October 11, 2012
Mr. Terence O'Brien
U.S. Securities and Exchange Commission
100 F Street, N.E.
Mailstop 4628
Washington, D.C. 20549
Re: | Legg Mason, Inc. |
Form 10-K for the Year Ended March 31, 2012 | |
Filed May 25, 2012 | |
Response dated August 1, 2012 | |
File No. 001-8529 |
Dear Mr. O'Brien:
The following responses to your comment letter dated August 8, 2012, were originally included in our response letter filed with the Securities and Exchange Commission on August 20, 2012, and have been revised, as follows below, to include certain information previously submitted with our request for confidential treatment pursuant to Exchange Act Rule 12b-4. All attachments referenced below continue to be subject to our request for confidential treatment.
Item 7. Management's Discussion and Analysis of Financial Condition and Results of Operations, page 23
Investment Performance, page 27
Comment 1:
We note your response to comment 2 in our letter dated July 18, 2012. While we understand providing performance information requested at the fund level would be cumbersome and would most likely not provide investors with useful information, it remains unclear why providing investors with the requested composite returns information for your investment strategies (i.e., large cap equity, small cap equity, et cetera), would not provide useful information. We note that several of your competitors provide the requested information in their periodic reports. As such, we continue to request that you disclose your investment strategies' composite returns, net of management fees, for the one, three and five year periods as of the end of your most recently completed fiscal year and/or quarter end, and from each investment strategy's inception, compared to their applicable benchmarks. If you continue to believe that this information is of no use to investors, we request that you expand upon your current disclosures to provide the following additional information:
• | a listing of your investment objectives by asset class in a table format; |
Page 2 of 17
• | the categories of investment vehicles sold by your principal distribution channels in a table format; |
• | the percentage of your AUM that have performed ahead of the benchmark comparison at the subset level of your three asset classes; |
• | the percentage of your AUM that have performed ahead of the corresponding peer group medians at the subset level of your three asset classes; |
• | rollforward and analysis of AUM by distribution channel for each period presented; and |
• | rollforward and analysis of AUM by client domicile for each period presented. |
Response:
Please note that we currently have approximately three hundred investment strategies that we manage for clients and we believe that including performance and comparative information for each strategy would be overwhelming for investors and would not provide them with meaningful disclosure for making investment decisions. However, we recognize the desire for us to disclose additional performance data, and we propose to disclose in future filings, similar to disclosures of other asset management firms, the absolute and relative performance compared to the applicable benchmark for a representative sample of funds within our AUM. This information would be representative of the significant investment strategies we use and will be net of management and other fees, for the one, three and five-year periods as of the end of each period presented, and from each fund's inception. As can be seen in the draft disclosure below, the representative sample of funds includes a total of 32 funds, and includes a representative sample of funds from each significant subclass of our investment strategies (i.e., large cap equity, small cap equity, etc.). The funds within this group are representative of the performance of significant investment strategies we offer that as of June 30, 2012, constituted approximately 61% of our AUM. The funds represent strategies that include a total of approximately $383 billion in assets under management as of that date. The only meaningful exclusion is our hedge funds of funds strategies, which involves privately placed hedge funds for which investment performance is not made publicly available consistent with current regulatory requirements (which represent only 3% of our assets under management as of June 30, 2012). Providing investment returns of funds provides a very good representation of our performance while avoiding the many complexities relating to factors such as multiple fee structures, bundled pricing and asset level break points that would arise in reporting for strategies or composites.
