Peeling away financial reporting issues one layer at a time

The Wild West of “Nonauthoritative” GAAP

Two posts ago, I provided two reasons for the PCAOB* to take the opportunity of its auditor’s report project to look closely at the meaning of, “fairly presented … in accordance with Generally Accepted Accounting Principles”:

  1. Truth in labeling — As I described in that earlier post, few readers of financial statements fully understand the technical meaning, in part because it no longer bears any relationship to common English usage.
  2. The Wild West of nonauthoritative GAAP — The vaguely specified methods set forth by the FASB for utilizing “non-authoritative” GAAP should be of great concern to the PCAOB.

In this post, I’m going to provide my views on #2, above.  To begin, I’ll first provide some background on the role of nonauthoritative GAAP in financial reporting under US GAAP.  Then, I’ll describe two cases in which the application of nonauthoritative GAAP should cause concern.  My presumption is that if I can stumble on these two cases, then there must be lots of others lurking in audited financial statements.


The latest pronouncement on the matter of nonauthoritative GAAP is Statement of Financial Accounting Standards No. 168 (2009).  As the last SFAS issued by the FASB, it pronounced that the Accounting Standards Codification would become the sole “authoritative” source of U.S. GAAP.  As to the  enduring question of what to do when authoritative sources do not directly (or by suitable analogy) address a particular transaction, ASC 105-10-05-3 enumerates the following potentially permissible nonauthoritative sources that could be referred to as GAAP:

  • Practices that are widely recognized and prevalent either generally or in the industry;
  • FASB Concepts Statements;
  • AICPA Issues Papers;
  • IFRSs;
  • Pronouncements of professional associations or regulatory agencies;
  • Technical Information Service Inquiries and Replies included in AICPA Technical Practice Aids;
  • Accounting textbooks, handbooks, and articles.

ASC Topic 105 makes clear that an issuer has broad discretion to choose amongst analogy to authoritative GAAP or from the above smorgasbord of nonauthoritative sources.  Although auditors would have some say on the matter, the PCAOB’s guidance to them is clear as mud.  Per AU 411 as amended by the PCAOB:

“The auditor’s opinion that financial statements present fairly … in conformity with [GAAP] should be based on his or her judgment as to whether (a) the accounting principles [whatever that means] selected and applied have general acceptance [whatever that means]; (b) the accounting principles are [merely] appropriate in the circumstances; (c) the financial statements, including the related notes, are [merely] informative … and (e) the financial statements reflect the underlying transactions and events … within a range of acceptable limits, that is, limits that are reasonable and practicable….”

I encourage you to think about how the auditors have applied AU 411 when reading the following two cases.

Case #1: Revenue Recognition by For-Profit Universities

About five years ago, I had occasion to review the revenue recognition policies of a for-profit university (FPU).  Below is what I found for what evidently constitutes “practices that are … prevalent … in the industry.”

Students are billed on a course-by-course basis.  They are billed on the first day of attendance, and a journal entry is made to debit A/R and credit deferred revenue for the amount of the billing.  The A/R is ultimately adjusted by an allowance for uncollectible accounts of around 30%, and the deferred revenue is recognized pro rata over the duration of the course.

Does prevalent industry practice justify recognition of executory contracts?  FPUs, like other enterprises engage in all manner of executory contracts; but, with the exception of some lease contracts, none are given accounting recognition before any performance has occurred.  Evidently, if FPUs act together to violate a broad-based convention, that’s okay, because it makes ‘industry practice.’    

But, recognition of executory contracts is not even the most glaring inconsistency with authoritative GAAP by FPUs.  Long-established and generally accepted (literally, for once) principles of revenue recognition run counter to what FPUs are doing.  Specifically, SFAC No. 5 enshrines the long-held view that revenue should not be recognized until it is earned and realizable (or realized), and practically all of the specifics in the ASC, not to mention SEC interpretive guidance, is consistent with this aspect of SFAC No. 5.

Is anything “earned” by an FPU until grades are recorded in transcripts?  I think it’s safe to stipulate two things: (1) from a student’s perspective, nothing is owed if the FPU doesn’t fulfill the obligation of recording a grade in a transcript that a student can use to document performance; yet (2)  there is no  “authoritative” GAAP that specifically speaks to this question.  Ergo:

“If the guidance for a transaction or event is not specified within a source of authoritative GAAP for that entity, an entity shall first consider accounting principles for similar transactions … and then consider nonauthoritative guidance.”  [ASC 105-10-05-2, emphasis supplied]

Call me cynical, but that language is squishy enough for an FPU to do pretty much anything it wants regarding the timing of revenue recognition.  So it does.

