corresp6-09.htm

[AT&T Inc. Letterhead]

AT&T Inc.

208 S. Akard

Dallas, Texas 75202

June 8, 2009

Securities and Exchange Commission

100 F Street NE

Washington, D.C. 20549

Re: AT&T Inc.

Form 10-K for the Year Ended December 31, 2008

Filed February 25, 2009

Form 10-Q for the Quarterly Period Ended March 31, 2009

File No. 1-8610

Ladies and Gentlemen:

We are in receipt of the letter from the Staff (the “Staff”) of the Securities and Exchange Commission (the “Commission”), dated May 18, 2009, to Mr. Richard G. Lindner, of AT&T Inc. (the “Company” or “we”). For your convenience, we have included the Staff’s comments herein and keyed our responses accordingly. We appreciate the Staff granting an extension until June 12, 2009.

We acknowledge that the adequacy and accuracy of the disclosure in the filing is our responsibility. We acknowledge the Staff’s position that comment or changes in response to the Staff comments in the proposed disclosure in the Forms 10-K and 10-Q do not foreclose the Commission from taking any action with respect to the filing. We also understand the Staff’s position that the Staff comments may not be asserted as a defense in any proceeding initiated by the Commission or any person under the federal securities laws of the United States.

The Company’s responses to the Staff’s comments are set forth below and revised disclosure has been written in bold type.

Item 3. Legal Proceeding, page 10

We note your disclosure regarding material pending legal proceedings. We also note news reports regarding other litigation to which you are a party, including Engers v. AT&T Management Pension Plan, filed in federal district court in New Jersey, and AT&T v. Hulteen, before the U.S. Supreme Court. Please provide an analysis as to why you have not discussed these cases in Item 3 of your Form 10-K. See Item 103 of Regulation S-K.

Response:

We did not include the two cases mentioned above because we believe the potential liabilities do not exceed the threshold set forth in Instruction 2 to Item 103 of Regulation S-K. With respect to the Hulteen case, the U.S. Supreme Court decided that case in our favor on May 18, 2009.

Management’s Discussion and Analysis of Financial Condition and Results of Operations

General

2.	Your ability to raise capital may be impacted by current credit ratings. With a realistic evaluation of your current financial situation, please discuss how the following items may impact current and future results:

·	changes in credit lines and credit availability;

·	parties with which you have credit lines;

·	whether backup credit lines remain available; and,

·	goodwill and license impairments

Please provide us with your proposed disclosures.

Response:

We understand the Staff’s focus on disclosure concerning capital markets access in the current economic environment. We are and expect to continue to be able to access both the short-term and long-term debt markets. As discussed in the liquidity section of our 2008 Annual Report and first quarter 2009 Form 10-Q, we have issued $7 billion in total of long-term debt during the fourth quarter of 2008 and the first quarter of 2009 as well as an issuance of over $2.5 billion that closed on April 30, 2009. Also as discussed in these filings, our existing credit facility remains available until July 2011 and has remained undrawn. We have not had a goodwill or license impairment subsequent to the initial adoption of FAS 142 in January 2002 at which time we recorded a cumulative effect related to the change in accounting standards for impairment. Additionally, an impairment of goodwill or licenses would not affect the debt/EBITDA ratio in the backup credit facility covenant. Accordingly, we believe that our existing disclosure adequately addresses our financial condition. While we believe our financial liquidity and ability to borrow have remained, and we expect to remain, stable during this time period, we monitor these trends. Should these trends turn unfavorable we would address them in an upcoming filing as may be required. We have addressed the potential impact of an unfavorable change in the financial markets on our operations in a Risk Factor addressing capital market access in our 2008 Annual Report.

Liquidity and Capital Resources, page 23

3.	Please provide a range for your capital expenditures for 2009 by project including those that you may decide to delay to a future period. Disclose the nature and amounts of your discretionary and non-discretionary capital expenditures. Please provide us with your proposed disclosures.

