This function computes RMSFE (Mean Squared Forecast Error Relative to the Benchmark)
rmsfe(x, pred_bench, y, ...)# S3 method for predbvhar
rmsfe(x, pred_bench, y, ...)
# S3 method for bvharcv
rmsfe(x, pred_bench, y, ...)
RMSFE vector corresponding to each variable.
Forecasting object to use
The same forecasting object from benchmark model
Test data to be compared. should be the same format with the train data.
not used
Let \(e_t = y_t - \hat{y}_t\). RMSFE is the ratio of L2 norm of \(e_t\) from forecasting object and from benchmark model.
$$RMSFE = \frac{sum(\lVert e_t \rVert)}{sum(\lVert e_t^{(b)} \rVert)}$$
where \(e_t^{(b)}\) is the error from the benchmark model.
Hyndman, R. J., & Koehler, A. B. (2006). Another look at measures of forecast accuracy. International Journal of Forecasting, 22(4), 679-688.
Bańbura, M., Giannone, D., & Reichlin, L. (2010). Large Bayesian vector auto regressions. Journal of Applied Econometrics, 25(1).
Ghosh, S., Khare, K., & Michailidis, G. (2018). High-Dimensional Posterior Consistency in Bayesian Vector Autoregressive Models. Journal of the American Statistical Association, 114(526).