Digital Financial Reporting

XBRL Cloud made a viewer available which allows anyone to have a look at SEC XBRL financial filings.  I spent some time experimenting with this viewer and point out 10 great features I found in the XBRL Cloud Viewer.  Experiment with these your self.  I am using the Apple 10-Q if you wanted to follow my examples below.

1. Pivot. While the SEC "interactive data" viewer is not really that interactive, the XBRL Cloud viewer is in fact interactive.  For example, go to the balance sheet (Condensed Consolidated Balance Sheets).  Grab the "Legal Entity [Axis]" and drag it down where it says "Concept" on the bottom of the balance sheet.  Also, notice the picture of the wrench and screwdriver on the tool bar, when you hover over it a yellow pop up shows "Configure".  Click on that.  This works similar to how a pivot table works in Excel.
2. Jumping around in document.  I used to call this "hypercube jumping", not really sure what to call it; but try this out.  Go to the balance sheet again as above.  Find the label "Property, plant, and equipment, net" and click on it. On the right side pane which pops up, click on the button "Linkage".  Notice that you see two networks listed.  One is the balance sheet, the other is condensed financial information.  Click on the condensed financial information line. Notice on the left how the viewer takes you to the details of property, plant, and equipment.  How cool is that!
3. Search.  This is a really nice feature!  Go to the Apple filing above, find the search text box in the upper right hand side area of the screen, type "respectively" (without the quotes), press the search button, and watch what happens.  The viewer finds the word respectively in the balance sheet labels and in a number of [Text Block]s (HTML fragments).
4. Fact tables view.  Now, a lot of people probably won't get exited about this; but it is a great modeling tool.  Click on "View" from the menu ribbon and select "Fact Table".  This shows you all the characteristics of all the facts of the selected component.  This is very helpful in creating appropriate modelling.  (It would be nice to be able to copy and paste the fact table information into Excel.)
5. Relations view.  As with the fact table above, you can view the relations.  Click on "View" and then "Taxonomy" from the menu ribbon.
6. Viewing detailed property information.  In both the fact table and relations view you can either click or double click on objects and then see the detailed information for that object in the right hand pane.
7. Viewing calculations.  Go to the balance sheet again, find total assets, click on the fact for one of the two periods shown, and then click on the "Calculations" button on the right.  Calculations are shown, the fact that they are consistent is shown, and they are not laid out upside down like many other software vendors lay them out!  Nice.
8. Sorting periods.  This is a little feature, but nice none-the-less.  If you go to the menu ribbon of the XBRL Cloud viewer, select "Options" and then "Period", notice that you can sort the periods in ascending or descending order.  Now, go to the balance sheet; change that option; and watch how the current period and prior period change order. SEC filers do this inconsistently and I end up looking at the wrong period periodically.  This helps that situation.
9. Can create direct links to the viewer.  Look at the URL of the viewer.  It may look a little funky, but if you read it closely, note the two parameters "title" and "instance".  You can change those and point to any SEC XBRL financial filing!  This is similar to what you can do with CoreFiling's Yeti.  Click hereto see what I am talking about (notice that you jumped to the concept "Assets, Total" in the US GAAP Taxonomy).  Think linked data.
10. It's free!  This is a free viewer.  You cannot beat that price.

That's all for now.

For the past several months I have invested significant effort in looking at the moving pieces of what it takes to appropriately verify that an SEC XBRL financial filing is a true and fair representation of the financial information of a reporting entity.  This document is an accumulation the majority of the information I gathered as part of that effort.

My next step is to distill this information in to a specific set of tasks/steps which can be used by an accountant creating an SEC XBRL financial filing which provide the verification which that accountant needs to be confident that they have achieved the appropriate result.  From there I will take a closer look at software applications which assist accountants in this verification process.

If you have any views, feedback, ideas, or other insight I would encourage you to send me an email and I will be sure to consider that information as I synthesize this information.

When you look at the information in the document above I believe that you will realize, as I have, that an SEC XBRL financial filing has a definitive, finite set of pieces or semantic objects: components, facts, characteristics, parenthetical explanations, relations, and the properties of those semantic objects.  These semantic objects are no different than the semantic objects of a financial report printed on a piece of paper.  The only difference between an XBRL financial report and a financial report expressed using a piece of paper is the medium of the expression; the information is no different.

While there is subjectivity and judgement related to what information needs to be expressed by a financial report and how to best express that information; the actual expression itself is very objective.  Besides, for something like XBRL to work effectively it seems that it must be objective otherwise the information exchange would never work.  Being explicit, being consistent, and being unambiguous are traits which computers need to work with financial information effectively.

Looking at SEC XBRL financial filings just got a lot easier!

XBRL Cloud has provided the ability to view any filing within a browser-based application.  All you do is click on an icon for that filing and you can view the filing. 

You can see not only the filing; you can see the taxonomy, fact table, details for all the information in the filing. 