We also propose to supplement and expand existing disclosures of AUM performance with:
• | a listing of our investment objectives by asset class in a table format; |
• | the categories of investment vehicles sold by our principal distribution channels (i.e., our global distribution operations (which primarily market to retail investors), and our affiliate distribution operations (which primarily market to institutions)), in a table format; |
• | a rollforward and analysis of AUM by principal distribution channel for each period presented. |
We note that we have already disclosed and will continue to disclose:
• | the percentage of our AUM in each asset class subset that have performed ahead of the benchmark comparison at the subset level of our asset classes; |
• | the percentage of our mutual fund AUM that have performed ahead of the corresponding peer group medians at the subset level of our asset classes. |
Page 3 of 17
We do not believe that AUM activity by client domicile would provide investors with additional meaningful information for making investment decisions, beyond the existing and proposed disclosures, and therefore, we do not propose to include such information in our disclosures. We note that we already provide AUM activity disclosure in the categories in which our investors are primarily interested and the relative levels of AUM by client domicile have not changed meaningfully over the past several years. In the future, if the relative levels of AUM by client domicile change in a manner that would make such information meaningful to investors, we will provide appropriate disclosures in our filings.
As described above, beginning with our September 30, 2012 Form 10-Q filing and in all future filings, we will revise our existing disclosures of performance information to include comparative disclosures similar to the below information as of June 30, 2012. Please note that the performance statistics below represent our current views and may be revised based on further analysis.
Average Annual Total Return (%) vs. Benchmark | |||||||
Inception Date | Return Type | 1-YR | 3-YR | 5-YR | 10-YR | Inception | |
Equity | |||||||
Large Cap | |||||||
LM CBA Aggressive Growth Fund | 10/24/1983 | Absolute | 1.28 | 19.49 | 0.69 | 6.67 | 11.59 |
Russell 3000 Growth | Relative | (3.77) | 1.94 | (2.1) | 0.54 | 2.32 | |
LM CBA Appreciation Fund | 3/10/1970 | Absolute | 6.60 | 14.30 | 2.12 | 5.75 | 10.08 |
S&P 500 | Relative | 1.15 | (2.1) | 1.90 | 0.42 | 0.06 | |
LM CBA Equity Income Builder | 11/6/1992 | Absolute | 9.29 | 14.35 | 0.83 | 6.93 | 7.77 |
Russell 3000 Value | Relative | 6.65 | (1.58) | 2.93 | 1.56 | 1.00 | |
LM CBA Fundamental All Cap Value | 11/12/1981 | Absolute | (3.81) | 11.96 | (2.66) | 3.75 | 9.77 |
Russell 1000 Value | Relative | (6.45) | (3.97) | (0.56) | (1.62) | (1.71) | |
LM CBA Large Cap Growth Fund | 8/29/1997 | Absolute | 3.92 | 12.70 | 1.07 | 4.97 | 5.63 |
Russell 1000 Growth | Relative | (1.84) | (4.8) | (1.8) | (1.06) | 1.88 | |
LM CM Value Trust | 4/16/1982 | Absolute | 0.22 | 10.25 | (9.37) | 1.14 | 11.31 |
S&P 500 | Relative | (5.23) | (6.15) | (9.59) | (4.19) | (0.05) | |
Legg Mason ClearBridge Large Cap Value Fund | 12/31/1988 | Absolute | 5.41 | 15.17 | (0.36) | 5.59 | 9.63 |
S&P 500 Value | Relative | 2.41 | (0.15) | 2.32 | 0.8 | 0.26 | |
Legg Mason BW Diversified Large Cap Value Fund | 9/7/2010 | Absolute | 6.36 | — | — | — | 15.70 |
Russell 1000 Value | Relative | 3.35 | — | — | — | 3.39 | |
Small Cap | |||||||
Royce Pennsylvania Mutual | 6/30/1967 | Absolute | (6.97) | 15.84 | 0.49 | 7.97 | 11.66 |
Russell 2000 | Relative | (4.89) | (1.96) | (0.05) | 0.97 | — | |
Royce Premier Fund | 12/31/1991 | Absolute | (6.54) | 16.34 | 3.89 | 10.85 | 11.86 |
Russell 2000 | Relative | (4.46) | (1.46) | 3.35 | 3.85 | 3.12 | |
