| Interpublic Group 1114 Avenue of the Americas New York, NY 10036
(212) 704-1200 www.interpublic.com |
May 10, 2010
|
VIA FAX AND EDGAR TRANSMISSION
Mr. Larry Spirgel
Division of Corporation Finance
U.S. Securities & Exchange Commission
100 F Street, NE
Mail Stop 3720
Washington, D.C. 20549
| Re: | The Interpublic Group of Companies, Inc. Form 10-K for the year ended December 31, 2009 Filed February 26, 2010 File No. 1-06686 |
Dear Mr. Spirgel:
Thank you for your April 26, 2010 letter to Frank Mergenthaler providing certain comments on the Annual Report on Form 10-K of The Interpublic Group of Companies, Inc. (“IPG” or “we”) for the year ended December 31, 2009 (the “2009 Annual Report”). We acknowledge the Division’s comments contained in the letter and, as discussed in the responses below, we propose to make certain changes in response to the comments in our Annual Report on Form 10-K for 2010 (the “2010 Annual Report”). Exhibit A to this letter presents our goodwill disclosure from the 2009 Annual Report, revised to show the proposed revisions as if they had been included in the 2009 Annual Report. The additional language to be included in the 2010 Annual Report appears in bold in Exhibit A.
Below are our responses keyed to your comments.
STAFF COMMENT No. 1 We note that goodwill represents 27% of your assets as of December 31, 2009. You disclose that changes in certain circumstances or conditions could result in future impairment charges. But it is unclear from your disclosure whether the goodwill in any of your reporting units is currently at risk of a potential impairment. Please disclose the following information for each reporting unit for which fair value is not substantially in excess of the carrying value:
| • | The percentage by which fair value exceeded carrying value as of the most recent step-one test; |
| • | The amount of goodwill allocated to the unit; |
| • | A description of the methodology used to determine fair value; |
| • | A description of key assumptions used and how the key assumptions were determined; and |
| • | A discussion of the uncertainty associated with the key assumptions and any potential events and/or circumstances that could have a negative effect on the key assumptions. |
Otherwise disclose in your critical accounting policies, if true, that none of your reporting units with significant goodwill is at risk of failing step one of the goodwill impairment test. Please provide us with your proposed disclosures.
COMPANY RESPONSE No. 1
IPG is a holding company of advertising agencies. All of our agencies share similar characteristics due to the common nature of their businesses and operating model. As a result, the methodologies used to determine fair value and the key estimates and assumptions used in our annual goodwill review are consistent for all of our reporting units. Therefore we have chosen to group our reporting units, for purposes of this disclosure, based on the percentage comparison of fair value to carrying value. The table on page 19 of our 2009 Annual Report discloses in the aggregate the number of reporting units whose fair value exceeded carrying value by less than 10%, 10% to 20%, or greater than 20%, as well as those units whose carrying value exceeded fair value. The aggregate goodwill associated with these reporting units is also noted in the table. See COMPANY RESPONSE No. 2 below for further information regarding the tables on Page 19.
Pages 18 and 19 of our 2009 Annual Report also provide the other disclosures requested by the staff’s comment. In particular:
| • | We disclose the methodology used to determine the fair values of our reporting units for our 2009 annual impairment review as using a combination of the income approach and the market approach. |
2
| • | We describe the key assumptions that apply to both the income and market approaches. Although the assumptions determined by management will vary based on specific factors of each reporting unit, the methodology used to determine fair value and the key assumptions and estimates are generally the same for all of our reporting units because they all operate within the advertising and marketing services industry. We disclose that our key assumptions include revenue growth, profit margins, terminal value growth rates, capital expenditures projections, assumed tax rates, discount rates, other assumptions deemed reasonable by management and relevant comparable company market multiples. We also disclose the range of discount rates used for our reporting units. As shown in Exhibit A, we propose to expand this disclosure in the 2010 Annual Report to provide the range of terminal value growth rates and revenue growth rates. |
| • | We disclose the uncertainties associated with our key assumptions and any potential events or circumstances that could have a negative effect on the key assumptions. As shown in Exhibit A, we propose to enhance this disclosure in the 2010 Annual Report by providing additional factors that could negatively effect key assumptions, such as significant client losses and changes in spending levels of our existing clients. |
STAFF COMMENT No. 2 We refer to the sensitivity analysis presented in connection with your discussion of the critical accounting estimates associated with goodwill. Please tell us in detail about the message you are trying to convey to your investors through three tables on page 19 representing the low-end, average and high-end valuation ranges.