Response:

We provided a range of our expected 2009 capital expenditures ($17-$18 billion) in the liquidity section of our first quarter 2009 Form 10-Q, and we discussed in the liquidity section of our 2008 Annual Report the allocation between segments of our capital expenditures during 2008. As noted in the Expected Growth Areas section of our Annual Report and the liquidity section of our first quarter 2009 Form 10-Q, we expect our growth areas to remain the same for 2009 and to continue to focus our capital expenditures in those areas. We do not manage our business in terms of discretionary and non-discretionary expenditures and our guidance is based on amounts that we fully expect to spend. (To the extent an expenditure would be mandated by regulatory requirements, we would comply with such an order.) We expect to remain within our targeted range for 2009 and we currently anticipate we will be including our expected range for 2010 in our 2009 Annual Report to Shareowners.

4.	Please provide a detailed analysis of how you anticipate meeting your cash flow requirements in the next two years.

Response:

As discussed in the liquidity section of our Annual Report and first quarter 2009 Form 10-Q, we expect to meet our cash flow requirements from funds from operations and incremental borrowings. As shown in our first quarter 2009 Form 10-Q, we have significant positive cash flow and have had ready access to the capital markets. We do not expect difficulties in sustaining a positive cash flow during the next two years. Accordingly, we believe our current disclosure does address our short term cash flow requirements.

Contractual Obligations, Commitments and Contingencies, page 26

5. Please provide us with your estimated range of the items not included in the table.

Response:

In this section of our 2008 Annual Report, the items we excluded in accordance with Regulation S-K generally fell into one or more of following categories: excluded because of immateriality; excluded because of uncertainty of timing of payment; excluded as postemployment benefit obligation; and excluded because they do not require the use of cash.

Cash items not included in the table but mentioned elsewhere in the Annual Report include capital leases with a present value of $167 million (noted in Note 8) and non-qualified retirement plans of $2.1 billion (noted in Note 11).

Minority interest (disclosed in our Form 10-Q as $403 million as of the end of 2008 and now included in equity) was one of the items excluded due to uncertain timing. Other excluded items which may require cash but whose timing is uncertain were insurance losses, asset retirement obligations, and other miscellaneous contingencies totaling between $400 and $500 million.

Excluded items not requiring cash included deferred compensation that will be paid in stock of $1 billion (noted in Note 11).

Most of the remaining liabilities not requiring cash related to deferred revenue and gains including the non-current portions of activation revenues deferred under SAB 104 and also included purchase accounting acquisition adjustments related to unfavorable lease obligations of $134 million. Collectively these totaled between $3.5 and $3.8 billion.

6. Expand your discussion to disclose in detail your debt covenants here and in Note 8.

Response:

Because of our strong credit ratings, we have relatively limited debt covenants. We believe our only significant debt covenant is a debt-to-EBITDA limitation contained in our backup credit facility, a facility that we have never drawn on. This covenant is discussed in the Liquidity section of our Annual Report and first quarter 2009 Form 10-Q as well as in Note 8 of the notes to the financial statements in the Annual Report. We do not believe that the other few covenants in our debt instruments which are incorporated as exhibits to our 2008 Form 10-K are material. In addition, all significant debt instruments are publicly available in filings under the issuing company and a listing of our individual debt issuances is posted on our website. Considering all of these factors, we believe our current disclosures are adequate.

Accounting Policies and Standards

Significant Accounting Policies and Estimates, page 20

Pension and Postretirement Benefits page 20

Please discuss in detail the development of the discount rates used for employee benefit plans and the methodology for developing the hypothetical yield curves, including the actual high-quality, fixed income corporate bonds used. Please provide us with the proposed disclosures. Please expand your disclosure in Note 11.

