
On these pages, the Deutsche Bundesbank makes real-time data sets of macroeconomic indicators (subject to revision) available which can be displayed and downloaded.
Please note the following.
The Deutsche Bundesbank's real-time database is extended on a regularly basis.
The indicators are compiled in lists of indicators, which, in turn, are categorised in a hierarchical tree structure.
From the homepage, you can search this tree for the desired list of indicators. Click to open the list of indicators.
The indicators in the list are linked to a page with a matrix view. Each indicator has a link which allows you to display the explanation of the indicator key, download the indicator directly and add it to the shopping cart. In addition, the number of existing vintages and the publication dates of the first and the last vintages are displayed.
Both the indicator key and the path are displayed for each indicator.
The number of existing vintages and the dates of the first and the last vintages are also displayed.
Other information that is available are details on the source and metadata sheets in PDF format.
This page has a link to the explanation of the indicator key and attributes (click
symbol),
to enable you to add items to the shopping cart, to download items directly and to access the help function.
When you first open the page, the "Display options" section shows only the current vintages. To see further vintages, you can use the "Display calculations from-to" entry fields (entry format: yyyy-mm-dd). The number of columns that can be initially displayed or displayed at any one time may change.
The vintages are displayed as columns in the table, so that, side by side, these vintages form a matrix. The column heading corresponds to the publication date of a vintage and is displayed in the format yyyy-mm-dd.
The row heading corresponds to the observation period.
Values displayed in bold are new values or changed values since the last calculation.
In order to improve readability of the value table in the case of very extensive number of reporting periods, related years are visibly grouped and can, if required, be displayed individually using the "Selection of the observation period for the table" function.
Reporting periods and vintages in the table are displayed in descending order, ie the top left corner shows the most recent period result of the latest vintage. This can be reversed by changing the sorting order to "ascending" so that the most recent period result of the latest vintage is displayed in the bottom right-hand corner.
Vintage-related attributes can be displayed below the caption row by clicking the "+" symbol in the table.
The vintages are displayed as columns in the table, so that, side by side, these vintages form a matrix. The column heading corresponds to the publication date of a vintage in the format yyyy-mm-dd.
The row heading corresponds to the observation period.
The following attributes are specified for a vintage.
| Data | Meaning | Possible values |
|---|---|---|
| Unit | Unit in which the vintage values are specified | eg DM, euro, DM/euro, 1995=100 |
| Dimension | Factor by which the data must be multiplied in order to attain the actual values | Ten, hundred, thousand, ..., billion |
| Recording method | Data recording method | Computerised (C), External source (E) or Manual (M) |
| Base year | Year for indices or prices | eg 1991, 2005; ... |
| Timespan | First and last dates entered | eg from 2000-01 to 2000-12 |
Vintage values are displayed with figures grouped into sets of three (ie 123,456.7 instead of 123456.7).
The value table can be copied to Excel using the cut-and-paste function.
A small window opens. The indicator key, path and attributes are displayed for the indicator. The number of existing calculations and the publication dates of the first and the last calculations are also displayed.
This is followed by a table of the key components and a table of the attributes together with their values and the corresponding meaning.
The top part of the page shows the text search, while the bottom part shows the search via indicator key components.
The term that is searched for (eg "prices") is written in the text field. A term can be entered with initial or final spaces and in upper or lower case. The search is started by pressing the "Start text search" button.
The text search looks for the term in the text descriptions of the key codes. If the term should exist for a code, a hit list is shown containing the indicators bearing this code in the key.
If you enter several terms, the terms do not all have to appear in the same code. For example, when searching for "monthly prices", the term "monthly" appears in code M of the second key position and the term "prices" in several codes of the fourth key position. The search displays all time series that have these codes in the respective key positions.
The text search function searches in the key components only, not in the attributes.
This page shows the key structure of the real-time database. The eight indicator-related key components are displayed in the pre-defined order. A text field into which the search criteria can be entered is offered for each key component.
The key components "Statistical accounting system, indicator category, indicator identification" are highlighted. These three key components form a segment with a hierarchical structure. The best procedure for assigning these key components is from top to bottom.