COMPANY RESPONSE No. 2
On page 19 of our 2009 Annual Report we disclose that we use the average of our fair values for purposes of our comparison between carrying value and fair value of step-one of the impairment test. The tables on page 19 provide a comparison detailing the impact that changes in assumptions may have on the outcome of the first step of the impairment test. The analysis compares the fair value to the carrying value of our reporting units using three different fair value amounts: the average, low end and high end of the valuation range. The low end of the fair value range was determined by reducing growth rates by 0.5% and increasing discount rates by 0.5%, and the high end of the range was determined by increasing growth rates by 0.5% and decreasing discount rates by 0.5%. In each table we state the percentage by which fair value exceeded carrying
3
value or if carrying value exceeded the fair value for the reporting units tested. As noted above, we grouped the results of this analysis into the following categories: (a) fair value exceeded carrying value by i) less than 10%, ii) 10% to 20%, or iii) greater than 20%, or (b) carrying value exceeded fair value. For each category we quantified the number of reporting units in that category and the aggregate amount of goodwill allocated to the respective reporting units. We believe that the amounts in each category indicate to investors the maximum potential amount of goodwill at risk under the actual method used (displayed in the table headed “Average”) and the amount of goodwill that would potentially be subject to step two of a goodwill impairment review if we used the low or high end of fair values.
As shown in Exhibit A, we propose to expand the explanation of these tables in the 2010 Annual Report, including adding language that highlights in text form that the tables display, under varying valuation scenarios, the number of reporting units either whose carrying value exceeds fair value or whose fair value exceeds carrying value by less than 10%, and the aggregate amount of goodwill associated with each category.
We acknowledge that IPG is responsible for the adequacy and accuracy of the disclosure in the 2009 Annual Report; that comments of the Commission staff or changes to disclosure in response to such comments do not foreclose the Commission from taking any action with respect to the 2009 Annual Report; and we may not assert comments of the Commission staff as a defense in any proceeding initiated by the Commission or any person under the federal securities laws of the United States.
If you have any questions or require any additional information with respect to the above, please do not hesitate to contact me at (212) 704-1430.
| Sincerely, |
| /s/ Christopher F. Carroll |
| Christopher F. Carroll |
| Senior Vice President, Controller and Chief Accounting Officer |
4
EXHIBIT A
(Amounts in Millions)
Goodwill and Other Intangible Assets
As a result of changes in authoritative guidance, we account for our business combinations using the acquisition accounting method in 2009, while we utilized the purchase accounting method in prior years. Both accounting methods require us to determine the fair value of net assets acquired and the related goodwill and other intangible assets. See Note 1 to the Consolidated Financial Statements for further information. Determining the fair value of assets acquired and liabilities assumed requires management’s judgment and involves the use of significant estimates, including projections of future cash inflows and outflows, discount rates, asset lives and market multiples. Considering the characteristics of advertising, specialized marketing and communication services companies, our acquisitions usually do not have significant amounts of tangible assets as the principal asset we typically acquire is creative talent. As a result, a substantial portion of the purchase price is allocated to goodwill and other intangible assets.
We review goodwill and other intangible assets with indefinite lives not subject to amortization as of October 1 st each year and whenever events or significant changes in circumstances indicate that the carrying value may not be recoverable. We evaluate the recoverability of goodwill at a reporting unit level. We have 15 reporting units that were subject to the 2009 annual impairment testing. Our reporting unit structure has not changed from 2008 except for the movement of one reporting unit into another reporting unit due to an operational decision by management during 2009. In addition, Deutsch and Lowe were tested as separate reporting units during 2009 but will be combined into one reporting unit in 2010 as a result of their alignment. Our annual impairment reviews as of October 1, 2009 did not result in an impairment charge at any of our reporting units.
