Confidential Treatment is requested by Raytheon Company pursuant to 17 CFR 200.83. This response letter omits confidential information that was delivered separately to the Division of Corporate Finance. The omissions are denoted with three asterisks (***).
 | Michael J. Wood | Raytheon Company | ||
Vice President, Controller and | 870 Winter Street | |||
Chief Accounting Officer | Waltham, Massachusetts | |||
781.522.5833 | 02451-1449 USA | |||
781.522.6411 fax | ||||
July 26, 2013
Via EDGAR Filing and U.S. Mail
Mr. Lyn Shenk
Branch Chief
United States Securities and Exchange Commission
Division of Corporation Finance
100 F Street NE, Mail Stop 3561
Washington, DC 20549
Re: Raytheon Company
Form 10-K for Fiscal Year Ended December 31, 2012
Filed February 13, 2013
Form 10-Q for Fiscal Quarter Ended March 31, 2013
Filed April 25, 2013
File No. 001-13699
Reference is made to your letter dated July 12, 2013 providing further comments on our previous response letter dated May 23, 2013. We appreciate the Commission's assistance in ensuring our compliance with applicable disclosure requirements and enhancing the overall quality of our filings. Because of the sensitive nature of the information contained in the attachments to this letter, this submission is accompanied by a request for confidential treatment for such information, pursuant to Rule 83 of the Commission's Rules on Information and Requests, 17 C.F.R. Section 200.83. We are requesting confidential treatment for Attachments A, B, C and D to this letter, as indicated by [***], with respect to the Freedom of Information Act (the “FOIA”) and have filed a separate letter with the Office of Freedom of Information and Privacy Act Operations in connection with that request. Under Rule 83, each of Attachments A, B, C and D is marked “Confidential Treatment is requested by Raytheon Company pursuant to 17 C.F.R. 200.83.”
If any person (including any government employee who is not a member of the Staff) should request an opportunity to inspect or copy such documentation and information, we respectfully request that (i) you promptly notify our General Counsel of such request so that we may further substantiate this request for confidential treatment by writing to the General Counsel at the address noted above or by facsimile at (781) 522-6471, or by calling (781) 522-3000; (ii) you furnish the General Counsel with a copy of all written material pertaining to such request (including but not limited to the request itself and any determination by the SEC staff with respect to such request); and (iii) you give the General Counsel sufficient advance notice of any intended release so that we may, if we deem it necessary or appropriate, pursue any available remedies.
We have set forth below the comments in your letter, followed by our responses.
Comment 6:
Item 8. Financial Statements and Supplementary Data
Notes to Consolidated Financial Statements
Note 14: Pension and Other Employee Benefits
Defined Benefit Retirement Plan Summary Information, page 108
1. | You state that you consider a best estimate range of long-term future returns between the 25th and 75th percentile as “the narrowest range within which the actuary reasonably anticipates that the actual results, compounded over the measurement period, are more likely than not to fall,” with reference to Actuarial Standards of Practice 27. Please clarify for us how you evaluate the movement of your point estimate within the developed range over time. That is, please tell us specifically how you consider and weight such movement relative to changes in the other factors you identified in the bullet-point list on page 7 of your response when considering whether a change in your ROA assumption is appropriate. |
Confidential Treatment Requested by Raytheon Company
Raytheon Company
File No. 001-13699
July 26, 2013
Page 2 of 6
Response:
As part of our annual process for determining whether it is appropriate to change our long-term return on plan assets (ROA) assumption, we first review the existing long-term ROA assumption against the reasonable range of outcomes (i.e. long-term expected returns between the 25th and 75th percentile of statistically determined possible outcomes, as described in our previous response), reflecting the Company's most recent asset allocation assumption and our outside actuarial firm's (Mercer, a global consulting leader in talent, health, retirement and investments) capital market outlook that they use across their entire client base (which we review for reasonableness). We then supplement that analysis by reviewing our historical asset return trends and averages to identify any significant patterns that should be considered in evaluating our long-term ROA assumption.
The position of our 8.75% assumption within the developed range since that assumption was set in 2003 has been relatively consistent and we have not increased or decreased our 8.75% assumption, even as the range itself has moderately fluctuated both up and down over time. This approach is consistent with the objective of not reacting solely to near term market volatility in setting a long-term assumption. For example, while the top end of the range has moved from 8.8% in 2003 up to 10.65% at year-end 2010 and subsequently down to 9.41% by year end 2012, we have neither increased nor decreased our long-term ROA assumption over that period.
