February 9, 2007
Ms. Linda van Doorn
Senior Assistant Chief Accountant
Division of Corporation Finance
Securities and Exchange Commission
Mail Stop 4561
100 F Street, NE
Washington, DC 20549
RE: | Global Payments Inc. Form 10-K for the year ended May 31, 2006 Filed August 4, 2006 Form 10-Q for the quarter ended August 31, 2006 Filed October 6, 2006 File No. 1-16111 |
Dear Ms. van Doorn:
This letter sets forth the responses of Global Payments Inc. (the Company) to your comments with regard to the above-referenced filings. Our responses include the supplemental information that you requested in your letter dated December 22, 2006 and on our recent conference call. We appreciate your constructive comments to assist us in improving our filings, and we look forward to your response.
Form 10-K for the year ended May 31, 2006
Consolidated Financial Statements
Note 1 Summary of Significant Accounting Policies, page 45
Goodwill and Other Intangible Assets, page 49
1. | Comment: We have read your response to prior comment 1 and note that the annual attrition of your customer relationships ranged from 3% to 7%. Please further clarify whether your straight line amortization methodology has fully captured the attrition that has occurred in each asset group on a static pool basis. For all periods presented, please quantify the amount of amortization expense that would have been recognized had an accelerated method of amortization been applied and advise us of the actual amortization expense recognized. |
Response:
We believe that our straight-line amortization methodology has fully captured the attrition that has occurred for our significant customer-related intangible assets. To support this statement, we examined the three significant acquisitions that comprised approximately 90% of the net book value of these assets as of May 31, 2006, as summarized in the attached Exhibit A. Our historical customer acquisitions have focused on acquiring merchants and banks that either allow us entry into a particular |
market or that broaden our footprint in any existing market. The type of customer that we generally acquire is larger and more stable than those that we add to our business through internally generated sales activities. For each of these significant assets on a static pool basis, we examined recent annual cash flows and attrition related to these assets. We describe the results of this examination below.
The largest of our significant acquisitions relates to the purchase of approximately 200,000 merchant contracts in Canada. We refer to this acquisition as Canada, which was completed in 2001 (please see the notes to Exhibit A for a further description of this asset, which is comprised of two substantially similar acquisitions that were completed within seven months of each other). The financial models used to value these assets at the times of acquisition each assumed a 6% annual revenue attrition rate. To assess whether actual results have differed from this estimate, we calculated the annual revenue generated from the acquired merchants at the time of acquisition, and compared that amount to the recent annual revenue generated by those merchants that remain our customers (i.e., those that have not attrited).
Based on our analysis, we believe that our average annual revenue attrition for Canada has been consistent with our initial assumption of 6%. We attribute this relatively low level of attrition to the size of our Canadian market share (which we estimate is at more than 30%), the lack of other competitors (we believe there are only four other viable competitors in Canada), and the large concentration of stable national and regional merchants that we acquired. In addition, we believe that the actual cash flows produced by this asset (using certain assumptions and cost allocations that we believe are reasonable for this analysis) are higher than our initially expected cash flows for each period since the date of acquisition. We attribute this variance to an actual attrition rate that is substantially similar to our initial estimate, in addition to significantly lower actual costs than expected, primarily due to the integration of this asset into our existing infrastructure.
The next largest significant acquisition relates to the purchase of more than 200 bank contracts in the United States. We refer to this acquisition as MAPP, which was completed in 1996. The financial model used to value this asset at the time of acquisition assumed a 3% annual cash flow attrition rate. To assess whether actual results have differed from this estimate, we calculated the recent annual revenue generated by the remaining bank customers. We then calculated recent annual cash flows related to these customers using certain assumptions and cost allocations that we believe are reasonable for this analysis. We then compared these recent annual cash flow amounts to the annual cash flows generated by this asset at the time of acquisition.
Based on our analysis, we believe that our actual cash flow attrition for MAPP has been equal to or better than our initial assumption of 3%. We attribute this relatively low level of attrition to the nature of these customers (primarily large financial institutions) and the significant difficulty in switching from our technology platform to a competitors platform. In fact, our analysis showed a 0% attrition rate for the largest customer relationships that represented approximately 50% of MAPPs annual revenue at the time of acquisition. We further believe that the actual cash flows produced by this asset are higher than our initially expected cash flows for each period since the date of acquisition. We attribute this variance to the steady nature of
2
these customer relationships, in addition to significantly lower actual costs than expected as a result of cost savings from integration efforts.
