This code compares volatility forecasts from two models (A and B). It produces the t-statistics from Diebold–Mariano–West tests of equal predictive accuracy. The null is expressed as follows H0: Loss(model A) - Loss(model B) = 0. The sign of the t-statistics indicates which forecast performs better for each loss function: a positive t-statistic indicates that model A forecast produces larger average loss than the model B forecast, while a negative sign indicates the opposite. The statistics are displayed for various values of the scale parameter b of the loss function (equation 24, page 252), chosen by the user. The cases b = 0 and b = −2 correspond to the MSE and QLIKE loss functions, respectively.
Working Paper (2018)
Müller Gernot and Hettig Thomas