In addition to considering whether our ROA assumption remains within the range, we take into account other factors that are more judgmental and are used subjectively to validate whether a change is needed if our results remain within the reasonable range. Examples of factors that are taken into consideration include significant changes in investment strategy (e.g. if we adopt a Liability Driven Investment Strategy, or LDI), a recent history of consistently lower asset performance (e.g. 4-5 years of returns near the bottom of the range), etc.
2. | Please tell us in which year you first established 8.75% as the rate to be used for your domestic pension plans and provide us with the supporting calculations and documentation used in that year to establish the best estimate range both overall and by asset class. Please ensure your response and the documentation provided clearly demonstrate how the range of returns identified by asset class is translated into the overall range of returns used to evaluate the overall ROA assumption. |
Response:
We established our current expected long-term ROA assumption for our domestic pension plans of 8.75% in 2003, a change from our previous assumption of 9.5%. The long-term ROA assumption was changed because the previous assumption of 9.5% no longer fell within the reasonable range of expected outcomes over a long-term period (i.e. the 25th to 75th percentile range), as indicated in a copy of an unsigned letter sent to our former Chief Financial Officer (included as Attachment D). The letter states the range of results between the 25th and 75th percentile was 5.0% and 8.8%. This analysis was based upon an asset allocation profile of 53% domestic equity, 7% international equity, 30% fixed income, 2.5% private real estate, 2.5% private equity and 5% cash. After a thorough review of their respective files, neither the Company, our auditors, nor our outside actuaries were able to locate detailed data regarding the reasonable range of outcomes and the setting of the long-term ROA assumption other than a copy of the letter.
We believe, although we do not currently have any contemporaneous documentation, that the ROA was set at the high end of the range based upon the data available at the time regarding economic assumptions (which exclude alpha) and our strong historical asset performance. As a point of reference, the arithmetic and geometric historical returns from 1986 (the implementation date for ASC 715) through 2002, the time when our current long-term ROA established, were 9.2% and 8.6%, respectively.
3. | Please provide us the same supporting calculations using your building block method and documentation requested in comment 2 above to support the best estimate ranges on page 7 (overall) and page 9 (by asset class) of your response. |
Confidential Treatment Requested by Raytheon Company
Raytheon Company
File No. 001-13699
July 26, 2013
Page 3 of 6
Response:
Attachment A includes the supporting materials for the 25th to 75th percentile range results shown on page 7 (overall) and page 9 (by asset class) of our previous response. This support is automatically generated output from our outside actuary's system used to develop the ranges. A description of the approach taken to generate these results was provided on page 7 of our previous response.
4. | Please explain to us in greater detail what the table on the bottom of page 7 of your response conveys. |
Response:
The table represents the range of expected return outcomes on our domestic pension plan portfolio over a 20-year time horizon, based upon our outside actuarial firm's capital market outlook, which they utilize across their entire client base. For example, the results shown for the December 31, 2012 analysis in the table conveys that there is a 25% chance that the compounded annual rate of return over a 20 year time period for Raytheon's domestic pension assets would be 5.62% or lower (and correspondingly, there is a 75% chance that the compounded annual rate of return would be greater than 5.62%). Similarly, there is a 50% chance that the return over a 20 year time period could be
lower than 7.51%, and a 50% chance the returns could be higher. Finally, there is a 75% chance that the compounded annual returns would be 9.41% or lower, and a 25% chance that the returns would be higher.
5. | Please tell us the time period you use to analyze historical market returns for purposes of (a) developing ranges of expected returns overall and by asset class, (b) evaluating actual historical plan asset performance and (c) evaluating actual historical returns consistently above or below the selected ROA assumption. For each purpose, please tell us how you determined the most relevant period over which to consider historical market returns in light of the current interest rate and overall capital market environments. |
Response:
We employ a forward looking approach in developing the 25th to 75th percentile range. We do not use historical Raytheon market returns to establish this range. Instead, the range of expected returns shown on the table on page 7 of our previous response is based on the capital market outlook of our outside actuary applied to our investment allocation profile. They use these capital market assumptions, which are not specific to Raytheon, across their entire client base.
We do not have a single specific period for which we review our historical market returns. Instead, we look at the totality of the data to identify trends in our performance that would indicate that further evaluation of our assumption is needed. For example, if the last four years of returns were consistently 5% rather than the average return of 9.9% actually realized for that period, that would be an indication that our long-term ROA assumption of 8.75% may not be reasonable.
