Unlike BP, which eventually came to admit the actual rate at which oil has been spurting into the Louisiana Gulf, the FASB and IASB continually resist any acknowledgement that the gulf between the two financial reporting regimes will remain vast, even under their best case convergence scenario. That's why the big financial reporting news of the week is the Boards' joint announcement that their latest best case scenario is once again being reluctantly revised downward:
"The IASB and the FASB today announced their intention to prioritise the major convergence projects to permit a sharper focus on issues and projects that they believe will bring about significant improvement and convergence between IFRSs and US GAAP."
Yes, it is true that a strict reading of the announcement only reveals that the Boards are going to be taking a little more time to finish their grand endeavor. But if an additional six months is all they will require to realize the miracle of Convergence, then why prioritize? Why not just stretch out the existing timetable?
Quite surprisingly, an answer to my questions may be found in another recent document from an unexpected source: the FASB no less. Here is what the FASB stated in its recent financial instruments and hedging exposure draft:
"The FASB's main objective is to develop accounting standards that represent an improvement to U.S. financial reporting. What may be considered an improvement in jurisdictions with less developed financial reporting systems applying International Financial Reporting Standards (IFRS) may not be considered an improvement in the United States." (page 2)
Is this the first crack in the "convergence is inevitable" façade? Maybe so, but to those of us who have been skeptical about convergence from the outset, the faults in the real convergence foundation made themselves apparent years earlier.
For example, in my previous post I mentioned in passing that convergence of R&D accounting had been attempted in earnest and then dropped like a hot potato. In the remainder of this post, I'm going to delve into that topic more deeply because the abandonment of the R&D convergence project would have been perceived to be a convergence death knell by all but the politicians and opportunists. Moreover, a headlong rush to closure on convergence will undo all the good that was accomplished by the FASB when it steadfastly resisted changes to its R&D standard.
Why SFAS No. 2 Is Right for the US
The differences between SFAS No. 2, Accounting for Research and Development Costs, (ASC 730-10) and the provisions of IAS 38 that deal with R&D arguably constitute the clearest case for why bright line rules may be essential in a particular accounting jurisdiction, even though fuzzy rules are tolerated in another. I'll deal with the fuzzy rules of IAS 38 in a moment, but I'll start with an explanation for why the FASB, in 1974, made a rigid and simple rule the foundation for its first substantive accounting standard.
To set the stage, around the time that SFAS 2 was promulgated there were (very generally) four kinds of R&D-intensive companies: (1 and 2) IBM and AT&T, by dominating computer hardware and telecommunications respectively, were in classes by themselves; (3) there were a few large pharmaceutical companies whose patented drugs made them hugely profitable; and (4) there were numerous smaller technology-driven companies scratching and clawing for capital and market acceptance. For the purpose of evaluating R&D accounting rules, the critical factor was this: the conventional wisdom held that only those smaller companies capable of reporting positive net income in an IPO prospectus would have any chance for a successful stock offering.
In regards to accounting standards, prior to SFAS 2, accounting for R&D was "principles-based." This meant that R&D costs were supposed to be deferred (and matched with future revenue from a project) if the expenditure created a probable future benefit. The costs of unsuccessful projects would have been immediately charged to expense.
That was the way it was supposed to work, but practice turned out to be the polar opposite of theory; virtually everyone did it wrong. Smaller companies seeking a successful IPO would capitalize all R&D expenditures, because that was the only way to report positive net income. For those other R&D intensive companies—like IBM, AT&T and big pharma—being too profitable was generally not in their best interests, in part because 'obscene' profitability would attract the attention of regulators. Thus, the big boys expensed all of their R&D, even though, because of their market power and concentration of intellectual capital, they had many successful projects.
So, the perverse situation that the FASB chose to confront was one where the R&D projects with the highest likelihood of success were being expensed, while the literally go-for-broke ventures would defer their R&D expenditures until their auditors could no longer keep themselves from snorting out loud.
The FASB tried as hard as they could to develop a set of subjective criteria that would permit deferral of some R&D costs, but it ultimately concluded that no set of qualitative or quantitative criteria would result in comparability across companies. Thus was born the rule that all R&D expenditures must go straight to the income statement.
IASC Serves Up R&D Chicken Salad
Four years later, the IASC (predecessor to the IASB) issued IAS 9 (the pertinent portions of which became IAS 38), to provide that research costs should be distinguished from development costs; and that development costs should be capitalized after the issuer met the following criteria:
- Technical feasibility;
- Intention and ability to complete an intangible asset for use or sale;
- Ultimate ability to use or sell the intangible asset;
- Ability to specify how the intangible asset will generate future economic benefits;
- The ability to reliably measure the development costs.
Thus was resurrected the fictions that: (1) management could reliably determine which costs should be deferred, and (2) that deferred development costs provide useful indicators of value. The motive, of course, was to give issuers greater flexibility to manage their earnings.