The last of our three significant acquisitions relates to the purchase of a small number of bank contracts in the Czech Republic. We refer to this acquisition as MUZO, which was substantially completed in calendar 2004. The financial model used to value this asset at the time of acquisition assumed a 7% annual revenue attrition rate. This attrition was assumed at a higher rate than our previous experience with bank customers due to the expected loss of a certain large client that was to occur shortly after the acquisition. To assess whether actual results have differed from this estimate, we calculated the actual revenue, cash flow (using certain assumptions and cost allocations that we believe are reasonable for this analysis), and attrition relating to the top ten acquired bank contracts, which accounted for 92% of MUZOs calendar 2004 revenue. We performed this analysis for each fiscal year since the date of acquisition.
Due to our large market share in the Czech Republic (which we estimate is at more than 50%), the lack of other competitors (we believe there are only two other viable competitors in the Czech Republic), and the significant difficulty in switching from our technology platform to a competitors platform, we have not lost any of these top ten customer relationships as of February 9, 2007, including the certain large client that we had initially expected to lose. As such, our average annual attrition rate has been significantly less than 7%. As a result of significantly lower than expected attrition levels, in addition to costs that have been consistent with our initial estimates, our actual cash flows have been higher than initially expected for each fiscal year since the date of acquisition.
Based on these analyses, the actual attrition rates and cash flows from each acquisition on a static pool basis are either consistent with or more favorable than the attrition rates and the cash flow estimates included in the initial valuation of each customer-related intangible asset. Further, for each of these acquisitions, we have also assessed and confirmed the useful lives of these assets. For each asset on a static pool basis, the estimated cash flows (determined in a manner consistent with our cash flow analyses discussed above) over the remaining useful life of the asset exceed the carrying value of the asset. In addition, for each asset, the expected cash flows in the final years of the useful life are meaningful (i.e., they do not dissipate to an immaterial amount).
Regarding our amortization methodology, at the time of each of our significant customer-related acquisitions, we expected a stable stream of cash flows and, therefore, a stable pattern of economic benefit, as described in our letter dated December 13, 2006. As a result, we believe that straight-line amortization is an acceptable method for these assets, based on the guidance in APB 17 (for acquisitions completed prior to June 1, 2001) and paragraph 12 of SFAS 142 (for acquisitions completed on or after June 1, 2001). Per your request, however, we have attached Exhibit B and Exhibit C, which provide the amount of amortization expense for our significant customer-related intangible assets that would have been recognized had an accelerated method of amortization been applied during our fiscal years 2004, 2005, and 2006 and during our first and second quarters of fiscal year 2007. We have also compared these amounts to the straight-line amortization
3
expense recognized during these periods. Lastly, we have provided Notes on Exhibit B and Exhibit C to help you better understand the amounts presented.
As we described to you during our recent phone conversation, in determining amortization expense under the accelerated method, we utilized the expected cash flow for each year from the initial valuation model and divided that amount by the expected total cash flow over the estimated life of each asset (which for the Canada asset is longer than the straight-line period due to the life used in the initial valuation model). We multiplied that percentage by the gross book value of the asset to arrive at the amortization expense for that period. We believe this approach is in accordance with paragraph 12 of SFAS 142, as it measures the expected amount of benefit derived from the asset in each period. Further, as the actual cash flows from each of our significant acquisitions have exceeded the initial estimates, we believe that basing our approach on the information in the initial valuation is a reasonable and appropriate method.
As shown in Exhibit B and Exhibit C, the straight-line amortization method produced a similar amount of expense in all periods presented in our most recent Form 10-K and in our Form 10-Q for our most recent two quarterly periods. The same is also true on a cumulative basis for our fiscal year periods 1996 (the date of our earliest significant acquisition) through 2003. This similarity results from our use of a shortened expected life for straight-line amortization purposes for our Canada asset. This shortened expected life produces a higher amount of expense under the straight-line method compared to the accelerated method in all periods (including the early periods) over the life of the Canada asset. This variance, however, is largely offset by the additional expense produced under the accelerated method for our MAPP and MUZO assets compared to the straight-line method.
Per your request from our recent phone conversation, we have described below (and presented in Exhibit B and Exhibit C) the variances produced by the two methodologies discussed above. The annual expense variance between the two methods as a percentage of current year pre-tax income was 0.0% in fiscal year 2004, 0.1% in fiscal year 2005, and 0.0% in fiscal year 2006. On a cumulative basis from fiscal year 1996 to fiscal year 2003 (i.e., the impact to beginning fiscal year 2004 retained earnings), the after-tax variance between the two methods was 0.1% of ending fiscal year 2003 total equity. As such, we do not believe that these variances are significant.
