Apart from defining the data, XBRL brings power to data by establishing the linkages among specific data points. When XBRL data is read, we not only see the meaning of a data point say Cash but the composition of cash like balance in hand, with the bank or its position in the balance sheet etc., can also be known. Using these interrelationships between the data points, it is possible to validate the data being reported and hence enhance the data quality.
Formula Linkbase over Calculation Linkbase
For quite some time, the accuracy was limited to basic arithmetic checks i.e. addition and subtraction via calculation linkbase. With formula linkbase, many more checks and rules can be built. This can help in getting cleaner and better data. The rules can be mathematical, logical, or user-defined. Without getting into technical details, let us briefly understand what kind of checks can be built using the formula linkbase.
Additional data checks offered by Formula Linkbase for data analysis
The rules built in the formula linkbase can either be for the reported data that is included in the final XBRL document or for deriving new values. In the latter case, the values which are required for derivation should be present in the final XBRL document. As observed some of the most common types of rules modeled using the formula linkbase are
4. Mathematical operations especially other than the ones used in calculation linkbase E.g. EPS = Profit/Number of shares, Income tax = Profit * Rate of tax
5. Reconciliation of beginning and ending balances. E.g. Opening balance of cash + changes during the period = closing balance of cash.
6. Values provided for break-downs defined as dimensions, aggregate to the total value
Formula linkbase requires the final XBRL document as input. The rules and checks or even deriving new values are based on the data that gets reported in the final XBRL document. The output obtained after the processing of the formulae on the XBRL document could be either Boolean (i.e. true if the rule is passed and false if the rule is failed) or a newly derived value. There is a provision to define error messages. However displaying the result of validation (i.e. formatting, showing only “false” results, etc.) is a function of the application which is used to either generate the final XBRL document or validate the final document.
The trend seen so far has been creating rules based on values provided in a single instance document. The formula specification however has provision to model rules to check data across multiple instance documents. The time for validating data depends on the number of rules and the size of the instance document.
IRIS has more than 11 years of experience in XBRL taxonomy and Implementation. We have created taxonomies for various reporting domains such as prudential reporting for banking and insurance, business registrar requirements, and Islamic reporting requirements for banks in the Middle East. Our consultants have extensive experience in working with financial taxonomies such as IFRS and US-GAAP, regulatory reporting taxonomies such as CRD IV & Solvency II taxonomies of the EU, and non-financial taxonomies such as Global Reporting Initiative or Corporate Actions. You can visit here to know more about our XBRL Taxonomy Services.