dyn_adj: Map over Several Horizons to Extract Dynamic Adjustment...

In nelson16silva/wavcoreinf: Estimation and Evaluation of Wavelet Core Inflation Measures

Map over Several Horizons to Extract Dynamic Adjustment Parameter

Description

dyn_adj is a fast way to extract the parameter of the dynamic adjustment for a range of horizons.

Usage

dyn_adj(inf, core, H, p, ...)

Arguments

inf	A vector containing headline inflation.
core	A vector containing a measure of core inflation.
H	An integer that gives the maximum horizon.
p	An integer, lag in the regression (see details).
...	Additional parameters.

Details

For a model \pi_{t + h} - \pi_t = a_0 + \lambda_h(\pi_t - \pi^*_t) + \sum_{i = 1}^pa_i\pi_{t - i} + e_{t + h}, where \pi is the headline inflation and \pi^* a measure of core inflation, this function gives \lambda_h for h \in 1, 2 ..., H.

In core inflation's literature, parameters \lambda_h and \lambda^*_h of the following regressions are important:

• \pi_{t + h} - \pi_t = a_0 + \lambda_h(\pi_t - \pi^*_t) + \sum_{i = 1}^pa_i\pi_{t - i} + e_{t + h},

• \pi^*_{t + h} - \pi^*_t = a^*_0 + \lambda^*_h(\pi_t - \pi^*_t) + \sum_{i = 1}^pa^*_i\pi^*_{t - i} + e^*_{t + h}.

A good core inflation measure should imply \lambda_h < 0 and \lambda^*_h = 0.

Value

A tibble with the dynamic adjustment parameter and p-value for the t test of \lambda^j_h = 0 for j = ( , *). Row 1 of the tibble is h = 1, row 2 is h = 2 and so on. For estimating \lambda^*_h, put core as the first argument of the function and headline inflation as the second.

See Also

dyn_adj_est, dyn_adj_best, dyn_adj_pred

Examples

inf_head <- coreinf_br[["ipca"]]
inf_corems <- coreinf_br[["ipcams"]]
inf_coredp <- coreinf_br[["ipcadp"]]
dyn_adj(inf_head, inf_corems, 4, 5, recursive = TRUE)
out <- purrr::map(list(inf_corems, inf_coredp), ~ dyn_adj(inf_head, .x, 4, 5, recursive = TRUE))
purrr::map(out, dplyr::summarise_all, mean)

nelson16silva/wavcoreinf documentation built on Feb. 17, 2025, 7:10 p.m.