This is definitely worth checking out. 

https://edgardashboard.xbrlcloud.com/edgar-dashboard/dashboard.do

More information later after I have gotten a chance to fiddle with it.

Clearly software plays a role in the verification of an SEC XBRL financial filing.  In fact, the more computer software does for you, the less manual work you need to perform (which reduces cost) and the easier it is to achieve the quality level most accountants desire.

Continuing my assessment of how to verify an SEC XBRL financial filing; here is a preliminary list of the functionality verification software needs to provide accountants and others in support of the process of verifying an SEC XBRL financial filing:

• Quality XBRL validation (XBRL 2.1, XBRL Dimensions).  What you are creating, after all, eventually ends up as XBRL and you need to be sure your XBRL technical syntax is correct.
• Provides SEC Edgar Filer Manual (EFM) rules validation.  The SEC EFM places specific additional technical syntax restrictions on how you are allowed to use XBRL.  You will need to validate your filing against these rules.  Other semantic rules are imposed by the EFM.  Many of those can be automatically verified. (Keep in mind that other rules cannot be automatically verified and must be manually verified.)
• Logical structure verification (such as US GAAP Taxonomy Architecture rules).  The US GAAP Taxonomy specifies how to construct your model. It uses [Table]s, [Axis], [Member]s, [Line Items], networks, concepts, and abstract concepts in specific and logical ways.  Creating illogical structures is, well...not logical.  Software can easily verify that you are creating logical, and not illogical, models.
• Semantics engine for processing business rules (preferably but not necessarily XBRL Formula). US GAAP has specific semantics.  Assets = Liabilities + Equity is one example.  The SEC Regulation S-X imposes semantics.  Industry financial reporting practices have specific semantics. And finally, each reporting entity has its semantics.  How are you going to verify that all these semantics are correct if your verification software does not have a semantics engine?
• Core US GAAP financial integrity rules validation. The proof section of the Financial Report Semantics and Dynamics Theory shows that there is a core set of US GAAP financial report semantics which every SEC filer's filing has.  These 8 core rules are specified in XBRL Formula, so if you have a semantics engine which supports XBRL Formula, your software can provide this verification.
• Ability for business users to create business rules.  Clearly reporting entities have their own unique semantics and therefore will need to be able to create business rules.  Even very basic SEC XBRL financial filings have roll forwards and member aggregation rules which XBRL calculations simply cannot be used to verify.  As such, business users will need to create these rules.  Emphasis here on business user.  Creating these rules does not have to be complicated.
• Views of semantic objects, relations, properties understandable by business users. Many verification steps are manual, things humans will need to do.  To do this, they need to see the SEC XBRL financial report in ways which are understandable and helpful to them.
• Appropriate navigation between semantic objects, relevant properties, and relevant relations. These objects are both related to one another and have detailed properties which business users need to have a look at to be sure they are set appropriately. To do that, users need to be able to navigate from say facts to the characteristics of those facts and then to the properties of those characteristics in rational, helpful ways.
• Leverage information models to create proper semantic renderings.  For renderings of the information to be understandable, they must be shown correctly in the rendering.  There are exactly two reasons for bad renderings: (1) bad models, (2) bad rendering engines.  One thing most rendering engines do not do well at this time is leverage the information models of the information which was, well, modeled. Roll ups, roll fowards, adjustments, variances, hierarchies, and other information models have characteristics which software vendors could leverage in their rendering software to improve what business users see and have to therefore work with. If you know your model is good (see bullet point three above), question the capabilities of the rendering engine of the software application you are using.
• SEC interactive data rendering.  While it is much more likely that an SEC XBRL financial filing will be viewed within software other than the SEC interactive data viewer; users still want to understand how their filing will look on the SEC web site.
• Support for tracking of manual validation tasks and steps. Not all tasks or steps can be automated.  So, verification of an SEC XBRL financial filing will always be a combination of automated and manual tasks/steps.  Software needs to help you manage these manual verification tasks/steps.
• Comprehensive and useful set of verification reports and appropriate verification evidence package. There are many different reports which can be provided which can make you both more efficient and more effective in your verification tasks/steps.  Software should support this. Further, particularly CPAs providing agreed upon procedures, work where an internal auditor or your process requires quality documentation (i.e. think SOX); you will need an evidence package.
• Collaborative, multi-user.  Although not required, creation of an SEC XBRL financial filing tends to be collaboration and thus software which offers this feature can be helpful.
• Transparency into what the software is automatically verifying so manual verification work can be properly planned.  Verification of an SEC XBRL financial report should not be a "black box". At the very least, you need to understand exactly what verification steps software performs so that you can properly plan your the manual tasks/steps required to supplement your automated verification. As such, you need to be able to see the specific automated rules software is performing.  Today this is particularly important as software may perform different sets of validation rules.
• Comparison between multiple versions of a report in order to understand differences, for example to manage last minute changes. A very helpful feature is the ability to compare different versions of an SEC XBRL financial filing to understand changes between the two versions.  If you did extensive work in verifying your SEC XBRL financial filing and then there are a few last minute changes which need to be made; how can you be sure some other change was not unintentionally or maliciously introduced?  Software which compares multiple versions can help you maintain control of your filing.

