Skip to contents

Run Monte Carlo Study for CGARCH(1,1) Estimation with ADCC Support

run_cgarch_monte_carlo.Rd

Performs a Monte Carlo simulation study to assess the accuracy of Copula GARCH parameter estimation. Computes bias, RMSE, and coverage probabilities for DCC/ADCC correlation dynamics parameters.

Usage

run_cgarch_monte_carlo(
  n_sim = 100,
  n_obs = 500,
  k = 2,
  true_alpha = 0.05,
  true_beta = 0.9,
  true_gamma = NULL,
  omega = NULL,
  alpha_garch = NULL,
  beta_garch = NULL,
  copula = "mvn",
  true_shape = 8,
  shape = NULL,
  transformation = "parametric",
  confidence_level = 0.95,
  verbose = TRUE,
  seed = 12345
)

Arguments

n_sim

Number of simulation replications

n_obs

Number of observations per replication

k

Number of series (default 2)

true_alpha

True DCC alpha parameter

true_beta

True DCC beta parameter

true_gamma

True ADCC gamma parameter (default NULL for standard DCC)

omega

Vector of GARCH omega parameters (default: rep(0.05, k))

alpha_garch

Vector of GARCH alpha parameters (default: rep(0.10, k))

beta_garch

Vector of GARCH beta parameters (default: rep(0.85, k))

copula

Copula distribution ("mvn" or "mvt")

true_shape

True shape parameter for MVT copula (default 8)

shape

Alias for true_shape (for backward compatibility)

transformation

PIT transformation type ("parametric", "empirical", "spd")

confidence_level

Confidence level for coverage (default 0.95)

verbose

Print progress

seed

Base seed for reproducibility

Value

List with:

estimates

Matrix of estimates (n_sim x n_params)

std_errors

Matrix of standard errors (n_sim x n_params)

bias

Bias for each parameter

rmse

RMSE for each parameter

coverage

Coverage probability for each parameter

persistence

Vector of persistence values per replication

summary

Summary data frame