Compute Coefficient of Variation for Bootstrap Outputs

Calculates the coefficient of variation (CV) for each dimension of bootstrap functional outputs, providing a measure of estimation stability.

Usage

compute_func_out_cv(
  func_outs,
  names = NULL,
  cv_thresholds = c(Stable = 0.3, Moderate = 0.6)
)

Arguments

func_outs

List of functional outputs from tsbs(), or a matrix.

names

Optional character vector of names for output dimensions.

cv_thresholds

Named numeric vector with thresholds for stability classification. Defaults to c(Stable = 0.3, Moderate = 0.6).

Value

A data frame with columns:

Name

Dimension name

Mean

Bootstrap mean

SD

Bootstrap standard deviation

CV

Coefficient of variation (SD/Mean)

Stability

Stability classification based on CV thresholds

Details

The coefficient of variation (CV) is defined as SD/Mean. Lower CV values indicate more stable estimates. Default thresholds classify estimates as:

• Stable: CV < 0.3

• Moderate: 0.3 <= CV < 0.6

• Unstable: CV >= 0.6

Examples

if (FALSE) { # \dontrun{
# After running tsbs() with a portfolio function
result <- tsbs(x, bs_type = "ms_varma_garch", func = risk_parity_portfolio, ...)

# Assess stability of bootstrap weights
stability_df <- compute_func_out_cv(result$func_outs, names = c("SPY", "EFA", "BND"))
print(stability_df)
} # }