Don't Tell Anyone, but R&D Convergence Was a Complete Failure
Soon after the Boards inked their agreement to pursue convergence in 2002, they divided the work into two piles: high-priority projects for which they believed convergence could be achieved in relatively short order; and everything else. R&D was put in the first pile, apparently because the differences between US GAAP and IFRS were seen as particularly straightforward and tractable.
Right. After about four years of back-and-forth wrangling from polar opposite points of view, the following statement was buried deep in the bowels of the FASB's website (it's long gone of course, and for all I know the only version that exists is the one I salted away for a moment such as this):
"…The Boards noted that elimination of the differences between IFRSs and US GAAP could involve consideration of fundamental issues and that those issues were part of a longer-term research project on intangibles being led by the Australian Accounting Standards Board. Nonetheless, the Boards agreed that they should explore possibilities to eliminate some IFRSs/US GAAP differences in the short-term. They instructed the staff (a) to consider the criteria for capitalisation of costs under SFAS 86 … to see if they could be used to make the criteria for internally-generated intangible assets in IAS 38 more operational and (b) to consider whether there are any aspects of US GAAP that could be moved closer to IAS 38, e.g. the requirements related to initial recognition of intangible assets acquired in transactions other than a business combination."
In other words, the FASB ran up against a brick wall erected by chicken salad chefs from Australia and who knows where else. While the R&D accounting project may have been formally suspended, it was effectively sentenced to die from inattention—and to be buried in an unmarked grave.
Depending on one's point of view, one may regard SFAS 2 as either a rules-based standard or a principles-based standard (the principal being that deferred costs per se should not be put on the balance sheet as assets). But however one views the quality of SFAS 2, the FASB clearly saw, to its significant credit, that the IASB's hazy, lazy criteria for capitalization would be a step backwards from what the FASB struggled to achieve in its earliest days; and the FASB would only accept a converged solution that constituted an improvement.
It could have even been the case that the FASB actually accepted at face value the propaganda from the opportunists that convergence was supposed to enhance comparability, and so they resisted change when comparability would, in their judgment, not be enhanced. I readily grant that I may be asking for too much, but the IASB's obstinate adherence to an R&D treatment that defeats comparability should have raised huge red flags for the FASB. If R&D accounting, which was seen as both important and low hanging fruit, could not be converged, then there might be little reason to expect that US investors would be best served by efforts to converge US GAAP and IFRS.
But now that we are supposedly on the cusp of convergence, whither R&D accounting? This crazy race to deliver something, anything, to the SEC by 2011 would completely undo the FASB's principled stance in opposition to the IASB's approach to accounting for R&D. What was formerly a high priority topic has been studiously ignored for years; for acting in any other way is to openly recognize what a charade the convergence dance has become, especially as we must endure in its final throes.
I wish I could be as sanguine as Shankar Venkataraman of Georgia Tech, who made this comment after the FASB acknowledged in their financial instruments ED that what is good for the goose may not always be good for the gander:
"This should be music to your ears, Tom. Of course, you recognize that the FASB cannot be as blunt as you are. Nonetheless, this seems to be a HUGE departure from the idea in the not-so-distant past that convergence was inevitable, no?"
Perhaps that lone acknowledgement may have signaled some shift in outlook, but there is still a long way to go before I'll concede that the FASB and I might be finally singing the same tune on convergence. I will say, however, that I never felt I was merely a member of a vocal minority (so sayeth the convergence opportunists). To the contrary, I firmly believe that the vast majority of stakeholders in the U.S. capital markets feel as I do; I'm just a lot noisier than most.
I think another, more obvious red flag was the inability to converge FAS 109 and IAS 12. IAS 12 was drafted using FAS 109 as a starting point, tax was identified as a "short-term" convergence project in the original Norwalk Agreement, and yet 9 years later the two boards have drifted further apart since the issuance of FIN 48 and the IASB's contradictory amendmends to IAS 12.
What would have the potential to kill convergence altogether would be for a number of large multinational companies (along the lines of GE or JP Morgan) to publicly say that they would have no problems with issuing two sets of numbers (GAAP and IFRS) or reconciling their books to IFRS the way foreign private issuers previously had to reconcile to GAAP. It would serve several purposes. First, it would embolden the SEC to declare that they have done enough and that 2 GAAPs are better than the 25 or so that we had 10 years ago. Second, widescale reconciliations of GAAP to IFRS would help expose more of the idiosyncrasies and inconsistencies within both sets of standards, possibly driving more change for the better. Finally, it might be enough of a sop to the Big 4 and AICPA (who would at least be able to promote limited IFRS services to larger clients vs. the almost zero projects happening now) so that they pull back on support for the larger convergence effort.
Posted by: KPO'M | June 05, 2010 at 02:05 PM