Skip to contents

Load a monetary policy shock series

Source: R/mp_shock.R
mp_shock.Rd

Loads a named shock series bundled with the package as a tidy data frame with class mp_shock. Optionally filters by date range.

Usage

mp_shock(series, start = NULL, end = NULL)

Arguments

series

Character(1). Name of the series. See mp_list() for available series.

start, end

Optional Date or character (YYYY-MM-DD) filters on the returned series. If NULL, returns the full bundled span.

Value

A data frame with class c("mp_shock", "data.frame") and columns:

  • date: Date, first day of the observation month.

  • shock: numeric, the shock value in the units published by the source (see mp_source() and the per-series help files for units and scaling conventions).

  • series: character, the series identifier.

Some series carry additional columns. bauer_swanson returns mps_raw alongside shock (the orthogonalised surprise); miranda_agrippino_ricco returns info; wu_xia returns shadow_rate and effr.

Details

Aggregation. All event-study series (nakamura_steinsson, bauer_swanson, gss_target, gss_path, jarocinski_karadi_mp, jarocinski_karadi_cbi, miranda_agrippino_ricco) are bundled at monthly frequency by summing FOMC-event-level surprises within each calendar month. Months with no scheduled FOMC meeting are coded zero, matching the convention in Gertler and Karadi (2015) and the authors' own maintained releases.

Bu, Rogers, and Wu (2021, Journal of Monetary Economics 118) argue that no-meeting months should be coded NA rather than zero when estimating proxy-SVAR or LP-IV models, to avoid downward-biased variance in weak-instrument F-statistics. mpshock does not apply this adjustment; recode after loading if needed.

Scaling. Units differ across series. nakamura_steinsson is rescaled to one-year nominal Treasury-yield equivalents; bauer_swanson and most others are in raw percentage-point surprises. See each series' help file.

See also

mp_list(), mp_source(), mp_align(), mp_to_quarterly().

Examples

ns <- mp_shock("nakamura_steinsson")
head(ns)
#> <mp_shock> nakamura_steinsson [monthly, 6 obs, 2000-02-01 to 2000-07-01]
#>        date      shock
#>  2000-02-01  0.0215744
#>  2000-03-01 -0.0028683
#>  2000-04-01  0.0000000
#>  2000-05-01  0.0335943
#>  2000-06-01 -0.0061341
#>  2000-07-01  0.0000000

# Filter to a specific window
ns_gfc <- mp_shock("nakamura_steinsson",
                   start = "2007-01-01", end = "2009-12-31")