Finding verification software with all these features can be challenging today.  But, if you start asking for these features because you believe you need them; it helps motivate software vendors to provide these features.  What I personally do is use multiple software applications to achieve my verification needs.  This may not always be possible or practical; but it is better than trying to "roll your own" in Excel.

The CEO and CFO of a reporting entity needs to be able to look at an SEC XBRL financial report for them to be able to sign off on that report. Clearly they look at the "paper" or HTML versions of these financial reports which get filed with the SEC.  There are even signatures on those reports that attest to this fact. Or saying this another way, why would a CEO or CFO not need to be able to review these SEC XBRL financial reports if they wanted to?

So, what; we just give the CEO and CFO the same software that accountants who create these reports use to review and verify what they have created. That should work.  Or maybe we give them the software used by the third party accountants who do verification work to be sure the report is created correctly.

Why not?  Well, that verification software used by accountants and third party accountants is a bit too hard to use perhaps.  So, make it easier to use.  And if you can make it easier to use so that the CEO and CFO can use it, why can't it be made easier for the accountants and third party accountants?

Let's throw in the investors and analysts who use this information; why would a lot of the information that they see be different than the information used by the CEO, CFO, accountant or third party accountant reviewing the financial report.  What about the lawyers involved in the process of getting that SEC financial filing out.

What, specifically, is so hard to use about verification software or other software used to look at the information contained in these SEC XBRL financial reports?  Take the SEC interactive data viewer.  What is so hard about using that?  Actually, the SEC interactive data viewer, while it leaves a lot to be desired, provides some very good clues as to what the CEO, CFO, accountants, third party accountants, investors, analysts, lawyers, and others might need.  Here is a link to an Apple 10-Q as viewed within the SEC XBRL interactive data viewer which you can use to help you think about this. 

It seems to me that this is some of the stuff that any one of those business users would need to see would include: all components, all facts within a component, all characteristics of a fact, all the properties of each characteristic, all the relations between characteristics and facts (such as computations of roll ups, roll forwards, etc).  These are all the semantic objects, properties of those objects, and relations between objects within the financial report. 

Let me walk you through what I am talking about.

First off, business users need to see all relevant information about reported facts within a financial report. Clearly all the components of the report are relevant, you can see all those on the left hand side within the Apple 10-Q as viewed by that SEC interactive data viewer. 

Select one of the components, let's try the "Condensed Consolidated Statement of Operations".

If you click on that component, you see Apple's statement of operations.  You can even "click" on a line items of that component and see information about characteristic of the fact you are looking at.  Can you see all the information about that line item concept characteristic?  No; you cannot see the references to the accounting literature for that concept, clearly that might be nice.

But what about the other information about a reported fact.  Click on the fact.  Well, the SEC previewer does not let you do that (rather you can click all you want, the viewer just won't do anything).  But what if it did?  What if you could click on any fact and see all of the relevant characteristics of that fact?  That would be nice.  Remember that component you clicked on to get to this page?  It would be nice to see the properties of that component also, all of them.

What about the characteristics which exist but you cannot see in the SEC viewer because the viewer simply does not show them?  If you look in the columns, you see the SEC viewer shows two column: for the current quarter and the same quarter for the prior year.  You cannot click on the period and get any information.  But what if you could?

There are other characteristics which are not shown.  You don't see that both the current period and prior period columns both relate to the legal entity "consolidated group".  You can see the CIK number which identifies the reporting entity.  Both of those are characteristics of each fact in the condensed consolidated statement of operations.  But what if you could click on any fact and get all the information about every characteristics of that or click on every characteristic and get the properties of that characteristic.

Basically, as I see it what makes software hard to use these days for CEOs, CFOs, accountants, third party accountants, analysts, investors, lawyers, and other business users is poor software design.  Today's software does not show you all the information you do need to see, none if the information that is irrelevant and therefore you don't need to see it, organizes the information you want and need to see logically.

There are many reasons software vendors have done this.  The biggest reason is that the US GAAP Taxonomy does not make its semantics clear.  Therefore, software vendors had to fall back on the XBRL technical syntax because they understand that. This situation needs to be corrected.

Full-featured verification software used when creating SEC XBRL financial reports will have far more features than are needed by a CEO, CFO or other business user who needs to look at one of these digital financial reports.  However, those additional features are efficiency tools.  The ability to look at every semantic object, every property of each of those semantic object, all the relations and the relations properties is core and fundamental to any business user looking at any SEC XBRL financial filing for any reason what-so-ever.