Response:

As discussed in Note 11 to our Annual Report, we determine our discount rate for U.S. pension and postretirement welfare plans primarily based on a December 31, 2008 hypothetical high-quality corporate bond yield curve. This yield curve is determined by the plan’s actuaries at J.P. Morgan. The J.P. Morgan Aa above median yield curve is developed from more than 300 corporate bonds, all in Barclay’s Point database, with a market value in excess of $200 billion. All of the bonds share these characteristics:

• US dollar denominated

• rated Aa1, Aa2 or Aa3 by Moody’s or rated AA+, AA or AA- by Fitch or S&P

• fixed coupon

• noncallable

• nonconvertible

• not index linked

• coupon frequency not “issued at a discount”

We did not include any bonds with yields ranging from 13% to 23.5% that would have increased the aggregate value of the bonds we were looking at by more than $6 billion.

Prior to feeding bond data into the model, the actuaries added any accrued interest to the “clean price” used to quote prices in the market. Accrued interest is calculated by taking the fraction of the coupon period that has elapsed, using a 30/360 US (NASD) day basis, times the coupon payment. These bonds were fed into the J.P. Morgan yield curve model which is very similar to a Cairns model.

We reviewed the curves produced by the actuaries’ testing, including the forward curves that were implicit in the overall curve. As a check, we also looked at another commonly used publicly published curve (the Citigroup Pension Discount Curve – Above Median) including that curve’s implicit forward rates. We believe that the J.P. Morgan curve produced the more reasonable “curve shape” as well as the most reasonable view of forward rates. The other curve implied forward rates that were very low, including some that were actually negative, for years in the 20 to 30 year range.

Accordingly, we determined that the curve developed by J.P. Morgan was an appropriate basis for measuring U.S. pension and postretirement benefit obligations at December 31, 2008. Applying this curve to projected cash flows under the plans produced a weighted average effective discount rate of 7.11% for U.S. pension obligations and 7.08% for U.S. postretirement benefit obligations. As we have consistently used discount rates rounded to the quarter percent, we selected a rate of 7.00% to measure U.S. pension and postretirement benefit obligations as of December 31, 2008.

In future filings, we will expand the discussion in the footnotes to provide some more information on the composition of the yield curve and then repeat that information in the MD&A section. The proposed disclosure is as follows:

Discount Rate Our assumed discount rate of x.xx% at December 31, 200x, reflects the hypothetical rate at which the projected benefit obligations could be effectively settled or paid out to participants on that date. We determined our discount rate based on a range of factors, including a yield curve comprised of the rates of return on [several hundred] high-quality, fixed-income corporate bonds available at the measurement date and the related expected duration for the obligations. These bonds were all rated at least Aa3 or AA- by one of the nationally recognized statistical rating organizations, denominated in U. S. dollars, and neither callable, convertible nor index linked. For the year ended December 31, 200x, we in/de/creased our discount rate by x.xx%, resulting in a de/in/crease in our pension plan benefit obligation of $x,xxx and a de/in/crease in our postretirement benefit obligation of $x,xxx. For the year ended December 31, 200x (the prior year), we in/de/creased our discount rate by x.xx%, resulting in a de/in/crease in our pension plan benefit obligation of $x,xxx and a de/increase in our postretirement benefit obligation of $x,xxx. Should actual experience differ from actuarial assumptions, the projected pension benefit obligation and net pension cost and accumulated postretirement benefit obligation and postretirement benefit cost would be affected in future years.

Asset Valuation and Impairments, page 20

8.	We note that goodwill accounted for 27% of total assets as of December 31, 2008. Tell us and disclose when you performed your impairment tests.

In light of the significance of your goodwill balance, we expect robust and comprehensive disclosure in your critical accounting policies regarding your impairment testing policy. This disclosure should provide investors with sufficient information about management’s insights and assumptions with regard to the recoverability of goodwill. Specifically, we believe you should provide the following information:

·	Provide a more detailed description of the steps you perform to review goodwill for recoverability.

·	Disclose a breakdown of your goodwill balance as of December 31, 2008 by reporting unit.

·	Disclose any changes to your reporting units or allocations of goodwill by reporting unit and the reasons for such changes.

