“Reducing Complexity” — The Latest Red Herring from the FASB

I don’t have exact figures, but I’m fairly certain that the AICPA’s annual Washington, DC conference on SEC-related developments may be the place where you can find the highest number of accountants crammed into a single hotel ballroom.  And that’s not counting those getting their CPE hours from watching the web feed.

If I had been one of the attendees who paid around $1,500 to attend a financial reporting fest with “developments” in the title, I would have expected that the FASB chair Russell Golden to use his time slot to talk about actual developments.  For example: 

• Why has convergence with IFRS been such a big flop?
• What has the dithering on financial instruments, leasing and revenue recognition actually accomplished?

But instead — and understandably, given all the downers of the year past — Mr. Golden regaled the crowd with his thoughts about “the future of the FASB and the issues that we’ll be addressing in the months and years ahead.”  I take umbrage with many of his remarks (e.g., his mischaracterization of the FASB’s mission, and the lame excuses given for tabling the conceptual framework project, which was begun almost ten years ago), but in the interests of brevity I’ll focus on the topic of reducing “complexity” of accounting standards.

Complexity is Simple

Mr. Golden would have us believe that “complexity” in accounting is a difficult concept, and that two ways (if he has more, he isn’t saying for now) of thinking about it are: (1) a standards could too dense and complicated; or (2) that the costs of producing information numbers could exceed its informational value.

I don’t buy it.  Accountants may not be rocket scientists, but accounting is far from rocket science.  If accounting standards are unclear, it’s probably because they were meant to be just that.  Over time, the FASB generously spiced its pronouncements with slippery notions of management intent, pretentious terminology, and gaps in procedures. Which of these ambiguities would Mr. Golden want to resolve, and how would he go about it? Your guess is as good as mine.

As for the costs of producing numbers from complex rules, Mr. Golden does make a valid, albeit relatively insignificant, point.  The un-acknowledgeable elephant in the room are the costs borne by investors from value-destroying decisions made by management to achieve an artificial accounting result. This is the real cost of complexity, it has been the theme that has kept my blog going over the years, and surely Mr. Golden won’t go anywhere near it.

The examples of accounting topic that Mr. Golden gave in his own speech will serve to make my point:

Accounting for liabilities and equity — In 2007, the FASB was actually on the right track when it expressed a preference for a simple and elegant solution: that only basic ownership interests should be classified as equity, and everything else on the right-hand side of the balance should would be a liability.  But, the IASB wouldn’t go along.  So here we are seven years later, without even so much as an active project on a severely broken topic, because the IASB is unwilling to curb management’s predilection for complex financial instruments, whose sole use is to smooth and/or hide earnings.

Perhaps though, there is a ray of hope in this statement from Mr. Golden:

“But let me be clear, any decision we make [to reduce complexity] would be weighted toward what investors need. If a change would impair an investor’s ability to do his or her work, then we would not consider this to be an appropriate agenda item.”

Here we have a topic for which Mr. Golden has acknowledged that simpler would be better.  We also know that the FASB has previously supported a solution that would actually help “… an investor’s ability to do his or her work,” that the IASB does not support.  Will Mr. Golden side with investors and reduced complexity, or will he remain silent, in deference to the convergence robocops?

You probably know what I think.

Hedge accounting — The FASB has shown little concern for the fact that transactions entered into for the purpose of qualifying for “special” hedge accounting (i.e., income smoothing) often cost more than non-qualifying hedging strategies that could actually be more effective in reducing the kinds of risks that shareholders actually care about.  The biggest source of complexity in hedge accounting is the highly flawed quantitative rules for determining “hedge effectiveness,” but less rigorous rules might only create more negative economic consequences.

The far better, and far less complex, solution would be to do away with hedge accounting altogether, and to require fair value measurements for financial instruments.  But, because of anti-fair value FASB members like Mr. Golden, that’s just one more opportunity for complexity reduction that will go by the wayside.

Goodwill impairment — I don’t favor the IASB’s approach, but the FASB’s solution to goodwill impairment (and long-lived assets in general) is far more complex. Yet, despite all of the convergence talk, the FASB wouldn’t dream of considering the IASB approach because managers in the U.S. want to keep their ‘big bath‘ accounting for impairments.  So long as impairments of long-lived assets are recognized long after the actual value destruction has occurred, management will try to soft pedal the conditions that led to problems until financial statement recognition is nigh upon them.

Revenue recognition — Mr. Golden proudly trumpets the forthcoming revenue recognition standard (that nobody wants) for its core principle that will “apply to everyone” — except that it won’t.

Under that core principle, revenue is recognized from an arrangement after a performance obligation is satisfied.  But, despite its purported universality, the boards have found it necessary to create numerous scope exceptions, most notably to require separate, highly complex standards for lessors and insurers.

Moreover, the core principle of the revenue standards does little or nothing to keep things simple, even for arrangements within its scope.  For example, it doesn’t clearly specify what a “performance obligation” actually is; or more important, how to measure it.  Hence, the 14 years of deliberations and at least two more years of implementation guidance to come.

More important even than the lack of clarity and limited scope of the revenue recognition standard are the exceptions within the standard itself.  Most recently, for example, the boards decided, in contravention of its putative core principle, that not all satisfied performance obligations would trigger revenue recognition.  You would, in addition, have to be able to conclude that, taking credit risk into account only, collection of what you are owed is “probable” (whatever that means — and could mean something different under IFRS than US GAAP) at the inception of the contract.  If not, you would have to wait until all cash is received and it is nonrefundable; or, if the contract is cancelled, the amounts received are non-refundable.

If anyone thinks that this looks like a step toward less complex accounting, kindly chime in. I think it’s just another dog’s breakfast.

* * * * * * *

$1,500 should be enough to get some minimal accountability from the leader of the FASB.  Without accountability, it’s hard to be hopeful that rosy and enthusiastic promises for a better future will materialize in any significant way.

If Mr. Golden were genuinely interested in making a serious dent in the complexity of accounting, then he should go about it a lot differently.  First, he should evaluate whether a complex rule generates negative economic consequences to present shareholders.  In too many cases, this is self-evident: if an accounting rule looks fishy, then management is probably abusing it.  Second, a compelling case for complexity should be made for a rule to survive in its present form.   

As a first cut, here’s what any forthright assessment of these two steps will produce:

• We don’t need any rule that results in the recognition of “other comprehensive income”;
• We don’t need anything in shareholders’ equity other than common stock and retained earnings;
• We don’t need stock compensation rules that understate the ultimate cost of share awards;
• We don’t need complex and arcane rules for recognizing asset impairments;
• We don’t need the concept of constructive obligations;
• We don’t need deferred taxes;
• We don’t need hedge accounting;
• We don’t need operating leases, or two methods for measuring and reporting lease costs;
• We don’t need the concept of a functional currency;
• We don’t need interest cost capitalization;

I know that there are some people who will disagree with some of the items on my list.  But, that’s not as important as agreeing that accounting is far too complex.

And, that Mr. Golden’s proposals won’t affect the real problems of complexity in any significant way.