In summary, based on the above analysis, we believe that our straight-line amortization approach to date is an acceptable method for these assets, considering the applicable guidance in APB 17 and paragraph 12 of SFAS 142. We understand, however, that you believe the accelerated method that we described above is a more appropriate approach. As such, we propose adopting this accelerated method as of December 1, 2006 (the first day of our third quarter of fiscal 2007) on a prospective basis for our significant customer-related intangible assets. Further, in accordance with paragraph 14 of SFAS 142, we intend to continue to track actual cash flow and attrition information for our significant customer-related intangible assets to ensure that the attrition rates experienced and cash flows generated are either consistent with or more favorable than our initial estimates, and, thus, that the pattern of amortization and the remaining useful lives are still appropriate.
4
Impairment of Long-Lived Assets, page 49
2. | Comment: We have read your response to prior comment 2 and note that your asset groupings are organized according to geographic business due to certain interdependencies that exist. We believe that it would be preferable to assess your customer lists and merchant contracts for impairment on an individual basis. If you continue to use the straight line method, then the assets should be tracked and the impairment analysis should be performed at the individual customer-related intangible asset level. If management believes the implementation of this model would be impractical, we would not object to the adoption of an accelerated method of amortization. Please advise us. |
Response:
As described in our letter dated December 13, 2006, the implementation of a lower level impairment testing model would be impractical given the interdependencies that exist between these assets and our operations. To summarize our prior explanation, as part of our acquisition strategy, we strive to fully integrate our acquisitions into our existing infrastructure, as we believe these acquired customers will deliver long-term benefits to us (i.e., the type of customers that we acquire generally do not have high attrition rates). As a result, we have a significant amount of interdependencies that impede our ability to produce identifiable cash flows specific to these assets.
Nevertheless, in our response to comment 1 above, we refer to cash flows relating to these assets on a static pool basis. These cash flows were calculated using certain assumptions and cost allocations that we believe are reasonable. If we were to use these calculated cash flows to assess impairment, there would be no impairment of these assets on a static pool basis at this time.
Due to the lack of precise, identifiable cash flows for these assets on an individual basis, we believe that it is appropriate under paragraph 10 of SFAS 144 to continue testing for impairment at a geographic level. We will continue, however, to use a diligent and comprehensive approach to track these assets on an individual basis using all available sources of data and financial indicators for purposes of identifying indications of possible impairments and in assessing useful lives. In addition, as described in our response to comment 1 above, we intend to adopt an accelerated method of amortization on a prospective basis for our significant customer-related intangible assets.
5
Finally, the Company hereby acknowledges that it is responsible for the adequacy and accuracy of the disclosures in the filings, staff comments or changes to disclosures in response to staff comments do not foreclose the Commission from taking any action with respect to the filings, and the Company may not assert staff comments as a defense in any proceeding initiated by the Commission or any person under the federal securities laws of the United States.
Again, we appreciate your assistance in enhancing the overall disclosure in our filings. Please do not hesitate to contact me or Martin Picciano, our Chief Accounting Officer, at 770-829-8000 with any questions or further comments.
Sincerely, |
/s/ Joseph C. Hyde |
Joseph C. Hyde Chief Financial Officer Global Payments Inc. |
6
Exhibit A
Customer-Related Intangible Assets
As of May 31, 2006
(in millions)
Acquisition Name |
Customer Type |
Calendar Year |
Gross Book Value |
Accumulated |
Net Book Value |
% of Total NBV | ||||||
Canada |
Merchants | 2001 | $89.9 | ($26.9) | $63.0 | 51% | ||||||
MAPP |
Banks | 1996 | 61.4 | (20.8) | 40.6 | 33% | ||||||
MUZO |
Banks | 2004 | 8.0 | (1.1) | 6.9 | 6% | ||||||
Subtotal |
159.3 | (48.8) | 110.5 | 90% | ||||||||
All Other |
Various | Various | 91.6 | (78.6) | 13.0 | 10% | ||||||
Total |
$250.9 | ($127.4) | $123.5 | 100% | ||||||||
Notes:
The Canada asset consists of two acquisitions completed in March 2001 and October 2001. These businesses were substantially similar in all respects, including a shared customer base. In many circumstances, one of these businesses provided processing services for VISA volume for a particular merchant customer, while the second business provided the same service for MasterCard volume for the same merchant customer.
