About financial ratios:
There would be the need of modifying the current model 1.0.2 to enhance the current structure for RATIOS. Below some considerations about how to approach according to ES:
every MS and public administration has its own practice; i.e. different schools for the same type of ratio, therefore it is not easy to come up with a clear and objective nomenclature different countries may interpret the same ratio in different ways. However, in the end, the ratio is adimensional, meaning there isn’t any unit measure associated to it. Possible solutions could be:
- XBRL approach: a ratio is represented as a numerator divided by a denominator in a structured and machine-oriented way. The most complex case could be that the numerator and/or the denominator are the result of a complex addition (or formula);
- Codified ratios: Use some of the ratios defined by the European Banks (see for instance the BACH system supported by 11 European Central Banks at ). This would take advantage of the support that the main countries already using them could provide. Also, these countries have databases with temporal values, percentiles, etc. The solution would consist in using the exact terminology and values used by these banks. This database is currently managed (kept) by the Bank of France;
- Self-maintenance of the ratio taxonomy: Alternatively, if any of the two previous options were appropriate, the Contracting Authority would define its own taxonomy with different codes per ratio. In this case the CA could chose the type of ratio applicable to the specific sector the CA is interested in.
Amongst the three options the second one is probably the simplest as it would take advantage of a well-proven (and free) system managed by trusted European institutions with a large experience in financial subjects.
As pointed above, the ratios are adimensionals, as they are the quocient between two amounts expressed in one single unit (the currency), but must be referred to a well defined temporal context (a period, with a start and end dates; or a specific point in time, e.g. a date). In general the temporal context will be a specific date, although for some ratios both the numerator and the denominator are "varying flows" (e.g. "profit margin"), in which case the temporal context is the period (generally an annual period, not necesarily the "natural" year) through which numerator and denominator have been measured ("net income" and "sales", for example).
From the technical perspective, none of the above cases would impact on the the XSD Schema; the impacts would be on (i) the XML instantiation; the GUIs, (ii) the business rules currently expressed in Schematron.
Proposed milestone: v1.1.0
The figure below would illustrate a data-structure that would cover any of these possibilities. Ideally if the RatioType code is used the Numerator and Denominator values should not be used (and vice-versa). This would be controlled via Schematron. However one could also consider that Numerator, Denominator and RatioType could simultaneously be specified; the con: possible inconsistencies wouldn't be so easy to validate. Let's recall you that if the RatioType is used, the code issuer agency should be specified (compulsorily) amongst the code attributes (so to identify XBRL, European Banks, the CA, other).
This proposal will have to be readjusted to match the final decisions on the above three possibilities.
![eds-financialratio](https://cloud.githubusercontent.com/assets/16134316/19927535/5cdd6186-a0fa-11e6-97fb-9629b42518b7.png)