Monetary policy factors

This page contains the monetary policy factors constructed using the procedure in Altavilla, Brugnolini, Gürkaynak, Motto, and Ragusa (2019) (ABGMR).

The factors are extracted from the Euro Area Monetary Policy Event-Study Database and rotated to provide information about different channels of conventional and unconventional monetary policy. A detailed discussion of the factors rotation and the identification strategy is in the appendix of (ABGMR).

These three vox.org articles might also be useful:

  1. The Euro Area Monetary Policy Event-Study Database

  2. Monetary policy in action: Multiple dimensions of ECB policy communication and their financial market effects

  3. Financial market reactions to monetary policy signal

Downloads

The factors are in two distinct csv files:

  • press_release_factors_YYYY-MM-DD.csv

  • press_conference_factors_YYYY-MM-DD.csv

The suffix YYYY-MM-DD in the filename denotes the last GC’s surprises used in the factor extraction.

Download
Press release Press conference Target Timing FG QE
2021-12-16 2021-12-16
2021-10-28 2021-10-28
2021-09-09 2021-09-09
2021-07-22 2021-07-22
2021-06-10 2021-06-10
2021-04-22 2021-04-22
2021-03-11 2021-03-11
2021-01-21 2021-01-21
2020-12-10 2020-12-10
2020-10-29 2020-10-29
2020-09-10 2020-09-10
2020-07-16 2020-07-16
2020-06-04 2020-06-04
2020-04-30 2020-04-30
2020-03-12 2020-03-12
2020-01-23 2020-01-23
2019-12-12 2019-12-12
2019-10-24 2019-10-24
2019-09-12 2019-09-12
2019-07-25 2019-07-25
2019-06-06 2019-06-06
2019-04-10 2019-04-10
2019-03-07 2019-03-07
2019-01-24 2019-01-24
2018-12-13 2018-12-13
2018-10-25 2018-10-25
2018-09-13 2018-09-13

Notes

  1. The factors used by ABMGR are:

    There is a difference between the factors used in ABMGR and the one you can download from this page. The difference is due to a data revision. The 1-month surprise for the Governing Council of September 4, 2014, has been revised to reflect a stale quote. This revision appeared first in the EA-MDP of January 2022.

  2. The factors are normalized as in ABMGR. More specifically, the target and the timing factors are normalized to have unit effect on the 1-month and 6-month OIS surprises, respectively; the fg and qe factors are normalized to have a unit effect on the 2-year and the 10-year yields, respectively.