Hearing on Pooling Accounting

Prepared Testimony of Mr. Dennis Powell
Vice President and Corporate Controller
Cisco Systems

10:00 a.m., Thursday, March 2, 2000 - Dirksen 628


My name is Dennis Powell. I am Vice President and Corporate Controller for Cisco Systems, Inc. Prior to Cisco, I was with a Big Five accounting firm for 26 years serving many technology companies, as well as having responsibility as a National SEC Reviewer.

Cisco is the worldwide leader in networking for the Internet with revenues currently approximating $17 billion per year. We are a multinational corporation with more than 26,000 employees in 200 offices and 55 countries. In the U.S., we have significant operations in California, Texas, Massachusetts and North Carolina.


I would like to start with some background on Accounting for Business Combinations. The two methods of accounting - "Purchase" and "Pooling of Interests" - have been generally accepted in practice since 1945. In 1970, the Accounting Principles Board studied and discussed the pros and cons of the two accounting methods, and issued APB16 "Business Combinations", which concluded that they "find merit in the purchase and pooling of interests methods of accounting for business combinations and accepts neither method to the exclusion of the other." (APB16, paragraph 42).

This viewpoint was reaffirmed in 1994 by a task force commissioned by the American Institute of Certified Public Accountants to study the usefulness of financial reporting. This report, entitled "Improving Business Reporting - A Customer Focus" concluded, after three years of study, that the existence of the two methods is not a significant impediment to users' analysis of financial statements and "A project to do away with either method would be very controversial, require a significant amount of FASB time and resources, and in the end is not likely to improve significantly the usefulness of financial statements."

In rereading APB16, I found it interesting that arguments for and against the pooling versus purchase methods of accounting haven't changed over the past 30 years - we are still debating the same issues. However, while time has not changed the fundamental arguments for each method, time has changed the magnitude of the implications from the defects of the purchase method described in APB16, such as the difficulty in identifying and determining the fair value of intangible assets, including goodwill, and the accounting for goodwill after the business combination is completed.

These problems of the purchase method are still with us, and the implications today are much more severe than they were in 1970. In 1970, most of an acquisition price was allocated to tangible, hard assets. Today, for knowledge-based technology companies, most of the acquisition price is allocated to intangible assets - and very little allocated to hard assets. For example, since 1993, Cisco has acquired 50 companies amounting to $19 billion. Of these acquisitions, only $900 million, or 5%, is attributed to hard assets - $18 billion or 95% would be left to allocate to intangible assets or goodwill. So because over time the nature of acquisitions has changed in the New Economy, the limitations of the purchase method have become more problematic. And yet the new FASB proposal would force all acquisitions to be accounted for under the purchase method, without having solved its defects.


One of the defects of the purchase method is the accounting for goodwill once it is recorded as an asset on the balance sheet. The FASB proposal requires that goodwill be treated as a wasting asset, and be amortized ratably over 20 years. This model incorrectly assumes that goodwill declines in value over time, which artificially reduces net income and misrepresents economic reality. In reality, the value of goodwill is dependent upon the success of the merger, and is not a function of time.

For example, we studied four technology mergers that occurred in 1996 and 1997, which were reported as poolings. We then recast the poolings as if they were purchases, and restated the financial statements for periods after the acquisition to show the impact of goodwill amortization. Using the Computer Sciences merger with Continuum as an illustration (Attachment A), the impact of the goodwill amortization is to reduce actual net income by an average of 18% each year. This would suggest that goodwill has declined in value. However, over this same time period, goodwill actually increased by 42%. The results of all four companies in the study, as summarized on Attachment B, demonstrate the same impact. The purchase accounting model significantly reduced actual earnings by an average of 48% for the amortization of goodwill, which presumes that goodwill is declining in value. But in reality, the goodwill has significantly increased from the date of the merger by an average of 43%. In these examples, which are typical of successful mergers, the purchase model grossly misrepresents economic reality.

As a side note, an average of 90% of the purchase price have been allocated to intangibles, where the purchase model is inadequate to properly reflect the true values of the intangibles acquired.

On the other hand, not all mergers are successful. In one merger we studied, the goodwill declined from $7 billion to $1.6 billion in just three years after the acquisition.

Based on the above study, it is clear that in successful mergers, the presumption that goodwill is a wasting asset is not valid. Goodwill increases in successful acquisitions and declines rapidly in unsuccessful acquisitions. One thing goodwill does not do is decline ratably over twenty years - the FASB model simply does not report true economic performance. Similar observations may be made of other intangibles, such as brand names, where the value for many brands have increased - not declined - over time.


Regarding valuation of other intangible assets, my concern is that the FASB proposal obligates companies to identify and value all intangible assets, without giving adequate guidance on how these assets should be separately identified and valued.

For the past 1- years, I have been participating on an AICPA committee tasked with developing best practices around valuation of Purchased Research & Development. I learned from that experience that there are no standards in the valuation community to provide any consistency or reliability around the valuation of these intangibles.

For example, how do you separately value "customer service capability" or "presence in geographical locations" or "effective advertising programs"? How do you separately distinguish the value of "research & development" from "technological know-how"?

At risk is a loss of credibility in financial reporting. We must have more guidance and standards around how to identify and value intangibles.


The FASB has stated that elimination of pooling solves a comparability issue between purchase transactions and pooling transactions. But elimination of pooling simply trades one comparability issue with a set of new comparability problems. First, mandating the purchase method creates significant comparability issues between companies who grow from internal organic development and those who grow through acquisition.

For example, a company that generates significant goodwill from its internal operations will report no goodwill value while the company that acquires goodwill through a merger will report the "value" of the goodwill at the time of the acquisition. So, while both companies may have the same value of goodwill, only the company who obtained the goodwill through a merger will report any amount on its balance sheet.

Secondly, elimination of pooling prevents comparability within the same company - in comparing operations before the acquisition, which do not include the activities of the acquired company, to operations after the acquisition, which do include the activities of the acquired company. Eliminating pooling does not solve the comparability issue.

I believe the comparability issue would be more effectively addressed by correcting the inherent problems of the purchase method than by eliminating pooling accounting as an option.


In conclusion, the U.S. accounting rules for Business Combinations, which includes both the pooling and purchase methods, has for the past 50 years, generated and supported the strongest capital markets in the world. Before the FASB radically changes these accounting rules to a model that will certainly stifle technology development, impede capital formation and slow job creation in this country, the FASB should make sure the proposed new method is without question, the absolute, correct solution. In reality, the FASB's proposed standard does not improve the accounting - it merely changes it. Worse yet, the proposed changes require companies to use a purchase model that does not work for companies in the New Economy, where most of the acquisition value cannot be attributed to hard assets, forcing companies to report an arbitrary, artificial net income number that is irrelevant and misleading.

We believe the FASB should:

(1) Retain the pooling of interests method of accounting.

(2) Revise the purchase method to correct its deficiencies:

(a) Charge purchased goodwill directly to shareholders' equity or amortize it through comprehensive income;

(b) Limit the allocation of purchase price to only those intangibles that can be objectively and reliably valued;

(3) Engage a task force, which would include valuation experts, to develop adequate guidance on how to identify, value and account for intangible assets for New Economy companies.

Home | Menu | Links | Info | Chairman's Page