Skip to contents

Estimates an unbiased VAR(1) using stochastic approximation (Bauer, Rudebusch and Wu, 2012)

Source: R/BiasCorrection.R
Bias_Correc_VAR.Rd

Estimates an unbiased VAR(1) using stochastic approximation (Bauer, Rudebusch and Wu, 2012)

Usage

Bias_Correc_VAR(
  ModelType,
  BRWinputs,
  RiskFactors,
  Economies,
  FactorLabels,
  GVARinputs = NULL,
  JLLinputs = NULL,
  verbose = TRUE
)

Arguments

ModelType

character. Model type to be estimated. Permissible choices: "JPS original", "JPS global", "GVAR single", "JPS multi", "GVAR multi", "JLL original", "JLL No DomUnit", "JLL joint Sigma".

BRWinputs

list. Contains the necessary inputs for the BRW model estimation:

  1. Cent_Measure: "Mean" or "Median" (unbiased estimation type)

  2. gamma: Numeric. Adjustment parameter between 0 and 1. Default is 0.5.

  3. N_iter: Integer. Number of iterations for the stochastic approximation algorithm after burn-in. Default is 5000.

  4. N_burn: Integer. Number of burn-in iterations. Default is 15

  5. B: Integer. Number of bootstrap samples per iteration for calculating the noisy measure of the biased estimator's mean or median. Default is 50.

  6. check: Logical. Indicates whether to perform a closeness check. Default is TRUE.

  7. B_check: Integer. Number of bootstrap samples for the closeness check. Default is 100000.

  8. Eigen_rest: Numeric. Restriction on the largest eigenvalue under the P-measure. Default is 1.

RiskFactors

numeric matrix (Td x K). Time series of risk factors.

Economies

character vector. Names of the C economies included in the system.

FactorLabels

list. Labels for all variables in the model.

GVARinputs

list. Inputs for GVAR model estimation (see GVAR). Default is NULL.

JLLinputs

list. Inputs for JLL model estimation (see JLL). Default is NULL.

verbose

logical. Flag controlling function messaging. Default TRUE.

Value

Bias-corrected VAR parameters based on the framework of Bauer, Rudebusch and Wu (2012). The list contains:

  1. KOZ_BC: estimated intercept (K x 1);

  2. K1Z_BC: estimated feedback matrix (K x K);

  3. SSZ_BC: estimated variance-covariance matrix (K x K);

  4. dist: root mean square distance (scalar);

General Notation

  • Td denotes the model time series dimension.

  • C number of countries in the system.

  • K denotes the total number of risk factors.

References

Bauer, Rudebusch and, Wu (2012). "Correcting Estimation Bias in Dynamic Term Structure Models"
This function offers an independent R implementation that is informed by the conceptual framework outlined in Bauer, Rudebusch and Wu (2012), but adapted to the present modeling context. Related Matlab routines are available on Cynthia Wu's website (https://sites.google.com/view/jingcynthiawu/).

Examples

# \donttest{
data(RiskFacFull)
Factors <- t(RiskFacFull[1:7, ])

BRWinputs <- list(
  Cent_Measure = "Mean", gamma = 0.4, N_iter = 1000, N_burn = 100,
  B = 10, check = 1, B_check = 5000
)

Economies <- "China"
N <- 3
ModelType <- "JPS original"
FactorLabels <- NULL

BRWpara <- Bias_Correc_VAR(ModelType, BRWinputs, Factors, Economies, FactorLabels, verbose = FALSE)
# }