Replication Codes

This repository contains the codes used to replicate the results in Marques and Trucíos (2026).

Empirical Application

• Empirical_Application.R computes one-step-ahead conditional variances and one-step-ahead VaR and ES for the daily returns series of the constituents of the Dow Jones Industrial Average Index.
• Tables_App.R Performs the Model Confidence Set procedure for the empirical application results (also generates the Tables 5 in the main manuscript and 10 in the Supplementary Material).
• Tables_App_VaR_ES.R Performs the calibration tests and apply the MCS to the scoringh functions for the VaR and ES (also generates Table 6)

Five-minute realized variances are freely available from the CaPiRe database. Daily returns were obtained from Economatica.

Monte Carlo Simulation

• MonteCarloSimulation.R runs the one-step-ahead forecasting experiment. To use the code, modify the parameters accordingly, or execute it in batch mode using, for instance, the following command:
  R CMD BATCH "--args GARCH-N BR" MonteCarloSimulations.R MonteCarlo_GARCH-N_BR.txt &
  (You can change BR to US, FALSE to TRUE.)
• Tables_MonteCarlo.R reproduces the results shown in Tables 2 and 3 of the main manuscript as well as Tables 4 - 9 in the Supplementary Material
• Aux_MonteCarlo.R reproduces Table 1 in the Supplementary Material.
• MonteCarloSimulations_Larger_Sample reproduces Table 2 in the Supplementary Material.

Auxiliary Functions

• DGPs.R defines the data-generating processes used in the simulations.
• Utils_GARCH-GAS-SV.R and utils.cpp contain additional functions for model estimation and forecasting.
• Descriptive_Statistics displays the descriptive statistics in Table 4.

References

Marques, F. and Trucíos C. (2026). "Daily Volatility Forecasting with Off-the-Shelf Models: A Comparative Study Under Stress". Submitted