Royce Total Return Fund | 12/15/1993 | Absolute | (2.82) | 15.61 | 0.65 | 7.11 | 10.49 |
Russell 2000 | Relative | (0.74) | (2.19) | 0.11 | 0.11 | 2.77 | |
Royce Low-Priced Stock | 12/15/1993 | Absolute | (18.13) | 13.47 | 0.33 | 7.39 | 11.6 |
Russell 2000 | Relative | (16.05) | (4.33) | (0.21) | 0.39 | 3.88 | |
Royce Special Equity | 5/1/1998 | Absolute | 2.87 | 14.46 | 4.27 | 8.07 | 9.03 |
Russell 2000 | Relative | 4.95 | (3.34) | 3.73 | 1.07 | 3.61 |
Page 4 of 17
Average Annual Total Return (%) vs. Benchmark | |||||||||||||||||
Inception Date | Return Type | 1-YR | 3-YR | 5-YR | 10-YR | Inception | |||||||||||
Fixed Income | |||||||||||||||||
Global Fixed Income | |||||||||||||||||
LM BW Global Fixed Income | 9/30/2003 | Absolute | 7.81 | 11.45 | 8.88 | — | 7.53 | ||||||||||
Citi World Gov't Bond | Relative | 5.14 | 6.10 | 1.57 | — | 1.82 | |||||||||||
LM BW Global Opportunities Bond | 11/1/2006 | Absolute | 8.04 | 12.16 | 8.85 | — | 7.91 | ||||||||||
Citi World Gov't Bond | Relative | 5.36 | 6.81 | 1.54 | — | 1.82 | |||||||||||
Legg Mason Australian Bond Trust | 6/30/1983 | Absolute | 12.11 | 10.81 | 8.76 | 7.31 | 2.72 | ||||||||||
UBS Australian Composite Bond Index | Relative | (0.28 | ) | ) | 2.25 | 0.6 | 0.53 | (0.74 | ) | ) | |||||||
Legg Mason Core Plus Global Bond Trust | 2/28/1995 | Absolute | 13.11 | 13.2 | 7.44 | 6.85 | 7.84 | ||||||||||
Barclays Global Aggregate (AUD Hedged) | Relative | 1.52 | 3.21 | (2.12 | ) | ) | (1.42 | ) | ) | (0.93 | ) | ) | |||||
Western Asset Global Multi Strategy Fund | 8/31/2002 | Absolute | 4.34 | 10.93 | 7.46 | — | 7.58 | ||||||||||
50% Bar. Global Agg./ 5% Bar. HY 2%/25% JPM EMBI + | Relative | (1.63 | ) | ) | 0.53 | (0.59 | ) | ) | — | 0.64 | |||||||
Western Asset Global High Yield Bond Fund | 2/22/1995 | Absolute | 3.31 | 14.88 | 5.37 | 8.34 | 7.64 | ||||||||||
Barclays Global High Yield | Relative | (3.6 | ) | ) | (1.16 | ) | ) | (3.28 | ) | ) | (2.62 | ) | ) | (2.01 | ) | ) | |
U.S. Taxable Fixed Income | |||||||||||||||||
Western Asset Core Bond Fund | 9/4/1990 | Absolute | 7.19 | 12.58 | 6.92 | 6.20 | 7.68 | ||||||||||
Barclays US Aggregate | Relative | (0.28) | 5.65 | 0.13 | 0.57 | 0.63 | |||||||||||
Western Asset Core Plus Fund | 7/8/1998 | Absolute | 7.87 | 12.68 | 7.80 | 7.10 | 7.04 | ||||||||||
Barclays US Aggregate | Relative | 0.40 | 5.75 | 1.01 | 1.47 | 1.04 | |||||||||||
Western Asset Corporate Bond Fund | 11/6/1992 | Absolute | 8.24 | 13.49 | 5.55 | 5.43 | 6.85 | ||||||||||
Barclays US Credit | Relative | (1.3) | 3.40 | (2.04) | (1.11) | (2.88) | |||||||||||
Western Asset Emerging Markets Debt | 10/17/1996 | Absolute | 7.07 | 13.74 | 8.97 | 12.51 | 11.53 | ||||||||||
JPM EMBI Global | Relative | (3.83) | 0.27 | (0.47) | 0.77 | 1.25 | |||||||||||
Western Asset High Yield Fund | 9/28/2001 | Absolute | 4.37 | 16.36 | 6.44 | 8.81 | 7.95 | ||||||||||
Barclays US Corp High Yield | Relative | (2.9) | 0.08 | (2.01) | (1.35) | (1.54) | |||||||||||
Western Asset Inflation Index Plus Bond | 3/1/2001 | Absolute | 11.34 | 9.57 | 7.89 | 7.02 | 7.14 | ||||||||||
Barclays US TIPS | Relative | (0.32) | (0.06) | (0.55) | (0.21) | (0.22) | |||||||||||
Western Asset Intermediate Bond Fund | 7/1/1994 | Absolute | 5.97 | 8.67 | 6.69 | 5.99 | 6.58 | ||||||||||
Barclays Intermediate Gov't/Credit | Relative | 0.55 | 2.86 | 0.68 | 0.91 | 0.57 | |||||||||||
Western Asset Short Term Bond Fund | 11/11/1991 | Absolute | 1.13 | 6.12 | 1.93 | 2.42 | 4.00 | ||||||||||
Citi Treasury Gov't/Credit 1-5 YR | Relative | (1.35) | 2.49 | (2.79) | (1.75) | (2.08) | |||||||||||
Western Asset Total Return Unconstrained | 7/6/2006 | Absolute | 4.06 | 9.83 | 5.87 | — | 5.8 | ||||||||||
Barclays US Aggregate | Relative | (3.41) | 2.90 | (0.92) | — | (0.74) | |||||||||||
Western Asset Adjustable Rate Income | 6/22/1992 | Absolute | 0.79 | 5.57 | 0.3 | 1.2 | 3.00 | ||||||||||
