Compute Standard Errors for DCC(1,1) Parameters — dcc11_standard_errors • tsbs

Main user-facing function for SE computation. Dispatches to appropriate method based on the method argument.

Usage

dcc11_standard_errors(
  params,
  std_resid,
  weights,
  Qbar,
  distribution = "mvn",
  use_reparam = FALSE,
  method = c("hessian", "sandwich")
)

Arguments

params: MLE parameter estimates
std_resid: T x k matrix of standardized residuals
weights: T-vector of observation weights
Qbar: k x k unconditional covariance matrix
distribution: "mvn" or "mvt"
use_reparam: Logical: parameters in (psi, phi) space?
method: SE method: "hessian" (default) or "sandwich"

Value

List with: If method="hessian":

hessian: Hessian matrix of NLL
info: Observed information matrix (= Hessian for NLL)
vcov: Variance-covariance matrix (inverse of info)
se: Standard errors
eigenvalues: Eigenvalues of Hessian
eigenvectors: Eigenvectors of Hessian
condition_number: Condition number of Hessian
params: Parameters
param_names: Para,eter names
method: "hessian"

If method="sandwich":

se: Standard errors
vcov: Variance-covariance matrix (inverse of info)
bread: Bread
meatinfo: Meat
vcov: Variance-covariance matrix (inverse of info)
params: Parameters
param_names: Parameter names
method: "sandwich"

Details

For reliable beta inference in high-persistence DCC models, consider using bootstrap SEs or profile likelihood CIs from dcc_inference.R.