Only numbers, letters, underscores, spaces and commas are accepted as permissible characters. A comma is the separator between individual codes in the entry. A code can be entered with initial or final spaces and in upper or lower case.
After each text field that is still empty, it is stated how many codes are used for this key component in existing indicators. This number depends on the pre-defined search criteria for the other key components.
Example: Without search criteria, there are three codes for "Area (country, country groups)". If the search criterion for the frequency is "M", there is only one code for "Area (country, country groups)". If the search criterion for the frequency is "D", there are zero codes for all other search criteria, since there is no indicator for a frequency of "D".
Moreover, the number of indicators (absolute and as a percentage of the whole) found for the current search is displayed.
Comment: All numbers are updated whenever the search criteria are changed. A change is made by pressing the "Reset search" button, by returning from the code assistant or after manually changing a text field and exiting.
By pressing the
symbol after a text field, all existing codes are displayed on a separate page (see code assistant) and made available for selection.
Each text field is followed by the
symbol. By pressing this button, the corresponding text field is locked and the
symbol is displayed. The content of the text field can then no longer be deleted by pressing the "Reset search" button. The
symbol is used to unlock the text field again. The content can then be deleted again.
The "Reset search" button can be used to delete all the codes that have already been entered.
The page offers an "Activate/deactivate autocomplete" mode. If this is activated and only one code exists for the key component, the system enters this code automatically. This mode is deactivated by default. This function can be "activated" or "deactivated" by clicking the link (link text changes accordingly).
The search can be started by pressing the "Start key search" button.
This page displays the name and the description for a selected key component as well as all codes (in alphabetical order) with their descriptions.
By default, only those codes are displayed that have indicators in the given selection (ie dependent on the already determined search criteria for the other key components). (In theory, these can also be no codes!)
You can display all theoretically possible codes, ie those maintained in the Bundesbank code list, via the "Display all codes" link (the link text then changes to "Display existing codes only").
If a code is displayed without a description, this means that the code is not defined in the code list.
Each code is preceded by a checkbox field which can be used to select or discard the code as search criterion. If a search criterion already exists when opening the page, the checkboxes are marked accordingly.
By pressing the "Apply" button (search conditions are applied) or the "Cancel" button (search criteria are not applied), you are directed back to the "Search" page. Here, the current search criterion is entered in the corresponding text field.
The indicators can be downloaded in CSV and Excel format. CSV is a text-based format which can be read by conventional spreadsheet programs such as Excel. In CSV format, there is one download file for each sheet; in Excel format, the sheets belonging to a download are made available in an Excel folder.
You can download the data matrix of any indicators directly by clicking the "CSV" link in the list of indicators or in the HTML display of the indicator.
Alternatively, you can download up to ten indicators onto your hard drive at any one time. If your browser is Javascript-enabled and accepts cookies, you can also make use of our shopping cart function. To do this, select the desired indicators for download by clicking the symbol in the "Shopping cart" column in the indicator list view.
Then navigate to the download form. There you will find that the selected indicators have already been entered. The entries are linked to the individual indicators. Each indicator has a link to the explanation of the indicator key
.
The content of the list can be resorted as required. By pressing the
symbol, the respective indicator is moved downwards, whereas the
symbol moves the respective indicator upwards.
The individual indicators can be removed from the list by clicking the
symbol. Another way of removing the indicators from the list is to activate the checkboxes preceding the unwanted indicators and then press the "Remove selected items" button.
The download is started by pressing the "Proceed to direct download" button.
The "Download in matrix format" or "Download in snapshot format" buttons direct you to the "Download options" page.
Owing to the size of the source files, it may not be possible to compile the bundle. If this is the case, use the links to the individual files or set a limit on the download period.
Depending on the selected format, this page shows the options for a matrix or snapshot download.
The standard file format is CSV and the standard language is the language of the page.
You can choose between English CSV (entries are separated by commas, a period is used as the decimal separator) or Excel, or German CSV (entries are separated by semicolons, a comma is used as the decimal separator) or Excel. The standard and default language for the direct download is the language that you have chosen for the display. Please note that the language you select determines whether your Excel program displays the file correctly after opening it.
