Modeling Financial Time Series with S-PLUS®, Second Edition


S+FinMetrics™ 2.0

Description | Functions | Webcasts | Purchase | Support

Description of S+FinMetrics™

S+FinMetrics™ is an S-PLUS module conceived by Andrew Bruce, Doug Martin, Jiahui Wang, Eric Zivot and developed by Insightful Corporation which provides advanced analytic-rich software for modeling, analyzing, and visualizing financial market data. The software offers the most comprehensive, modern, and flexible analytic tool available for precise, predictive econometric modeling of financial time series. Version 1.0 of the module (which operates under S-PLUS 6) includes the following functionality:

  • Easy-to-use Trellis plots for multivariate time series

  • Time series manipulations such as missing value interpolation, disaggregation, differences, distributed lags and polynomial distributed lags

  • Rolling sample statistics such as variance, maximum, and minimum

  • Moving average operators for both regularly spaced and irregularly spaced time series

  • Common technical analysis measures and indicators

  • Statistical tests for normality, autocorrelation, heteroskedasticity, multicollinearity, GARCH effects, and long memory

  • Extreme value theory models based on generalized extreme value and generalized Pareto distributions as well as copulas

  • Autoregressive distributed lag regression models

  • White and Newey-West corrections for heteroskedasticity and serial correlation

  • Robust estimation of REG-ARIMA models and robust detection of level shifts, trend breaks, and outliers

  • Rolling and recursive regression

  • Generic rolling models for back-testing

  • Long memory fractional ARIMA and SEMIFAR models

  • Univariate GARCH models including long memory FIGARCH and FIEGARCH models

  • Multivariate GARCH models

  • Linear and nonlinear systems of regression equations

  • Classical and Bayesian vector autoregression models

  • Tests for unit roots and cointegration

  • Vector error correction models

  • State space models and efficient estimation, prediction, smoothing, and simulation using the Kalman filter

  • Statistical multifactor models for large data sets based on asymptotic principal components

  • Term structure interpolation

New features in version 2.0 of S+FinMetrics™ (which runs under S-PLUS 7.0) include:

  • Variance ratio tests, efficient unit root tests and tests for nonlinearity

  • Threshold AR, smooth transition AR and Markov switching AR models as well as Markov switching state space models

  • Simulated solutions to systems of stochastic differential equations

  • Generalized method of moments estimation

  • Gallant and Tauchen's semi-nonparametric conditional density estimation and efficient method of moments estimation

S+FinMetrics™ incorporates functions from S+GARCH, the EVIS library of functions for modeling extreme values created by Alexander McNeil, the EVANESCE library of functions for modeling extreme values and bivariate copulas created by Rene Carmona and Julia Morrison, the SsfPack C library of state space modeling functions created by Siem Jan Koopman, and the SNP and EMM FORTRAN libraries created by Ronald Gallant and George Tauchen. S+GARCH was originally developed by Zhuanxin Ding, Hong-Ye Gao, Doug Martin, Jiahui Wang and Yihui Zhan. The S+FinMetrics™ function arima.rob was written by Ana Bianco, Marta Garcia Ben, Elena Martinez and Victor Yohai. The S+FinMetrics™ long memory modeling functions FAR, FARIMA, SEMIFAR and fgarch were developed by Jan Beran, Andrew Bruce, Don Percival, Alan Gibbs and Jiahui Wang and supported by NSF grant DMI-9801614 to Insightful Corporation (formerly MathSoft, Inc.). Much of the new functionality in version 2.0 of S+FinMetrics™ was supported by the NSF SBIR Phase II grant DMI-0132076 to Insightful Corporation. The S-PLUS implementation of Gallant and Tauchen's SNP and EMM FORTRAN libraries was accomplished by Jiahui Wang, Bob Thurman, Michael Sannella, Ying Gu and Eric Zivot, with the generous help and support of George Tauchen. Hu McCulloch kindly provided the term structure data included with S+FinMetrics™, and James MacKinnon provided data sets for the response surface critical values for the Dickey-Fuller and Phillips-Ouliaris distributions.

Complete Function List

  • S+FinMetrics™ 1.0 function reference manual.

  • New features in S+FinMetrics™ 2.0 include:

    • improved coupla functionality (numerical hessian computed for fit.copula, compare fitted copulas using information criteria, improved method functions for copula objects)

    • efficient unit root tests, variance ratio tests, tests for nonlinearity (BDS, Tsay threshold, Hansen threshold), tests for serial correlation in regression residuals

    • improved state space model functionality (numerical hessian and sandwich hessian computed from new function SsfFitFast, improved method functions, support for Markov switching in transition equation system matrices, support for affine term structure components, Kim's approximate filter for Markov switching state space models)

    • nonlinear regime switching models (threshold AR models, smooth transition threshold models, Markov switching AR models, Markov switching state space models.

    • numerical simulation of univariate and multivariate stochastic differential equations (Euler's methods, strong order 1 scheme, weak order 2 scheme)

    • generalized method of moments estimation of linear and nonlinear models with the new gmm function (allows user specified weight matrices, efficient weight matrices robust to heteroskedasticity and autocorrelation, continuous updating estimation)

    • Gallant and Tauchen's semi-nonparametric (SNP) conditional density models and efficient method of moments estimation (automatically fit SNP models, implement EMM using user specified simulator functions without re-compiling the Gallant and Tauchen  FORTRAN code)


The following are a list of archived Insightful webcasts by Eric Zivot that illustrate the functionality and use of S-PLUS and S+FinMetrics™:

The following are a list of archived Insightful webcasts by industry professionals and academics that highlight the use of S-PLUS and S+FinMetrics™ for financial applications:

How to Purchase S-PLUS and S+FinMetrics™

Insightful Support Pages

Last updated: March 20, 2006

home | news | contents | scripts | errata | additions | S+FinMetrics | order