skip to Main Content

Ungarbling Financial Accounting

Alas, I have not posted very much of late.  But, it’s only because I am trying to find as much time as possible to write my book describing a simple, comprehensive and robust basis of accounting that any public company could use; and that would produce more relevant information for investors than U.S. GAAP or IFRS.  Progress has been slower than I had hoped, but to-date I have drafted the chapter explaining the objectives of my “Shareholder-Oriented Financial Accounting” (S-OFA), and am now drafting what I call the foundational constraints of S-OFA.

I decided to take time away from book writing for this post for two reasons.  First, I thought that it could help me to organize my thoughts about one particular foundational constraint that I have not heretofore written about:  the “garblings” (i.e., distortions and confusions) of numbers and terms that are ubiquitous in U.S. GAAP and IFRS.

Second, I continue to struggle to find a title for the book.  The latest working title is Shareholder-Oriented Financial Accounting: Liberation from the Politics of Capital Markets Regulation.  It doesn’t feel to me that the part after the colon hits the sweet spot. All suggestions – serious, funny, or otherwise — are welcome.  Respond either by commenting on this post, or email me.

Ungarbling the Garbling in GAAP

Garbling occurs to some degree in practically all media, but in financial accounting it’s both intentional and very intense.  Let me explain.

The U.S. GAAP/IFRS approach to measuring assets (and liabilities) is frequently and far too simplistically described as a “mixed attribute” system.  Historic cost may be mistakenly considered to be an attribute, but that is a digression. (Historic cost is merely an attribute of a past transaction, as opposed to the asset itself.)  A legitimate reportable “attribute” is generally understood to be either: a measure of value if sold (e.g., “fair value”); cost to replace; or present value of future benefits.

But, IMHO, too few understand that some (the majority of?; an increasing number of?) assets are “measured” with algorithms, which cannot possibly produce a “faithful representation” of any of the aforementioned financial attributes.  That is because there is only one way to describe the numbers generated with these algorithms:  by nothing more than the algorithms themselves.  I will refer to the numbers produced by these algorithms as “garblings.”

One of my favorite examples of a garbling — one that I have used in many posts (see here for the latest, and here for the collection — is the “translation” of land that is held by a foreign subsidiary.  (I’ll explain why I put “translation” in quotes later.) According to U.S. GAAP (and IFRS), the algorithm that produces the number reported on the consolidated balance sheet would be historic cost in the foreign currency multiplied by the current exchange rate.  SFAS 52 even states that this process of “translating” the land into the reporting currency is nothing more than a mechanical process (therefore, leading to a translation “adjustment” as opposed to a gain or loss).

Other examples abound, with leases and loans being two recent examples.  Together, they reveal that GAAP is not even a “mixed attribute” information system that, at its very best, adds together the historic cost of some assets (non-attributes) with the fair value of an interest rate swap; and presents the apples/oranges summation as “total assets.”  It is even less than that.  U.S. GAAP and IFRS are largely collections of garblings.

It seems to me that this is unique among what I might call language-based reporting systems, e.g., newspapers, radiological reports.  Statements from these systems could require specialized knowledge to be accurately understood, but it is clear that within these systems intentional garbling is seen as an act of bad faith.  The defenders of garbling in financial accounting argue that garbled numbers are informative nevertheless, because they approximate, or are correlated, with attributes.  Moreover, garbling could be excused because the cost of estimating an attribute is too high, and/or its reliability is too low. To these arguments, I have the following responses:

  • Even under the best of circumstances, there is no way to evaluate the deviation or degree of correlation between a garbling and an attribute. U.S. GAAP could even require a garbled cash balance!
  • Algorithmic garblings have ZERO representational faithfulness, which is purportedly one of the two fundamental qualitative characteristics of financial accounting (the other being relevance).
  • It is costly for users to learn all of the garbling algorithms, and few (zero?) users of financial statements actually do.  On the other hand,  reading a listing of assets/liabilities reported using one consistent attribute is not much more challenging than reading the newspaper.

Translating land held by a foreign subsidiary at current exchange rates actually provides two instances of garbling.  Thus far, we have considered the garbling of a number, but U.S. GAAP garbles words as well, and “translation” is a prime example.  That term would ordinarily imply that there is an attempt to retain meaning (e.g., translating from Spanish to English); but, as I already mentioned, the FASB admits that by reporting an amount for land by multiplying an historic cost in a foreign currency by a current exchange rate is nothing more than a mechanical process.

In addition to translation, here is a sampling of term garblings that I have written about in postings over the years, and that S-OFA will not permit:


Garbling in GAAP is like dirt.  It’s everywhere.

I will be trying my best to make S-OFA as clean as a whistle — both its words and its numbers.

Back To Top