1: The XBRL language
1.1) Origins The application of computer science to business and accounting sciences has had one of its latest advances in the XBRL language. Known are the errors and waste of time generated by the re-typing of the Financial Statements issued by the different entities, either for their subsequent analysis or information filing. For this reason, in 1998 Charles Hoffman developed the XBRL language, starting from his work with the XML language. The acronym XBRL is the acronym for eXtensible Business Reporting Language (Extensible Business Reporting Language). The underlying idea of this initiative was none other than to standardize the format with which financial information is distributed among the different providers and consumers.
The standard is administered by an international non-profit consortium (XBRL International Incorporated) made up of approximately 450 organizations, including regulators, government agencies, consultancies and software developers.
XBRL aims to standardize the format of business and financial information that circulates digitally.
For this, it is based on the definition of taxonomies, a set of metadata that describes the data to be reported, the format and structure they have, as well as the relationships between said data.
Technically these taxonomies are XML schemas, which must comply with the standards established by the XBRL specification, published by XBRL International, the version is currently in force.
1.1 of that specification
On the other hand, the data that we want to report, that is, the economic facts of an entity and a specific time period, are represented by what we call XBRL reports (Instances in English) that are technically XML files, which will refer to the taxonomy, schema XML, on which they are based.
1.2 How XBRL works
XBRL is a computer programming add-on that tags each segment of automated business information with an identification code or marker. In most cases, the accounting software will insert the labels automatically. If your chart of accounts lacks the XBRL feature, you can add the tags using a free add-on program or customized software tagging tools.
Identification markers stay with the data when it is moved or changed. Thus, no matter how you (or, more accurately, your application software, such as a spreadsheet or word processor) format or change the information, the markers remain attached to it. Thus a number identified as a representation, for example, of profit in US dollars will always be recognized in this way. Typical labels include financial identifications such as assets, current assets, and receivables.
If the XBRL program does not contain ID markers that meet the needs of your business, you can easily create your own markers and add them because the program is fully customizable.
According to the language of accounting, a series of elements or labels, with autonomous and own rules to be fulfilled at the time of processing the data in the preparation of the report is called a data dictionary, which has been built and has already been it has what is technically called a Taxonomy, which must be universally accepted.
2. The digitization of information exchange processes: purchasing and billing processes.
Now, undoubtedly, the XBRL language allows streamlining the exchange of information and the subsequent analysis of the Financial Statements, helping Managers, Investors, Analysts, etc. that will not be seen in the tedious task of retyping, for its conversion from written document to binary information, the different accounting reports.
But what happens at the operational level? That is, what progress has occurred at the level of exchange of basic information, billing, purchases, issuance of payments, etc.
We believe that at the daily operating level, it is exactly the same as prior to the adoption of the XBRL language for Statements and Accounting Reports. In the entire administrative accounting process, there is a double entry, when not, a triple or quadruple entry of the same records.
In the purchasing and billing processes, the administrative officials of the supplier and buyer companies enter exactly the same thing, some checking out the items and recording the movements in the debtor accounts, others recording the entry of items and the movements in With creditor accounts, there is a phenomenal duplication of tasks that comes at a huge cost to the management of related companies.
Let's look at an example of a common purchasing process, namely:
a) Determination of the need for purchases: suppose that we are computerized and the system reports the item (s) that we must add to our stock. We then proceed to call, send an email or establish a communication mechanism with our suppliers to place the order through a Purchase Order.
b) Preparation of the invoice and / or remittance: under the assumption that the price and payment conditions have already been accepted, an official of the supplier must issue a receipt and / or invoice selecting the items ordered and assigning them a price according to a list Dadaist.
Later the accounting system will make the style entries.
c) Registration of goods entry and invoice received: exactly the same but in the opposite direction, instead of low stock, high and instead of debtor, creditor, some administrative official, you must re-type item by item, quantity by quantity of the purchase invoice received.
As we see above, as in the financial reports that had to be entered again, at the administrative level, on a day-to-day basis this situation is repeated thousands of times, item codes, quantities, prices, re-entered in sellers and buyers, high of stocks that exactly correlate with stock removals.
It would seem that the old accounting principle of "double entry" had been misinterpreted as "enter everything twice"
Then the question arises, could not be formulated a mechanism of entry - exit of goods that does not involve the re-digitization and duplication of tasks?
The big problem: the article codes.
All this immense duplication of tasks exists because the codes of the sellers and buyers are not compatible, that is, the same item, it is called in one way in the supplier and in a different way in the buyer, if not in different ways in the different branches of the buyer. It is as if a person, every time he travels, changes his name in each country he visits, therefore he must know where he is to know his name, it sounds ridiculous, but that is what we do at an administrative level.
