Navigation Path: Home > Statistics > Monetary and financial statistics > Monetary statistics > Historical time series

Historical time series on monetary aggregates

This section contains historical time series on euro area monetary aggregates and loans. Two groups of series are made available.

Reference series on monetary aggregates (M1, M2-M1, M2, M3-M2 and M3), loans to the private sector and net external assets, starting in January 1980 (January 1983 for loans), which are used by the ECB in its monetary analysis. These series are made available both for the reference euro area and for the Euro 11 plus Greece (i.e. also including Greece prior to its entry into the euro area).
Additional series on monetary aggregates (M1, M2-M1 and M3) covering the 1970s, which are made available for research purposes.

Data
Background

Data

Reference series on monetary aggregates (start date: 1980), net external assets (start date: 1980) and loans (start date: 1983) ( 1 )

Euro area:

Non-seasonally adjusted data: pdf 55 kB, part1 en; 47 kB, part2 en; 39 kB, part3 en; csv data ( zipped ) 35 kB, en last update 26 April 2013
Seasonally adjusted data: pdf 58 kB, part1 en; 49 kB, part2 en; 39 kB, part3 en; csv data ( zipped ) 37 kB, en last update 26 April 2013

Additional research series on monetary aggregates (start date: 1970) ( 2 )

Non-seasonally adjusted data:
pdf 191 kB, en ; csv data, 15 kB, en
Seasonally adjusted data:
pdf 211 kB, en ; csv data, 17 kB, en

( 1 ) Reference series are updated regularly following the publication of the ECB's Monthly Bulletin.
( 2 ) Additional series series are a one-off statistical product and are not regularly updated.

Background

Data refer to end-of-month outstanding stocks as consolidated at the euro area level and to the associated flows, indexes of adjusted stocks and growth rates (except for the net external assets for which only the stocks are displayed). Both non-seasonally adjusted and seasonally adjusted versions of the data are presented. Please note that, in general, the flows of the additional series made available for research purposes are derived from the difference in stocks. However, where relevant patterns associated with non-financial transactions could be identified, flows may differ from the difference in stocks.

Sources and data availability for the reference series

Three types of source have been used to derive the reference series:

from January 1999 the balance sheet of the euro area MFI sector as reported in accordance with the statistical requirements set out in ECB Regulation ECB/2008/32;
best estimates of national contributions to the euro area aggregates. These data take account, as far as possible, of the cross-border positions of MFIs within the euro area;
national monetary aggregates and net external assets, which are largely non-harmonised.

For the Euro 11, source (1) is available with the same degree of detail dating back to September 1997, while Greece has provided comparable data with this level of detail dating back to March 1998. Time series going further back in the past are constructed from sources (2) and (3) (detailed overview of data availability for the reference series).

National historical data prior to January 1999 have been converted on the basis of the conversion rates vis-à-vis the euro which were irrevocably fixed on 31 December 1998. Accordingly, Greek historical series prior to January 1999 have been converted using the GRD/ECU exchange rate as at 31 December 1998. The historical series provided by Greece for the period from January 1999 to December 2000 have been converted using the prevailing GRD/EUR exchange rates for the reference periods.

Methods used in compiling the additional research series

The basic procedure consists in the backcasting to January 1970 of the national reference series using a suitable non-harmonised series as an anchor. The suitability is verified by comparing the closeness of the reference series to the non-harmonised series over a sample covering the 1980s. Once the suitability is confirmed, the backcasting takes place by means of a preliminary linearisation of the series, i.e. filtering out the effects of level shifts, additive outliers and transitory changes by means of intervention variables. Where no anchor series was available (either because no series at all was available or because the available non-harmonised series was considered to be unsuitable), a gross-up procedure was followed. (technical notes on the procedure used)