Describe the significant estimates and assumptions you employed under the discounted cash flows methodology to determine the fair value of your reporting units in your impairment analysis. For example, you should disclose the discount rates for each reporting unit and how those discount rates were determined, including your consideration of any market risk premiums. In addition, we believe your disclosure should address your estimates of future cash flows, as follows:

1)	Provide a discussion of your historical cash flow growth rates and explain how your historical growth rates were considered when determining the growth rate to be utilized in your cash flow projections.

2)	Disclose the growth rate you need to achieve in your cash flow projections in order to avoid having a goodwill impairment charge.

3)	In view of the current economic environment, discuss how you considered the uncertainties inherent in your estimated future growth rates. For example, you should explain if and when you anticipated a recovery of the economy in your growth rates used in your cash flows analysis.

·	Describe changes to the assumptions and methodologies, if any, since your last annual impairment test.

For further guidance, refer to Release No. 33-8350 “Interpretation: Commission Guidance Regarding Management’s Discussion and Analysis of Financial Condition and Results of Operations.” Please provide us with the proposed disclosures.

Response:

Consistent with FAS 142, Goodwill and Other Intangible Assets, and as disclosed in Note 1 and Note 6 of our 10-K and in our MD&A section, we test goodwill for impairment on an annual basis. For fiscal year 2008, the goodwill impairment test was performed as of the beginning of business on October 1, with the actual testing taking place in the fourth quarter, consistent with prior years. We would also test more frequently if warranted by specific facts and circumstances.

In future filings we will place more information into the MD&A section of our document, including repeating certain items which are included in the footnotes to the financial statements including the notation that for much of our goodwill, our reporting units coincide with our operating segments. Following is an example of the type of disclosure that we would include in next year’s Form 10-K assuming we neither change our reporting units nor our method of determining fair value. We monitor both so changes in how we operate our business could lead to changes in reporting units and economic or other circumstances might warrant changing the method of evaluating fair value.

Goodwill is not amortized but is tested for impairment annually in the fourth quarter. Goodwill is tested on a reporting unit basis, and our reporting units generally coincide with our segments, except for certain operations in the Other segment. If there are indications of significant decreases in fair value, testing may also be done sooner than the annual test.

Goodwill impairment testing is a two step process. The first step involves determining the fair value of the reporting unit and comparing that to the book value. If the fair value exceeds the book value, then no further testing is required. If the fair value is less than the book value, then a second step is performed.

In the second step, the fair values of all of the assets and liabilities of the reporting unit, including those that may not be currently recorded, are determined. The difference between the sum of all of those fair values and the overall reporting unit’s fair value is a new implied goodwill amount that is compared to the recorded goodwill. If implied goodwill is less than the recorded goodwill, then an impairment to the recorded goodwill is recorded. The amount of this impairment may be more or less than the difference between the overall fair value and book value of the reporting unit. It may even be zero if the fair values of other assets are less than their book values. Goodwill is the only asset that may be impaired when testing goodwill.

As shown in Note x, more than 95% of our goodwill resides in the Wireline, Wireless and Advertising Solutions segments. For each of those segments, publicly traded companies whose services are consistent with those primarily offered by the segment exist, giving a market indication of enterprise value. Enterprise value is the sum of a company’s equity and debt values. One standard valuation technique is to determine enterprise value as a multiple of a company’s operating income before depreciation and amortization. We determined the multiples of the public companies and then calculated a weighted average of those multiples. Using those weighted averages, we then calculated fair values for each of those segments to determine if additional testing was required and, in all circumstances, no additional testing was required.

In the event of a x% drop in the fair values of the reporting units, the fair values would have still exceeded the book values of the reporting units and additional testing would still have not been required.

If we decide to change either reporting units or the method of doing the testing, we will discuss any change and give the reasons for the change. If the sensitivity test would require a phase 2 test, we would explain that additional testing.