But, determining the revenue recognition trigger is still a mere quibble compared to the “realizable” question.  A typical company estimates its allowance for doubtful accounts to be around 2% of gross accounts receivable; but an FPU’s allowance could be 30%. If, after an allowance is initially accrued, it is discovered that the actual uncollectible amount was one-third higher than the accrual, the premature recognition of earnings by the typical company may or may not be material; but for an FPU, it will be a cataclysmic income statement event.

Industry practice is industry practice, but that doesn’t make it good practice to come to a conclusion that collectibility is reasonably assured when a small change in the estimate of uncollectible billings can have such a material effect on the financial statements.  Yet, over a sizable number of FPUs and their auditors, they see nothing in the ASC to stop them.

Case #2: Accounting for Tax-Advantaged Projects 

Investments for which tax credits are available (e.g., solar, wind, affordable housing) are often set up as IRS-approved “partnership flip structures.” Without getting into a lot of detail, “tax equity” investors receive nearly all of the cash distributions (mainly from tax benefits) in the earlier years of the project.  A “sponsor” is responsible for operating the project and receives the bulk of the cash distributions occurring near the end of the project’s planned life.

Tax equity partners must account for their investment using the equity method. The sponsor must consolidate the partnership — since it meets the definition of a “variable interest entity” and the sponsor is deemed to be the primary beneficiary. The sponsor measures the non-controlling interests (i.e., the tax equity) using the mirror image of the equity method.

Nonauthoritative GAAP comes into play because authoritative GAAP presumes that an “ownership percentage” can be calculated by investors for measuring their share of the partnership’s earnings under the equity method.  This is not possible for investors due to the complexities of the contractual provisions for allocating taxable income/loss and cash flows.

In response to the problem, the AICPA’s Accounting Standards Executive Committee issued an exposure draft in 2000 that proposed  the “hypothetical liquidation based on book value” (HLBV) method for determining a varying ownership percentage each period. Again, I don’t think it is necessary to get into the details, but basically what happens under HLBV is that: the business is unrealistically assumed to have been liquidated at each balance sheet date (probability is essentially zero of that actually happening in most years) at its book value (usually a huge understatement of economic value); and the hypothetical net proceeds from liquidation are distributed to investors according to the cash flow “waterfall” set forth in the partnership agreement.  Each equity investor’s “percentage of ownership” for the period is calculated as its relative share of the waterfall.

We could debate the merits of HLBV, and perhaps I could even explain it a little better (for what it’s worth I think it’s a whacky approach to a problem that boils down to changes in the present value of expected contractual cash flows); but for the purposes of this post it is sufficient to know that the ED was never finalized.  In other words, HLBV was explicitly rejected by an authoritative accounting standard setter, but is still permissible to be used as nonauthoritative GAAP because … lots of issuers do it.

* * * * * * *

As the forgoing examples illustrate, nonauthoritative GAAP can be a crapshoot.  Surely there is a better way to determine an appropriate accounting treatment than from an unordered list of ingredients for coleslaw?

Of course, there is.  Standard setters have always had the ability to produce a set of recognition and measurement principles that no issuer should be permitted to violate when filling a gap in GAAP.  I’m pretty sure that most of my readers have a pretty good idea why they haven’t done that.

At minimum, and the reason I am writing this post, is that the application of nonauthoritative GAAP should trigger special disclosures:

  • The notes to the financial statements and MD&A should prominently identify the sources of nonauthoritative GAAP, and explain why it was necessary to find and apply a nonauthoritative source.
  • The notes and MD&A should also identify and quantify the revenues, expenses, assets and liabilities that were subject to nonauthoritative GAAP treatments.
  • The audit report should also include a paragraph that alerts the reader to the use of nonauthoritative GAAP and the note wherein the details are disclosed.
  • The auditor should specifically state that the use of the specific nonauthoritative GAAP  is a reasonable choice under the circumstances, and preferable to the application of authoritative GAAP by analogy.

