Statistical processing

Source data

The data sources for financial accounts are prepared in a collaboration between Statistics Denmark and Danmarks Nationalbank, where each institution is responsible for preparing a number of sources. Overall, Statistics Denmark is responsible for preparing sources based on the companies' accounts, general government as well as information on households, while Danmarks Nationalbank contributes with a number of primary statistics that describe the various parts of the financial sector in detail.

The most important sources are:

· Insurance and pension statistics (prepared at Danmarks Nationalbank) · Investment fund statistics (prepared at Danmarks Nationalbank) · Statistics on monetary financial institutions (prepared at Danmarks Nationalbank) · Securities statistics (prepared at Danmarks Nationalbank) · Balance of payments statistics (prepared at Danmarks Nationalbank) · Accounting data block (prepared in Statistics Denmark) · Enriched accounting data block (prepared at Danmarks Nationalbank) · Information on cooperative housing (prepared in Statistics Denmark) · Information on employee options (prepared in Statistics Denmark) · Information from the wealth statistics (prepared in Statistics Denmark) · Information on general government (prepared in Statistics Denmark) · Information on the net lending / net borrowing calculated in the non-financial sector accounts (prepared in Statistics Denmark)

Frequency of data collection

Data for the statistics are collected quarterly with the exception of a few sources that are collected annually. Annual data comes i.a. from accounting statistics, income statistics and digital annual reports (XBRL data).

Data collection

Data for these statistics is provided through internal deliveries.

Data validation

Data validation takes place in several phases, as part of the data processing:

1. The first phase takes place even before we receive data from our primary statistics, as input data has already been validated and quality assured by the suppliers
2. The next phase occurs when we have received data in the form of an additional validation and transformation of input data. Transformation means that input data is classified and coded according to ESA 2010.
3. In the third phase, corrections are made to our primary statistics, whose errors have been identified during phase 2. After which we return to phase 1 and the process is re-run.
4. In the fourth phase, a data validation of the sources in relation to each other takes place. As financial accounts are a reconciled system, the internal consistency in itself acts as quality control in the ongoing production, as different sources with the same information can be compared.
5. In the final phase, corrections may be added if this cannot be implemented in our primary statistics in the short term. Subsequently, these corrections will be implemented in the primary statistics in connection with the next revision of these.

Data compilation

The processing of data for these statistics consists primarily of integrating data from the many data sources, so that a harmonized data set is obtained, where the collected data is transformed into a coherent financial accounts system. Danmarks Nationalbank's and Statistics Denmark's joint system for financial accounts is therefore built around a set of data sources in a uniform format. The general data processing consists of three phases: preparation of sources, quality assurance and reconciliation.

** Preparation of sources ** In the preparation phase, bids are prepared for all sources at Danmarks Nationalbank and Statistics Denmark, respectively. The responsibility for the preparation of the individual source blocks can be seen in the Sources section. In addition to data validation, the process consists of transforming the individual sources into a uniform format. Prepared source blocks are exchanged between the institutions.

** Quality assurance ** In financial accounts, a source prioritization is made for each combination of asset sector, liabilities sector and instrument based on a balancing of the quality of the individual sources. Prioritization of sources is necessary because there may be multiple source bids on the same crossing, but no source has bids on all crossing. As an example, there is a crossing that describes the banks' total lending for insurance and pensions. For this crossing, the statistics on monetary financial institutions (MFIs) will have a bid on the size of their lending and the insurance and pension statistics will have a bid on the size of the loan they have taken out at MFI institutions. If these two sizes are not the same, the source prioritization will determine which bid to prioritize. The quality assurance phase is an iterative process where there will be the possibility of corrections to the primary statistics, which are discovered by putting the primary statistics opposite each other on each individual cut.

Reconciliation Source prioritization alone cannot ensure consistent numbers. Since not all sources are consistent for each source bid, the sums from the prioritized sources may differ from the sum in input from the primary statistics. The sum means the total of either the asset or liability side of each instrument. To ensure consistency to selected totals, a residual calculation is also made. The residuals are calculated as the difference between the reconciled source bids and the sum of source bids from the source being voted against. The residuals are then distributed to selected sectors. The distribution is determined via a predetermined voting table and will be within the same instrument or transferred to another instrument.

After consistency is ensured in connection with the source prioritization, a check of current errors is made for all cuts. A stream contains errors if the stream identity: primo value + streams = end value is not met. That a current identity is not fulfilled may be due to a method change in the source statement, current errors originating from the primary source or that the current error is formed on the basis of the source prioritization and flow of residuals. If there is a power failure, the transactions are calculated residually so that the identity holds.

Finally, the financial part of the national accounts is reconciled with the net acquisition of receivables from the non-financial national accounts.

Adjustment

No corrections are made to data beyond what has already been described during data validation and data processing.