Citi T-Bill 6-Month | Relative | 0.69 | 5.39 | (0.82) | (0.72) | (0.36) | |||||||||||
Western Asset Mortgage Defined Opportunity Fund Inc. | 2/24/2010 | Absolute | 7.94 | — | — | — | 13.69 | ||||||||||
Barclays Global Aggregate Corp. | Relative | 4.51 | — | — | — | 7.61 |
Critical Accounting Policies and Estimates, page 46
Goodwill, page 52
Comment 2:
We note your response to comment 8 in our letter dated July 18, 2012. To help us better understand your conclusion that the estimated fair value of your one reporting unit exceeds the carrying value by more than 51% as of December 31, 2011, considering your market capitalization was less than total stockholders' equity by 39.9% as of December 31, 2011, please address the following:
Page 5 of 17
• | We note your reference to the disclosures you provided in your fiscal year 2009 Form 10-K regarding the use of a control premium to bridge the gap between the estimated fair value of your reporting unit and your market capitalization. However, this disclosure was discontinued beginning with your fiscal year 2010 Form 10-K even though total stockholders' equity continued to exceed market capitalization. Please provide us with the disclosure you intend to include in your next periodic report to explain to investors how you estimated the reasonable control premium considered in your last goodwill impairment test, and confirm to us that you will continue to provide investors with an explanation for the control premium estimated for future goodwill impairment tests to the extent that total stockholders' equity exceeds market capitalization. Otherwise, please explain to us how you determined the estimate of a control premium is not a material estimate as of your most recent goodwill impairment testing date. |
• | Please provide us with the report you used to assess your estimated control premium for your most recent impairment test. Please also explain to us how you considered your specific facts and circumstances when comparing the information related to recent transactions. |
• | Please provide us with your reconciliation of the estimated fair value of your reporting unit as of your most recent goodwill impairment testing date and your market capitalization. In this regard, your disclosures in the fiscal year 2012 Form 10-K indicate the only difference is the control premium. However, your response notes that corporate costs and debt are excluded from the estimated fair value of the reporting unit and are also contributing to the difference. Further, please confirm that you will revise your disclosure in future filings to disclose the other reconciling factors between the estimated fair value of your reporting unit and market capitalization. Please provide us with the disclosure you intend to include in your next periodic report. |
• | As previously requested, please provide us with your detailed discounted cash flow calculations you prepared as of your most recent goodwill impairment testing date. It should be clear from the documents you provide to us that you are testing goodwill for impairment in the same manner as described in your disclosures. Please also help us understand the material assumptions you made in the discounted cash flow calculations and how you determined these assumptions were reasonable. |
Response:
Overview of Goodwill Analysis
Legg Mason's goodwill impairment testing requires that we estimate the implied fair value of the reporting unit and compare that value to the carrying amount for that unit. In estimating the implied fair value of our reporting unit for these purposes, we primarily utilize a discounted projected cash flow analysis. This analysis requires that we project the future cash flows from the reporting unit and apply a discount rate to estimate the present value of the projected future cash flows.