6. | Please quantify the changes in the actual pension asset allocation between the date at which you first established 8.75% as the ROA assumption to be used for your domestic pension plans and the current allocation. |
Response:
We first established our long-term ROA assumption in 2003. Over the period 2003 - 2012, we have maintained an investment strategy with a focus on maximizing portfolio growth at an appropriate level of risk. While we have made tactical adjustments to our actual asset allocations to react to near term market conditions and to optimize return, the basic allocation between equities and alternative investments vs. fixed income investments has been fairly consistent over this time period. For example, in 2003, our analysis (as described in our outside actuary's letter referenced in question 2, above) was based upon an expected allocation of 65% in equity and alternative assets, comprised of 53% US equities, 7% international equities, 2.5% private real estate, and 2.5% private equity, with the remaining 35% in fixed income and cash. By the end of 2012, the allocation used to evaluate our long-term ROA assumption was based upon an expected allocation of 65% in equity and alternative assets, comprised of 31% US equities, 16% international equities, 4% private equity, 3% private real estate, and 11% of other alternative assets, which include absolute return
Confidential Treatment Requested by Raytheon Company
Raytheon Company
File No. 001-13699
July 26, 2013
Page 4 of 6
funds. Our shift to more international and more alternative investments from 2003 to 2012 has generally offset the reduction in domestic equity.
7. | Please explain to us how you developed the “net alpha adjustment” of 0.5% for 2012 indicated in the portion of the table on the top of page 10 of your response. Also, explain to us how this was factored into the expected returns and 25th to 75th percentile ranges for private equity and absolute return funds shown in the table on page 9 of your response. |
Response:
The net alpha assumption of 0.5% was based on a gross alpha assumption of 0.74% offset by an assumed reduction for active management investment expenses of 0.21%. The alpha assumption was developed based on a review of the long-term history of active management in each asset class and reflects an estimate of the potential for successful active management based on the median historical results. The assumption was developed as a weighted-average of their expected alpha for each asset class in Raytheon's asset allocation. Alpha assumptions were already embedded into the expected returns for private equity and absolute return fund categories as a component of our outside actuarial firm's independent capital market outlook. Therefore, no additional alpha assumption was considered for these asset classes. Note that the table on page 9 of our previous response that shows the expected returns and percentile ranges for each asset class does not include any adjustment for alpha. The alpha adjustment is made at the portfolio level and is reflected in the table shown on page 7 of our previous response.
8. | Please explain to us whether the actual average annual ROA amounts of 9.08% and 8.9% indicated in the paragraph on the top of page 8 of your response represent arithmetic or geometric returns. If arithmetic, provide us with the geometric returns. Also, quantify for us the differences, if material, between the asset allocations underlying these historical returns and the asset allocations underlying your expected return for 2012. |
Response:
The actual average annual ROA amounts of 9.08% and 8.9% and all of the other historical averages described in our previous response represent the unadjusted arithmetic average return for the historical periods represented. Attachment B provides the historical asset performance as well as the arithmetic and geometric averages for each period presented in our previous response. Consistent with the application of the long-term ROA under ASC 715, where the long-term ROA is applied to a discrete annual period for purposes of determining pension expense, we look at the arithmetic average of our historical annual asset performance as a representation of the average return for an independent year.
As stated in our response to question 5 above, we look at the totality of the historical data to evaluate if there have been any notable trends in our performance. Once our long-term ROA has been determined to be within the 25th to 75th percentile range of results, we utilize the multiple data points of historical averages and patterns of returns to confirm reasonability of results compared to past history. As part of this review, we take into consideration potential outliers in the historical asset return data to ensure that our conclusions are not being influenced by unique economic events. For example, when we review our historical data over long-term periods (e.g. the 27 years since the implementation of ASC 715) and short-term periods (e.g. the past five years), we examined the actual average asset returns both with and without the 2008 market downturn since this was a unique event triggered by one of the most extreme financial crises in U.S. financial market history. Because the 2008 actual asset return of (25.1%) was approximately three standard deviations from our average asset return over the 27 year historical period, we believe it is useful to also examine and consider our data without the 2008 market downturn when reviewing historical performance. For reference, our historical average asset returns, excluding 2008, range from 9.7% over the past 5 years to 12.5% over the past 10 years on a geometric basis and 9.9% over the past 5 years to 12.6% over the past 10 years on an arithmetic basis.
We do not have actual asset allocation data for the entire 27 year period. Please see our response to question 6, describing changes in asset allocations from 2003 to 2012, which we believe answers the specific question you are asking.