We note that wireless licenses accounted for 17% of total assets as of December 31, 2008. Please tell us and disclose when you performed your annual impairment test and concluded that wireless licenses were not impaired. Tell us whether you performed subsequent interim impairment tests. If you did not, tell us why, addressing the factors in paragraph 8 of SFAS 144. You should discuss in your critical accounting policies and estimates the factors you considered in determining why no interim impairment testing under SFAS 142 was required.

In light of the significance of your wireless licenses balance, we expect robust and comprehensive disclosures in your critical accounting policies regarding your impairment testing policy. This disclosure should provide investors with sufficient information about management’s insights and assumptions with regard to the recoverability of wireless licenses. Specifically, we believe you should provide the following information:

·	Disclose the carrying value of the intangible asset for each unit of accounting.

Describe the nature of the valuation techniques you employed in performing the impairment tests. If you used discounted cash flow methodology, addressing EITF D-108, describe the method you used to isolate the cash flows associated with the intangible asset. If you used a hypothetical build-up or start-up method, describe qualitatively and quantitatively the significant estimates and assumptions used in your valuation method to determine the fair value of each unit of accounting in your impairment analysis.

Quantitatively and qualitatively describe in detail the changes in the estimates used in your assumptions to determine the fair value of your units of accounting since your last impairment test. In addition tell us and disclose how the assumptions in your most recent test were impacted by the current economic environment.

Provide a sensitivity analysis showing the impact on your impairment test resulting from a one percent change in each of your significant assumptions. For example, you should separately quantify the impact of a one percent decline in your revenue growth rates, one percent decline in your net cash flows and a one percent increase in your discount rate.

·	Provide a sensitivity analysis that discloses the impairment amount that would have resulted from hypothetical reductions in the fair value of your licenses at the time of your impairment testing.

Please provide us with the proposed disclosures.

Response:

Consistent with SFAS 142, Goodwill and Other Intangible Assets, and as disclosed in Note 1 and Note 6 of our 10-K and in our MD&A section, we test our wireless licenses for impairment on an annual basis. Also, consistent with EITF 02-7, we test the licenses on an aggregate basis, consistent with the management of the business on a national scope. For fiscal year 2008, the impairment test was performed as of the beginning of business on October 1, with the actual testing taking place in the fourth quarter, consistent with prior years. We would also test more frequently if warranted by specific facts and circumstances.

Those circumstances do not currently exist under the guidance of paragraph 8 of FAS 144. Market prices for FCC licenses have shown no significant decline in value, nor has there been a material change in our wireless business model or in the way we conduct our wireless business. We continue to invest in our network and expand our coverage area and we have not seen a significant adverse change in legal or regulatory factors that would impact our business model. Additionally, our historical and projected operation and financial trends do not show a need for interim evaluation. Finally, we have no history or current circumstances of cash flow losses in our wireless business.

Wireless licenses are not amortized but are tested for impairment annually in the fourth quarter. The licenses are tested on an aggregate basis, consistent with the management of the business on a national scope. If there are indications of significant decreases in fair value of the licenses, testing may also be done sooner than the annual test. There were no indications of this in 2009.

Consistent with prior years, we performed our test of the fair values using a discounted cash flow model (the Greenfield Approach), which assumes the licenses are the only assets owned at the outset (i.e., a "raw license"). This model then incorporates cash flow assumptions regarding investment in the network, development of distribution channels and the subscriber base, and other inputs for making the business operational. The assumptions which underlie the development of the network, subscriber base and other critical inputs of the discounted cash flow model were based on a combination of average marketplace participant data and our historical results, trends, and business plans. Operating metrics such as capital investment per subscriber, acquisition costs per subscriber, minutes of use per subscriber, etc. were also used to develop the projected cash flows. Since the cash flows associated with these other inputs were included in the annual cash flow projections, the present value of the unlevered free cash flows of the segment, after investment in the network, subscribers, etc., is attributable to the licenses. A portion of the terminal value of the segment, which incorporates an assumed sustainable growth rate, is also discounted and is likewise attributed to the licenses. The discount rate used to calculate the present value of the projected cash flows is based on the optimal long-term capital structure of a market participant and its associated cost of debt and equity. The discount rate utilized in the analysis is also consistent with rates we use to calculate the present value of the projected cash flows of licenses acquired from third parties.