Page 1 of Attachment 1 contains the projected discounted cash flow analysis. In this analysis, cash flows from the reporting unit are projected for each of the next 40 years based on an assumed growth rate for the year. The starting point for the projected cash flows, to which the applicable growth rate is applied for the first year in the calculation, is the annualized, year-to-date net income of the reporting unit plus noncash depreciation and amortization. The projected
Page 6 of 17
cash flow for each year is discounted to present value using the assumed discount rate, and the discounted cash flows are added together. These calculations are shown on the bottom of page 1 of Attachment 1. On the top right hand side of page 1 of Attachment 1, the total discounted cash flows, plus a present value tax benefit which is discussed below, are compared to the carrying value of the reporting unit.
Consistent with standard valuation practices for taxable transactions, the projected discounted cash flow analysis also factors in a tax benefit value determined using a standard formula. This tax benefit represents the discounted tax savings a third-party that purchased Legg Mason on the valuation date would receive from future tax deductions for the amortization of the purchase price over 15 years.
The discounted projected cash flow analysis is based on significant assumptions, including the calculation of the discount rate used to estimate the net present value of the projected cash flows and growth rates and related assumptions underlying the projections of future cash flows. These assumptions are discussed in detail later in this response. In general, we utilized a discount rate based upon a weighted average cost of capital analysis. This analysis utilizes a number of inputs, which are discussed later in this response, to produce an estimate of the company's cost of capital (a weighted blend of the costs of equity and debt capital). This analysis resulted in a discount rate of 14%, which is within a reasonable range of discount rates observed for investment advisory companies based on data from Morningstar.
Additional details of the discount rate calculation can be seen on pages 2-5 of Attachment 1. Page 2 of Attachment 1 shows the actual calculation of the discount rate using the weighted average cost of capital analysis. Peer company betas, size and equity premiums are inputs to the calculation of this discount rate, and support for the inputs we used in our calculation is included on pages 3 and 4 of Attachment 1. Page 5 of Attachment 1 includes detail of discount rates for asset management companies based on Morningstar data, which provides a market check of, and support for, the discount rate that Legg Mason used in its calculation.
The significant assumptions underlying the projected cash flows are summarized later in this response. In general, we developed projected annual market growth rates for equity (6% through fiscal year 2014 and 7% thereafter), fixed income (3%) and projected annual organic growth rates (net client asset flows) for our business (2%). These asset growth rates were used to project future revenues.
Page 6 of Attachment 1 provides details of our cash flow projections for each of our significant asset managers using the general growth assumptions discussed above, and also based on an analysis of historical experience at the asset manager. The growth rates for the first five years were developed based on a year by year assessment that considers current market conditions, our experience, our internal financial projections, relevant publicly available statistics and projections and discussions with our internal experts. Page 7 of Attachment 1 of the attached includes documentation of our consideration of actual market performance statistics.
Page 7 of 17
Page 8 of Attachment 1 of the attached is the “market cap reconciliation” and shows the results of our discounted projected cash flow analysis for the reporting unit (after deducting corporate debt and costs) compared to a market-based valuation of our equity value (a guideline transaction valuation method), which applies an average of peer transaction EBITDA multiples (multiples of EBITDA paid as purchase price in change of control transactions for comparable companies) to our EBITDA. We further assess the accuracy of the results from these valuation methods by comparing their appropriately weighted average to our market capitalization to determine an implied control premium (i.e., the difference between the implied valuation and the market capitalization). If the implied control premium is within a reasonable range observed by the market, it supports the reasonableness of our analyses. These analyses are discussed more in detail later in this response.