Confidential Treatment Requested by Raytheon Company
Raytheon Company
File No. 001-13699
July 26, 2013
Page 5 of 6
9. | In connection with the comment above, please tell us whether the other average historical returns used in your various analyses are based on arithmetic or geometric averages. In addition, explain when and how you use arithmetic versus geometric returns in the various analyses supporting your ROA assumption. Last, provide us with the historical returns that you used in your various analyses. |
Response:
As described in our response to question 8 above, all of the historical average returns provided in our previous response are arithmetic averages. Attachment B provides the historical asset performance as well as the arithmetic and geometric averages for each period presented in our previous response, as well as averages excluding 2008 since this was driven by such a significant and unusual financial crisis.
For purposes of developing the 25th to 75th percentile ranges, our outside actuary uses both arithmetic and geometric calculations, which are solely based upon external capital market assumptions, not specific to Raytheon and then applies these calculations to our asset allocation profile. The process begins with determining arithmetic rates of return for each asset class using a building block approach that incorporates current external market conditions of interest rates, equity pricing, economic growth and inflation, overlaid with forward external looking projections of normal inflation, growth and interest rates to determine expected returns. Standard deviations for each asset class, along with correlations between various asset classes, are then developed by our outside actuarial firm, based on external historical data not specific to Raytheon.
From there, a forward looking long-term geometric rate of return is developed for the portfolio from these inputs, using standard statistical techniques. The output is then used to determine the percentile distribution of results for the investment portfolio.
10. | If the best estimate range were to move such that the current 8.75% expected return was no longer within the range, please explain to us how you would determine and select the ROA to be used. |
Response:
While we do not use a formulaic method (i.e. this is a judgment based assumption), we would select an ROA assumption, that falls within the reasonable range of outcomes (i.e. long-term expected returns between the 25th and 75th percentile), based on all data available to us including historical data, building blocks, forward looking results, expenses and recognition of alpha for active management . We would seek to select an ROA that would not need further adjustment for the foreseeable future, assuming no changes to the asset allocation, or significant changes to the capital market outlook.
This process ensures that the underlying concept of the long-term return assumption being established to “reflect the average rate of earnings expected on the funds invested or to be invested to provide for the benefits included in the projected benefit obligation,” as provided under ASC 715, is not compromised by putting undue weight to short-term market conditions and volatility that may not persist over the long-term.
11. | Please describe to us your expectation of expected ROA for 2014 and beyond. |
Response:
Our current long-term ROA of 8.75% was evaluated at year end 2012 and we believe it is reasonable for the reasons described in our responses. We will follow our standard annual process to evaluate the long-term ROA for purposes of determining 2014 expense and re-evaluate our long-term ROA assumption at year end 2013.
12. | Please tell us the time horizon over which you expect to provide pension benefits and how this was factored into your ROA determination. |
Confidential Treatment Requested by Raytheon Company
Raytheon Company
File No. 001-13699
July 26, 2013
Page 6 of 6
Response:
Our domestic pension plan covers current retirees, terminated employees entitled to future benefits and active employees hired prior to January 1, 2007. As illustrated in the graph provided in Attachment C, we expect to provide benefits to participants over the next 50 years into the future. Given the nature of our benefit obligation, we take a long-term, consistent approach to establishing our asset allocation strategy and corresponding ROA assumption, which is not intended to be reactive to near term economic volatility but instead aims to represents a reasonable return that could be expected to be achieved over the long-term life of the defined benefit pension plan.
**************
As requested in your original comment letter, we acknowledge the following:
• | The Company is responsible for the adequacy and accuracy of the disclosure in its filings; |
• | Staff comments or changes to disclosure in response to Staff comments do not foreclose the Commission from taking any action with respect to our 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. |
We believe that our responses adequately address the matters referenced in your letter dated July 12, 2013. If you would like to discuss our responses or any further questions you may have, please do not hesitate to contact me at 781-522-5833.
Respectfully,
/s/ Michael J. Wood | |
Michael J. Wood | |
Vice President, Controller and Chief Accounting Officer | |
Principal Accounting Officer | |
cc: William H. Swanson
David C. Wajsgras
Jay B. Stephens
Confidential Treatment Requested by Raytheon Company
Confidential Treatment is requested by Raytheon Company pursuant to 17 C.F.R. 200.83
Attachment A
***
Confidential Treatment is requested by Raytheon Company pursuant to 17 C.F.R. 200.83
Attachment B
***
Confidential Treatment is requested by Raytheon Company pursuant to 17 C.F.R. 200.83
Attachment C
***
Confidential Treatment is requested by Raytheon Company pursuant to 17 C.F.R. 200.83
Attachment D
***