Carry out the following steps to set the language of the download.
In the "Reporting periods (rows)" section, you can set a limit on the period for which the indicators are to be displayed (entry format: yyyy-mm). These will then appear as rows in the resulting CSV or Excel file.
The standard and default setting for the download is the entire period available.
The "Adjust limits" option can be selected only if a period was selected.
In the "Publication dates (columns)" section, you can set a limit on the period for displayed vintages (entry format: yyyy-mm-dd). Each vintage within the set time period will appear as a column in the CSV or Excel file that is generated.
The standard and default setting for the download is the entire period available.
The "Adjust limits" option can be selected only if a period was selected.
The publication practice of statistical data is defined as "regular" if the publication frequency is about the same as the length of the reporting period. In the matrix view of the real-time data of an indicator, the regular publication scheme ideally implies a triangular structure at the bottom end.
The triangular structure is beneficial for numerous applications. However, it results automatically from the output format "All" only in an ideal situation, ie
An exception to 1) are, for example, the (monthly) retail sales, which are published twice a month by the Federal Statistical Office.
An exception to 2) are, for example, the orders received that could not be published on a regular basis in the first half of 2006 owing to strikes at statistical offices in the federal states.
In case 2), it is important to note that the triangle structure can be created conceptually by this output format, but not in reality, ie there will be "steps" at the bottom end of the issued data matrix.
In the data matrix, the omitted regular publication is generally documented by an empty column with the heading
However, it may be helpful to have the most recent available data record in the basically empty column. For this reason, you can choose the option "Insert previous vintage".
Snapshot time entries must be made in the format yyyy-mm-dd.
In the output file, you receive the indicator with the vintage that was available at the snapshot time.
For each indicator of the export list, the vintage that was available at the given date is selected. These are issued next to each other on a worksheet. There is therefore one sheet for each snapshot time that was specified in the download options.
The worksheet name of the Excel file and the name of the CSV file correspond to the given snapshot date and has the format yyyy-mm-dd.
The complete time series key appears as column heading. The row description corresponds to the observation period and has the format yyyy-mm-dd or yyyy-mm-dd
For every calculation, the attributes are issued at first, followed by the values in ascending order.
In the snapshot format, it may occur that different frequencies are written next to each other on a single worksheet. In this case, the vintages share the date column and are always given in the format yyyy-mm-dd. The date column is geared to the lowest available frequency. As a rule, the observation of an indicator with a higher frequency is entered in the first sub-period.
The time series of the real-time database are endowed with systematic keys that meet the Gesmes/TS key standard and enables systematic navigation and searches.
The key has the following structure:
| Key components | Example | Designates | Description |
|---|---|---|---|
| Frequency | M | Indicator | Monthly |
| Area (Countries, country groups) | DE | Indicator | Germany |
| Adjustment indicator | N | Indicator | Neither seasonally or working day adjusted |
| Acccounting system | I | Indicator | Business cycle and labour market indicators by branches |
| Indicator category | IP1 | Indicator | Production |
| Indicator identification | ACM01 | Indicator | Industry |
| Price or unit reference | C | Indicator | In constant prices |
| Calculation basis | I | Indicator | Index |
| Publication date: year | 2005 | Vintage | 2005 |
| Publication date: month | 11 | Vintage | November |
| Publication date: day | 07 | Vintage | 07 |
The key for an indicator consists of key components 1-8, which are written in succession and separated by a dot (there is an analogous time series key consisting of components 1-11).
consist of the first eight key components.
Example: M.DE.N.I.IP1.ACM01.C.I
Example: M.DE.N.I.IP1.ACM01.C.I.2008.01.09
Each time series has attributes that are not part of the time series key.
| Source | eg "Federal Statistical Office" |
| Methodology | Description of the indicator |
| Unit | Unit in which the vintage values are specified |
| Dimension | Factor by which the data must be multiplied in order to attain the actual values |
| Decimals | Number of decimal places |
| Recording method | Data recording method |
| Base year | Year for indices and prices |