We review intangible assets with definite lives subject to amortization whenever events or circumstances indicate that a carrying amount of an asset may not be recoverable. Recoverability of these assets is determined by comparing the carrying value of these assets to the estimated undiscounted future cash flows expected to be generated by these assets. These assets are impaired when their carrying value exceeds their fair value. Impaired intangible assets with definite lives subject to amortization are written down to their fair value with a charge to expense in the period the impairment is identified. Intangible assets with definite lives are amortized on a straight-line basis with estimated useful lives generally between 7 and 15 years. Events or circumstances that might require impairment testing include the loss of a significant client, the identification of other impaired assets within a reporting unit, loss of key personnel, the disposition of a significant portion of a reporting unit, significant decline in stock price or a significant adverse change in business climate or regulations.
The authoritative guidance for goodwill specifies a two-step process for goodwill impairment testing and measuring the magnitude of any impairment. The first step of the impairment test is a comparison of the fair value of each reporting unit to its carrying value, including goodwill. The sum of the fair values of all our reporting units is reconciled to our current market capitalization plus an estimated control premium. Goodwill allocated to a reporting unit whose fair value is equal to or greater than its carrying value is not impaired, and no further testing is required. Should the carrying amount for a reporting unit exceed its fair value, then the first step of the impairment test is failed and the magnitude of any goodwill impairment is determined under the second step, which is a comparison of the implied fair value of a reporting unit’s goodwill to its carrying value. Goodwill of a reporting unit is impaired when its carrying value exceeds its implied fair value. Impaired goodwill is written down to its implied fair value with a charge to expense in the period the impairment is identified.
The fair value of a reporting unit for 2009 was estimated using a combination of the income approach, which incorporates the use of the discounted cash flow method, and the market approach, which incorporates the use of earnings and revenue multiples based on market data. Historically, we have used a combination of the income and market approach for our analysis. However, the fair value of a reporting unit for 2008 was estimated using only the income approach as we determined that the market approach was not appropriate due to the deterioration and extreme volatility of the credit markets in the latter part of 2008. For 2009, we determined that the market approach was appropriate to use and we primarily applied an equal weighting to the income and market approach for our analysis. For the income approach, we used projections, which require the use of significant estimates and assumptions specific to the reporting unit as well as
those based on general economic conditions. Factors specific to each reporting unit include revenue growth, profit margins, terminal value growth rates, capital expenditures projections, assumed tax rates, discount rates and other assumptions deemed reasonable by management. For the market approach, we used judgment in identifying the relevant comparable company market multiples.
These estimates and assumptions vary between each reporting unit depending on the facts and circumstances specific to that unit. The discount rate for each reporting unit is influenced by general market conditions as well as factors specific to the reporting unit and the discount rates we used for our reporting units were between 12.0% and 15.0%. The terminal value growth rates for our reporting units were between 3.0% and 3.5%. The terminal value growth rates represent the expected long-term growth rate for the advertising and marketing services industry, incorporating the type of services the reporting unit provides, and the global economy. The revenue growth rates for our reporting units used in our analysis were generally between 5.0% and 6.0%. Factors influencing the revenue growth rates include the nature of the services the reporting unit provides for its clients, the geographic locations in which the reporting unit conducts business and the maturity of the reporting unit. We believe that the estimates and assumptions we made are reasonable, but they are susceptible to change from period to period. Actual results of operations, cash flows and other factors will likely differ from the estimates used in our valuation, and it is possible that differences and changes could be material. A deterioration in profitability, adverse market conditions, significant client losses, changes in spending levels of our existing clients or a slower or weaker economic recovery than currently estimated by management could have a significant impact on the estimated fair value of our reporting units and could result in an impairment charge in the future.
We have performed a sensitivity analysis to detail the impact that changes in assumptions may have on the outcome of the first step of the impairment test. Our sensitivity analysis provides a range of fair value for each reporting unit, where the low end of the range reduces growth rates by 0.5% and increases discount rates by 0.5% and the high end of the range increases growth rates by 0.5% and decreases discount rates by 0.5%. We use the average of our fair values for purposes of our comparison between carrying value and fair value for the first step of the impairment test.
