CointegrationAnalysisResults(params, cov, ...)

Estimation results for cointegrating vector estimators.

DynamicOLSResults(params, cov, resid, lags, ...)

Estimation results for Dynamic OLS models.

Bootstraps

IIDBootstrap(*args[, seed])

Bootstrap using uniform resampling.

IndependentSamplesBootstrap(*args[, seed])

Bootstrap where each input is independently resampled.

StationaryBootstrap(block_size, *args[, seed])

Politis and Romano (1994) bootstrap with expon distributed block sizes.

CircularBlockBootstrap(block_size, *args[, seed])

Bootstrap using blocks of the same length with end-to-start wrap around.

MovingBlockBootstrap(block_size, *args[, seed])

Bootstrap using blocks of the same length without wrap around.

LocalBlockBootstrap(block_size, locality, *args)

Bootstrap using blocks that are resampled from near their own position.

Block-length Selection

optimal_block_length(x)

Estimate optimal window length for time-series bootstraps.

Testing with Multiple-Comparison

SPA(benchmark, models[, block_size, reps, ...])

Test of Superior Predictive Ability (SPA) of White and Hansen.

MCS(losses, size[, reps, block_size, ...])

Model Confidence Set (MCS) of Hansen, Lunde and Nason.

StepM(benchmark, models[, size, block_size, ...])

StepM multiple comparison procedure of Romano and Wolf.

Long-run Covariance (HAC) Estimation

Andrews(x[, bandwidth, df_adjust, center, ...])

Alternative name of the QuadraticSpectral covariance estimator.

Bartlett(x[, bandwidth, df_adjust, center, ...])

Bartlett's (Newey-West) kernel covariance estimation.

Gallant(x[, bandwidth, df_adjust, center, ...])

Alternative name for Parzen covariance estimator.

NeweyWest(x[, bandwidth, df_adjust, center, ...])

Alternative name for Bartlett covariance estimator.

Parzen(x[, bandwidth, df_adjust, center, ...])

Parzen's kernel covariance estimation.

ParzenCauchy(x[, bandwidth, df_adjust, ...])

Parzen's Cauchy kernel covariance estimation.

ParzenGeometric(x[, bandwidth, df_adjust, ...])

Parzen's Geometric kernel covariance estimation.

ParzenRiesz(x[, bandwidth, df_adjust, ...])

Parzen-Reisz kernel covariance estimation.

QuadraticSpectral(x[, bandwidth, df_adjust, ...])

Quadratic-Spectral (Andrews') kernel covariance estimation.

TukeyHamming(x[, bandwidth, df_adjust, ...])

Tukey-Hamming kernel covariance estimation.

TukeyHanning(x[, bandwidth, df_adjust, ...])

Tukey-Hanning kernel covariance estimation.

TukeyParzen(x[, bandwidth, df_adjust, ...])

Tukey-Parzen kernel covariance estimation.

ZeroLag(x[, bandwidth, df_adjust, center, ...])

Zero-lag (No autocorrelation) kernel covariance estimation.

PreWhitenedRecolored(x, *[, lags, method, ...])

VAR-HAC and Pre-Whitened-Recolored Long-run covariance estimation.

CovarianceEstimator(x[, bandwidth, ...])

Base class of the kernel-based long-run covariance estimators.

CovarianceEstimate(short_run, one_sided_strict)

Covariance estimate using a long-run covariance estimator.

Data Sets

load()

Load the S&P 500 data used in the examples.

load()

Load the NASDAQ Composite data used in the examples.

load()

Load the VIX Index data used in the examples.

load()

Load the West Texas Intermediate crude oil price data used in the examples.

load()

Load the Brent and WTI crude oil price data used in the examples.

load()

Load the Core CPI data used in the examples.

load()

Load the AAA and BAA rates used in the examples.

load()

Load the Fama-French factor data used in the examples.

load()

Load the graduate school admissions data used in the examples.

load_file(file_base, filename)

Load data from a csv.gz file.

Utilities

test([extra_args, exit, append, location])

Run the tests of the installed package.