If either the projected rate of growth of cash flows or revenues were to decline by 1%, or if the discount rate were to increase by 1%, the fair values of the licenses, while less than currently projected, would still be higher than the book value of the licenses. If there had been a x% decrease in the fair value of the licenses, the fair value would still have exceeded the book value.

Notes to Consolidated Financial Statements

Note 4. Segment Information, page 47

10.

We note that you have a reportable segment, Wireline, for which you report significant data regarding two components, Voice and Data. Please tell us your consideration of the guidance in paragraphs 10 through 15 for SFAS 131. Furnish us with the reports provided to your chief operating decision maker (CODM).

Response:

In our Wireline segment, Voice and Data are descriptions of two types of products which we sell. Both product types are generally offered over the same network, utilizing the same personnel. So while the revenues may be separately reported, the joint nature of the costs do not allow for the type of reporting and review that is contemplated by paragraphs 10 through 15 of FAS 131.

We will supplementally provide under separate cover the reports requested.

Note 6. Goodwill and Other Intangible Assets, page 50

11.

We note that you have determined that your reporting units for purpose of testing impairments are your operating segments. Furthermore, we note that you have discussed the operations of two components of your Wireline segment-Voice and Data. Please note that paragraph 30 of SFAS 142 states an operating segment shall be deemed to be a reporting unit if all of its components are similar, if none of its components is a reporting unit, or if it comprises only a single component. Tell us why you believe that you do not have reliable financial and cash flow information at lower levels. Tell us how you prepare your regulatory reports and the state and other levels. Tell us the source(s) of the data that you have disclosed regarding the various components of your business.

Response:

As noted in our response to question 10, Voice and Data represent product groupings rather than separate lines of business as the costs associated with the products are generally common. In general, the discussion of our business below the Wireline level has been included to describe and provide analysis on various groupings of product sales. This information is drawn from many sources, primarily our multiple billing systems across various legal entities. We accumulate this revenue data by similar characteristics to be able to provide our shareholders with additional analytical detail below total revenue.

Our regulatory reporting is done under differing accounting principles (including one state that has never adopted FAS 87 for pension accounting) and is specific to legal entity rather than total operations in those jurisdictions. For example, in certain jurisdictions our broadband DSL product sales to the customer occur in a different legal entity than the local telephone company, which has the predominance of the telephone plant. Those consumer revenues would not be included in the local telephone company regulatory reports. Accordingly, information provided on regulatory reports is considered strictly compliance information rather than operational financial information to be utilized in running the company.

Note 11. Pension and Postretirement Benefits, page 57

12.	We note the different assumptions for your U.S. and non-U.S. benefit plans. Tell us and disclose in future filings the rationale for the different assumptions.

Response:

We have qualified pension plans in a dozen European, Asian and African countries, primarily European. As noted in Note 11, the total across all countries of the plan assets and the liabilities as of December 31, 2008 were both less than $800 million, relatively small compared to our domestic plans. Our discount rates assumptions for each country are reviewed and determined based on bond rates for each country, and our expected rates of return are based on the investment mix in the plan assets for each country including the effect of any investments required by the plans.

In future Form 10-Ks, we will enhance the disclosures to include the following paragraph leading to the assumptions table:

In determining the projected benefit obligation for certain non-U.S. defined benefit plans, we use assumptions based upon interest rates relative to each country in which we sponsor a plan. Additionally, the expected return is based on the investment mix relative to each plan’s assets. Following are the significant weighted-average assumptions:

Please do not hesitate to contact me at (214) 757-3220 if you should have further questions.

Sincerely,

/s/ John J. Stephens

John J. Stephens

Senior Vice President and

Controller