The following table shows the number of reporting units we tested in our 2009 and 2008 annual impairment reviews and the related goodwill value associated with the reporting units at the low end, average and high end of the valuation range for a) fair values exceeding carrying values by less than 10%, b) fair values exceeding carrying values between 10% and 20%, c) fair values exceeding carrying values by more than 20% and d) carrying values that exceed fair value.
| 2009 Impairment Test 1 |
2008 Impairment Test |
|||||||||||||
| Low End |
Low End |
|||||||||||||
| Fair value exceeds carrying value by: |
Number of reporting units |
Total goodwill at the reporting units |
Fair value exceeds carrying value by: |
Number of reporting units |
Total goodwill at the reporting units |
|||||||||
| Less than 10% |
3 | $ | 715.0 | Less than 10% | 3 | $ | 231.1 | |||||||
| 10% - 20% |
1 | 12.2 | 10% - 20% | 3 | 759.9 | |||||||||
| Greater than 20% |
4 | 1,493.7 | Greater than 20% | 7 | 1,899.1 | |||||||||
| Carrying value exceeds fair value |
5 | 341.8 | 2 | Carrying value exceeds fair value | 3 | 330.8 | 2 | |||||||
| Average |
Average |
|||||||||||||
| Fair value exceeds carrying value by: |
Number of reporting units |
Total goodwill at the reporting units |
Fair value exceeds carrying value by: |
Number of reporting units |
Total goodwill at the reporting units |
|||||||||
| Less than 10% |
6 | $ | 615.3 | Less than 10% | 5 | $ | 541.9 | |||||||
| 10% - 20% |
2 | 447.6 | 10% - 20% | 1 | 20.0 | |||||||||
| Greater than 20% |
4 | 1,493.7 | Greater than 20% | 10 | 2,659.0 | |||||||||
| Carrying value exceeds fair value |
1 | 6.1 | 2 | |||||||||||
| High End |
High End |
|||||||||||||
| Fair value exceeds carrying value by: |
Number of reporting units |
Total goodwill at the reporting units |
Fair value exceeds carrying value by: |
Number of reporting units |
Total goodwill at the reporting units |
|||||||||
| Less than 10% |
5 | $ | 430.4 | Less than 10% | 4 | $ | 535.7 | |||||||
| 10% - 20% |
4 | 638.6 | 10% - 20% | 2 | 26.2 | |||||||||
| Greater than 20% |
4 | 1,493.7 | Greater than 20% | 10 | 2,659.0 | |||||||||
| 1 | We did not test two reporting units in 2009 because we determined we could carry forward the fair value of the reporting unit from the last annual test, as the fair value significantly exceeded the book value. |
| 2 | We used the average fair value for purposes of our comparison between carrying value and fair value for the first step of the impairment test. |
The table above displays the information related to our 2009 annual impairment test. We use the average of our fair values for purposes of our comparison between carrying value and fair value for the first step of the impairment test. Our results for the comparison between carrying value and fair value at the average fair value indicated that one reporting unit’s carrying value exceeded its fair value, and there was $6.1 of goodwill associated with that reporting unit. As a result, we were required to perform the second step of the impairment test. Under the second step, the estimated fair value of the reporting unit was allocated among its assets and liabilities in order to determine the implied fair value of goodwill. We did not record an impairment charge in the fourth quarter of 2009, as we determined that the implied fair value of goodwill was greater than the carrying value for this reporting unit. Additionally, no carrying value adjustment was made to the intangible assets for this reporting unit as the estimated cash flows were greater than the carrying value of these assets on an undiscounted basis. Using the average fair value there were six reporting units whose fair value exceeded their carrying value by less than 10%, and there was $615.3 of goodwill associated with these reporting units.
Additionally, we performed a sensitivity analysis and reviewed the carrying values compared to the low and high end of the valuation range. Using the low end of the valuation range, the carrying values for five reporting units exceeded their respective fair values, and there was $341.8 of goodwill associated with these reporting units that would be subject to the second step of a goodwill impairment review. There were three reporting units whose fair values exceeded their carrying values by less than 10%, and there was $715.0 of goodwill associated with these reporting units. Using the high end of the valuation range, there were no reporting units whose carrying values exceeded their respective fair values. There were five reporting units whose fair values exceeded their carrying values by less than 10%, and there was $430.4 of goodwill associated with these reporting units.