Government contract cost accounting practitioners recognize that not all Cost Accounting Standards are created equal. Some Standards, such as CAS 406 or 407, are generally considered to be more straightforward and comprehensible than others; whereas some of those other Standards (e.g., CAS 415) are generally considered to be tough nuts to crack.

In other words, some Standards are more difficult than others to comply with. The two pension Standards—CAS 412 and 413—definitely fall into the “tough nuts to crack” category. They are considered by most practitioners to be perhaps the most difficult and complex Cost Accounting Standards to understand, let alone comply with. They are more difficult for a variety of reasons, not the least of which is that you need a really good actuary to help make the necessary adjustments between GAAP accounting and government contract cost accounting.

They are so complex that, when revised in 1995, the revisions initiated a “storm of litigation” that continues to this day. And the litigation involves big dollars, too. In many cases, there may be literally tens of millions of dollars at stake.

Take, for example, General Dynamics’ appeal of a Contracting Officer’s Final Decision asserting that GD was in noncompliance with the requirements of CAS 412 for its Fiscal Years 2005 through 2009. It was estimated that approximately $53 million (plus interest) rode on the outcome.

The issue was GD’s use of actual market returns on its pension plan assets instead of actuarial “expected” returns, for purposes of calculating its Forward Pricing Rates. Just to give you a hint of both the complexities and big numbers involved, let’s throw out some numbers from the June 2011 ASBCA decision.

• The market value of GD’s pension plan assets (as of 1 January 1994) was $2,122,371,000.

• The actuarial value of GD’s pension plan assets (as of 1 January 1994) was $1,697,968,000.

• The “expected” actuarial value of GD’s pension plan assets (as of 1 January 1995) was $1,772,424,000.

• The “projected” market value of GD’s pension plan assets (as of 1 January 1995) was $1,998,698,000.

Since 1986, GD consistently estimated the next year’s market value of its pension plan assets using the actual market performance from the preceding year. For Forward Pricing purposes, future years’ performance was estimated assuming an 8 percent return on assets. When GD updated Forward Pricing Rates during the year (as it did from time to time), it used actual Year-to-Date performance plus an estimated 8% return for the unknown balance of the year. The government acquiesced to GD’s methodology for 20 years, but in 2006 the government objected to GD’s methodology for the first time, asserting that it violated the requirements of CAS 412. The government’s issue was that CAS 412 allegedly required use of actuarial assumptions that reflect long-term trends, so as to avoid distortions caused by short-term fluctuations; whereas GD’s methodology reflected short-term results.

In its response, GD noted that, under the Truth-in-Negotiations Act (TINA) it was required to identify cost or pricing data that was “accurate, current, and complete”—and thus was required to update actuarial assumptions with actual cost information where known. GD also argued that its market return data was “historical fact” and not an actuarial assumption of any sort. GD also asserted that using actual market returns reflected its “best estimate” of pension costs to be incurred, and thus complied with the fundamental requirement of CAS 412.

Judge Peacock, writing for the ASBCA, found that a fact is not an assumption, when viewed in isolation. However, when viewed in the context of pension accounting, the rate of return of plan assets was an actuarial assumption subject to the requirements of CAS 412. Judge Peacock also found that there was no conflict between the CAS pension measurement requirements and the FAR Part 15 TINA requirements, writing—

There is no conflict between the CAS and FAR cost or pricing data requirements. Although the FAR does require submission of ‘accurate, complete and current’ cost or pricing data for consideration, it does not dictate the relative importance of submitted data or how that data will be used in cost estimation, negotiation and pricing. Appellant assumes that the most current data is also the most accurate and complete data. That may or may not be the case. Appellant's assumption is particularly problematic here because the RPFPRs [Revised Proposals for Forward Pricing Rates] specifically (and pension cost estimation generally) are intended to make projections a number of years into the future. Moreover, the government maintains that the relevant ‘current’ (as well as most accurate) data is that required for use in the CAS 412 measurement methodology.

GD lost. It filed a Motion for Reconsideration. It lost that Motion as well. Then it filed an Appeal at the Federal Circuit. It lost there as well. Notably the decision was divided, with what seemed to us to be a well-reasoned dissenting opinion from Judge Wallach.

Let’s first look at the majority ruling, which affirmed the ASBCA decision. As might be expected, the Appellate decision essentially reiterated the ASBCA’s logic. For example, Judge Lourie, writing for the majority, wrote that—

We agree with the government that General Dynamics’ use of midyear market values and the subsequent blended rate for the base year violate CAS 412-50(b)(4). First, both the midyear market value and the subsequent blended rate are actuarial assumptions. CAS 412-30(a)(3) defines an ‘actuarial assumption’ as an ‘estimate of future conditions affecting pension cost.’ As a matter of principle, we agree with General Dynamics’ proposition that the actual value of the plan assets on a given day is a historical fact, not an actuarial assumption. That historical fact, however, must be distinguished from the two decisions concerning which data point to use and how that data point affects the established rate. …

Contrary to General Dynamics’ assertion, the presumed accuracy of the midyear value in the base year does not make the use of that value and the subsequent blended rate compliant with CAS. Indeed, the ‘accuracy’ argument raised by General Dynamics ignores the fact that the forward pricing rate is not only for the base year, but for a projection from three to nine years into the future. Indeed, even if General Dynamics’ approach may be an accurate representation over the short term, that is only because it impermissibly reflects short-term fluctuations. General Dynamics’ method improperly locks in that short-term fluctuation causing a distortion that alters the level of growth throughout the rest of the projection.

Judge Wallach did not agree with the majority opinion. He wrote—

The ASBCA’s decision denying GD’s appeal rests on two invalid assumptions, either of which, if corrected, is sufficient to mandate reversal: First, GD’s use of current, intra-year data is not an ‘actuarial assumption’ within the meaning of CAS 412; Second, even if the data used is an actuarial assumption, the actuarial assumption does not result in ‘distortions caused by short-term fluctuations.’ CAS 412-50(b)(4). For each reason, the Government did not carry its burden to prove a CAS violation. …

There is no evidence that GD’s use of intra-year data results in distortions caused by short-term fluctuations. To the contrary, the record reveals that GD’s method has proven more accurate. We should not require companies to abandon decades-long practices that are compliant with the CAS for less accurate calculating methods suggested by the Government.

Unfortunately for General Dynamics, any future hopes rest with the Supreme Court. And, as we all know, SCOTUS does not agree to hear very many government contracts cases.

For the rest of you, consider this: General Dynamics is one of the largest defense contractors in the world. It has access to some of the best accountants, actuaries, and attorneys. And it still ran afoul of DCAA. GD’s situation ought to make you worried—because if GD can screw up CAS compliance, then so can you.