Using mrwd(x) one can fit a multivariate Random Walk with drift to x.
It would be nice to be able to fit a multivariate Random Walk model too (without drift). Such a naive model is useful in performance evaluation of in-sample forecasts when used as a benchmark.
Maybe adding an argument for specifying if a drift is wanted or not cold be a good extension?
Using
mrwd(x)one can fit a multivariate Random Walk with drift tox.It would be nice to be able to fit a multivariate Random Walk model too (without drift). Such a naive model is useful in performance evaluation of in-sample forecasts when used as a benchmark.
Maybe adding an argument for specifying if a drift is wanted or not cold be a good extension?