FANPAC MT

FANPAC MT 3.0

The Financial Analysis Application (FANPAC) provides econometric tools commonly implemented for estimation and analysis of financial data. The FANPAC application allows users to tailor each session to their specific modeling needs and is designed for estimating parameters of univariate and multivariate Generalized Autoregressive Conditionally Heteroskedastic (GARCH) models.

Features

  • Univariate ARCH, GARCH, ARMAGARCH,FIGARCH
  • Multivariate BEKK, DVEC,CCC,DCC,GO,FM,VAR
  • Normal, t, skew generalized t, multivariate skew distributions
  • Keyword interface

Supported models include:

  • BEKK GARCH model
  • Diagonal vec multivariate models:
  • -GARCH model
    -Fractionally integrated GARCH model
    -GJR GARCH model

  • Multivariate constant conditional correlation models:
  • -GARCH model
    -Exponential GARCH model
    -Fractionally integrated GARCH model
    -GJR GARCH model

  • Multivariate dynamic conditional correlation models:
  • -GARCH model
    -Exponential GARCH model
    -Fractionally integrated GARCH model
    -GJR GARCH model

  • Multivariate factor GARCH model
  • Generalized orthogonal GARCH model
  • Univariate time series models:
  • -GARCH model
    -OLS
    -ARIMA

Modeling flexibility provided with user-specified modeling features including (when applicable):

  • GARCH, ARCH, autoregressive, and moving average orders
  • Flexible enforcement of stationarity and nonnegative conditional variance requirements
  • Pre-programmed, user controlled Boxcox data transformations
  • Error density functions (Normal, Student’s t, or skew t-distribution)

GAUSS FANPAC output includes:

  • Estimates of model parameters
  • Moment matrix of parameter estimates
  • Confidence limits
  • Time series and conditional variance matrices forecasts

FANPAC tools facilitates goodness of fit analysis including:

  • Reported Akaike and Bayesian information criterion
  • Computed model residuals
  • Computed roots of characteristic equations
  • GARCH time series data simulation
  • Andrews simulation method statistical inference
  • Time series ACF and PACF computation
  • Data and diagnostic plots including:
  • -ACF and PACF
    -Standardized residuals
    -Conditional correlations, standard deviations, and variance
    -Quantile-quantile plots

  • Residual diagnostics including skew, kurtosis, and Ljung-Box statistics

Examples

  • GARCH model with Student’s t distribution. Click here.
  • Platform: Windows, Mac, and Linux.

    Requirements: GAUSS/GAUSS Light version 10.0 or higher.