Exhibit B
Significant Customer-Related Intangible Assets
Annual Amortization Expense Comparison
(in millions)
Straight-Line | Accelerated | Variance | % of Pre-Tax Income | ||||||||
Fiscal 2004 |
|||||||||||
Canada |
$ | 3.7 | $ | 3.1 | $ | 0.6 | 0.6% | ||||
MAPP |
2.0 | 2.6 | (0.6) | (0.5%) | |||||||
MUZO |
0.1 | 0.2 | (0.1) | (0.1%) | |||||||
Total |
$ | 5.9 | $ | 5.9 | ($ | 0.0) | (0.0%) | ||||
Fiscal 2005 |
|||||||||||
Canada |
$ | 3.7 | $ | 3.0 | $ | 0.7 | 0.5% | ||||
MAPP |
2.0 | 2.6 | (0.6) | (0.4%) | |||||||
MUZO |
0.5 | 0.9 | (0.4) | (0.2%) | |||||||
Total |
$ | 6.2 | $ | 6.5 | ($ | 0.2) | (0.1%) | ||||
Fiscal 2006 |
|||||||||||
Canada |
$ | 3.7 | $ | 2.9 | $ | 0.8 | 0.4% | ||||
MAPP |
2.0 | 2.6 | (0.6) | (0.3%) | |||||||
MUZO |
0.5 | 0.8 | (0.3) | (0.1%) | |||||||
Total |
$ | 6.2 | $ | 6.3 | ($ | 0.1) | (0.0%) | ||||
Straight-Line (After Tax) |
Accelerated (After Tax) |
Variance | % of FY03 Equity | ||||||||
Cumulative Basis FY 1996 to FY 2003 |
|||||||||||
Canada |
$ | 4.9 | $ | 3.8 | $ | 1.1 | 0.3% | ||||
MAPP |
9.5 | 10.1 | (0.7) | (0.2%) | |||||||
Total |
$ | 14.4 | $ | 14.0 | $ | 0.4 | 0.1% | ||||
Notes:
Numbers may not add due to rounding. Pre-tax income represents income before income taxes and minority interest. The three assets above accounted for approximately 90% of the net book value of our customer intangible assets as of May 31, 2006. The MUZO acquisition was substantially completed during FY 2004. As such, the FY 2004 amounts for MUZO reflect partial year activity. The Canada amortization assumes a constant foreign exchange rate from the time of the acquisition under both methods. Due to increases in the CAD / USD exchange rate, the actual amortization expense recorded for Canada in fiscal years 2004, 2005, and 2006 was $4.3 million, $4.6 million, and $5.0 million, respectively. These foreign currency differences would not impact the variances between the two methods shown above. Please see the Notes on Exhibit A for other information relating to these assets.
8
Exhibit C
Significant Customer-Related Intangible Assets
Fiscal 2007 Quarterly Amortization Expense Comparison
(in million)
Straight-Line | Accelerated | Variance | % of Pre-Tax Income | ||||||||
Fiscal 2007 Q1 |
|||||||||||
Canada |
$ | 0.9 | $ | 0.7 | $ | 0.2 | 0.3% | ||||
MAPP |
0.5 | 0.7 | (0.1) | (0.2%) | |||||||
MUZO |
0.1 | 0.2 | (0.0) | (0.1%) | |||||||
Total |
$ | 1.6 | $ | 1.5 | $ | 0.0 | 0.0% | ||||
Fiscal 2007 Q2 |
|||||||||||
Canada |
$ | 0.9 | $ | 0.7 | $ | 0.2 | 0.4% | ||||
MAPP |
0.5 | 0.7 | (0.1) | (0.3%) | |||||||
MUZO |
0.1 | 0.2 | (0.0) | (0.1%) | |||||||
Total |
$ | 1.6 | $ | 1.5 | $ | 0.0 | 0.0% | ||||
Notes:
Numbers may not add due to rounding. Pre-tax income represents income before income taxes and minority interest. The three assets above accounted for approximately 90% of the net book value of our customer intangible assets as of May 31, 2006. The Canada amortization assumes a constant foreign exchange rate from the time of the acquisition under both methods. Due to increases in the CAD / USD exchange rate, the actual amortization expense recorded for Canada was $1.3 million in both the first quarter and second quarter of fiscal 2007. These foreign currency differences would not impact the variances between the two methods shown above. Please see the Notes on Exhibit A and B for other information relating to these assets.
9