Juan is Juan everywhere ……
If we kept the name of the suppliers, that is, their codes, if there were not the more than widespread custom of re-coding each item in each purchase, simply with a flat file with the purchase data, the systems of each company could be fed from the same source.
The re-entry of data exists because every time an item is purchased in a company the internal code of the same is placed and the movement to stock must be registered by said code. If the codes were the same or a translator existed, the systems could enter “digital invoices” simply by reading the files sent by the provider.
The operation, if the process were digitized, would be:
Although the operating scheme is similar to the normal one, the great difference is found in the non-existence of re-coding of the articles and therefore the entry into the systems is carried out directly by “capturing” the data in digital form.
Ultimately, if sellers agree on the types of fields, formats and transfer codes of the articles, the typing at the time of purchase is totally eliminated. The aforementioned constitutes a very important advance when it comes to making administrative processes more efficient, thousands of hours of typing are saved, reconciliations of accounts can be made online, etc.
The same advance that XBRL implies for consolidated information can be applied to basic accounting information.
Let's see how the process would work:
1) Sending price list in digital file or web pages:
Suppliers would send a digital price list and item codes format, or generate a web page for the purpose of placing purchase orders.
2) Digital billing submission:
Once the order is accepted and as soon as the merchandise is sent, a digital invoice is sent online with the same data as the common invoice that accompanies the merchandise.
3) Information capture:
Once the merchandise arrives, it is verified and the digital invoice is controlled with the invoice sent.
Once this process is carried out, the computer systems, which "understand" the invoice of the suppliers, Juan is Juan and not Diego or Alberto, incorporate the invoices into the stock system and generate the accounting entry.
As you can see, the only time the codes had to be entered and invoices entered was when building the digital purchase orders or the website.
Finally, from the point of view of the tax collection office, it would be very easy to control that the tax amounts invoiced by one party are exactly those deducted by the other.
There was a qualitative leap?: incorporate intelligence
Now, although it is true that the previous proposal implies an additional advance to the administrative processes, it is in fact the application of the same concepts of standardization of formats and digitization that is applied in XBRL.
It saves operating time, eliminates re-fingerings, but in reality we think that it does not mean a big change in the processes, they simply become more efficient.
Therefore, the question arises, when is the qualitative leap made?
We believe that the big change occurs when you take advantage of the great potential of processing and storage of current systems. At current speeds, it is perfectly feasible to think of computer systems performing autonomous activities, today reserved exclusively for the various officials of the organization.
That is, if the systems are programmed with behavioral rules and connected to the internet, given the capacity for dialogue that exists between them, it is perfectly feasible that the price request from a supplier, even the purchase up to the limits that are fixed can be done autonomously.
Once the article codes are standardized or translated, a system, upon reaching the replenishment point, can perfectly interconnect with the supplier companies through the network, “dialogue” between the systems and order orders.
In turn, once they have been formulated and delivered, the systems can integrate the billing information that is sent to them into their accounting systems.
Obviously, a merchandise reception control must be established, as we mentioned before, so that merchandise that was not properly controlled is not incorporated into the systems.
Although interconnection of systems, autonomous dialogue and programmed response seems too futuristic a concept, it is exactly what is done in medical science by curing a programmable pacemaker or a cardiofibrillator. In other words, thousands of people in the world live thanks to the fact that there are systems that register, connect and act absolutely alone.
Therefore, we believe that in accounting science as in medical science, the great leap can be made, whether in purchasing scheduling, cash transfers or the commercial area. Systems must census, compare with their behavior rules and act ……
Artificial Intelligence applied to accounting processes is definitive, as defined by Elaine Rich and Kevin Niigth (1):
“Artificial intelligence (AI) studies how to make machines perform tasks that, for the moment, are performed by Human Beings "
In short, we must be able to emulate human behavior, through rules designed a priori applicable to accounting systems, by way of example:
Rule:
In short, if I do not have stock, I am in a period of the year that I need to meet the minimum stock and I can pay for it, then: I place the order. These concepts are what are called Expert Systems:
Variants of this type of problems can be found by combining several logical rules or they are also solved with the so-called Case-Based Systems, which apply past experience to solve a current case.
Therefore we conclude that as in a triple jump athlete, in the 80's we supported the first foot with the advent of PCs, in the 90's through the exchange of accounting information (XBRL) we supported the second, but the We will take a great leap forward when we are able to delegate to intelligent systems, tasks that until today we only rely on our own administrative officials of different hierarchy…