If the dubious concept of nonauthoritative GAAP must be allowed to persist, then, as in other areas where issuers have discretion, financial statement readers are entitled to know the gory details.


*I am a member of the Standing Advisory Group of the PCAOB.  The views expressed herein are my own and do not necessarily reflect the views of the Board or other Board members or staff.


  1. Reply Russ May 23, 2014

    Tom, you seem to believe that accounting is a black and white bean counting exercise. Quite to the contrary, it requires the exercise of significant judgment and, due to costs versus benefits, can in many cases be at best an effort to get close. There are certainly no shortage of problems with GAAP. There will always be problems with any approach to something this complex, especially when you factor people into the equation. People with varying knowledge of the subject, varying motivations and needs, and varying levels of integrity.
    There is not and never will be a perfect system. Different objectives for accounting information suggest different principles and methods.
    In your example, the FPUs “earn” the fee ratably over the time period of the course as courses are held, lectures are given, work and exams are graded. The fee for a course is not earned by entering a grade in a book, students are not buying a grade, they are buying a course. They may never complete the course, but will still owe the fee because they signed up and did not withdraw timely. Therefore, recognizing the revenue ratably over the term of the course is fair and reasonable. The revenue is recognized as the corresponding costs are incurred. There usually is a point of no return, where a student may want to withdraw but can’t. But waiting until that point to recognize revenue earned up to that point would not provide a fair representation either, since most students will complete the course and owe the entire fee. Therefore, judgment has to be exercised and estimates made. Yes, some of that revenue will not ultimately be collected or even due because of withdrawals, but that is a significantly smaller error than the error generated by recognizing no revenue until a grade is entered in a book, the error of recognizing losses from costs incurred each day, month, etc. up to the completion of the semester, with no recognition of revenue.

    • Reply Tom Selling May 23, 2014

      Hi, Russ:

      Your interpretation (which doesn’t deal with all of the problem areas that I raised) runs counter to analogous interpretations by the SEC staff. SAB 104 generally provides that all elements of an arrangement must be delivered in order for revenue to be recognized (unless certain elements are perfunctory, or an element has standalone use). A “course” consists of knowledge and a grade. Also, as I pointed out, from a student’s perspective, if the professor did not calculate and record a final grade (clearly, a non-perfunctory service) upon course completion, she would be entitled to her money back.

      Of course, I don’t believe that accounting is a black and white bean counting exercise, and I don’t low where you get that idea from. I hope that we at least agree that accounting a lot more complicated than it should be; and that there are lots of opportunities to make it simpler and better.

  2. Reply Russ May 27, 2014

    Tom: You can’t deal with all the problems you raise. There is no perfect. There may be room for improvement, and that opportunity is likely rising with many of the changes occurring today. Different perspectives and different objectives require different approaches and methods. So for any approach or method developed to be as broadly applicable as possible, within a consistent framework, and with appropriate cost/benefit considerations, anyone can easily poke holes and identify problems. While there is gain to be had from continuous improvement, there is frequently loss to be had too.
    You seem to identify problems, but don’t offer any specific solutions. I expect this is because you understand this. As soon as you offer a specific solution, there will be many problems identified therein from varying perspectives and objectives, because there is no perfect.
    To continue with the class objective. A “course” does not necessarily consist of a grade. That would be unfair to the college. A student signs up, goes to a handful of classes and then drops out. The college is entitled to compensation for services rendered. A professor delivered lectures, assigned work, graded some work. A facility was provided and utilities were consumed to deliver these services. Administrators administrated. To argue that the college is not entitled to proportional compensation is unfair. The costs versus benefits of calculating these allocations for every individual student are prohibitive. Thus there are compromises and shortcuts, arbitrary marking posts. Likewise, waiting to recognize the revenue that covers these incurred costs until the end of the course is misleading, unless the agreement is that there is no fee for the course unless a grade is awarded. But that is not the arrangement, and rightfully so. The student must perform also to “earn” the grade and it would be an unfair arrangement to make the college bear the risk of student nonperformance, after the college has performed and stands ready to complete their part of the bargain. The assumption that the revenue is only earned upon entering a grade in the book is flawed.
    We do agree that there are many opportunities for simplification, but they generally will involve compromise from many perspectives. Compromises that will leave an optimal solution open to complaints.

Leave a Comment