Pages 9 and 10 of Attachment 1 document the asset manager market transactions we considered to develop a peer transaction EBITDA multiple in order to determine whether the implied control premium was reasonable.
Pages 11 to 21 of Attachment 1 include our analysis of relevant market transactions to assess the reasonableness of the control premium implied by the fair value assessments made in our reporting unit goodwill impairment tests and include the report we used to assess our estimated control premium for these purposes. We have considered all special circumstances regarding Legg Mason in determining the comparability of the market data. Aside from smaller items that address value amounts and are discussed later in this response (see e.g., the discussion of synergies and our model discussed later in this response), we do not believe that there are specific circumstances relating to Legg Mason that make the data not comparable. While we know no obvious circumstances relating to these transactions to cause us to exclude them from our consideration of control premium values, we have excluded the outlying 1/3 of transactions to mitigate the chance that any transaction with special circumstances might influence our conclusions. From time to time, we also review asset manager transaction premium values (page 21 of Attachment 1), and find their values do not vary meaningfully from control premiums in other industries, but the asset manager population is more limited, and therefore we use more current transaction data from other industries.
Significant Assumptions
Legg Mason's impairment testing employs cash flow and other analyses where significant assumptions are used to estimate the values third party market participants would pay in a taxable transaction to acquire the subject asset(s). The Discounted Cash Flow Method (“DCF”) of the Income Approach was the primary methodology utilized in deriving the fair value estimate of the Reporting Unit. Legg also utilized the guideline transaction (“GLT”) method of the Market Approach, specifically enterprise value to EBITDA multiples to substantiate the estimate of value computed by the Income Approach in our reconciliation of the reporting unit to our consolidated market capitalization. Reliance on these valuation methods is consistent with valuation practices. During annual testing, Legg Mason is required to evaluate and, if necessary, revise the assumptions employed. Our assumptions, valuation methodology and inputs were reviewed by a nationally recognized independent valuation expert. The more significant aspects
Page 8 of 17
of our consideration of these assumptions are summarized below. We utilize data input as of October 31 and compare these to December 31 and consider any necessary updates.
◦ | Discount Rate |
The discount rate utilized in deriving the fair value estimate was based on a Weighted Average Cost of Capital (“WACC”) model that was derived using the Capital Asset Pricing Model (“CAPM”) for development of a market-based cost of equity capital. This is an acceptable approach that is consistent with valuation practices. Overall, our discount rates are not significantly different from the prior year. Market inputs to our discount rate determination (i.e. debt rates, risk-free rates, peer equity betas, etc.) would result in lower discount rates, but because most of our subjective company/asset risk factors have been increased compared to the prior year, the discount rates did not change significantly. Also, our growth assumptions are generally decreased compared to the prior year.
See below for comments on specific inputs into the CAPM and WACC model:
• | Risk Free Rate: [***] |
• | Beta Calculation: [***] |
• | Equity Risk Premium: [***] |
• | Cost of Debt: [***] |
• | Capital Structure: [***] |
• | Company Size Premium: [***] |
The discount rate increased to 14.0% for the combined operating segment, Global Asset Management, from a slightly lower weighted average discount rate of 13.4% last year (comprised of 13.2% for Americas and 14.0% International).
Our determination of discount rate includes an adjustment added for equity risk for Legg Mason relative to other asset management peer companies. The equity risk factor (alpha factor) is further adjusted to consider the relative risk associated with each Legg Mason business component.
Our consideration of relative company/asset specific risk for the Global Asset Management reporting unit included the following items:
• | Evolving distribution capabilities compared to peers with affiliated brokerage operations, established direct marketing, etc. |
• | Although flows are improving somewhat, there continues to be flow risk at our largest affiliates, primarily Western and ClearBridge. |
• | Business streamlining initiative. |
______________________________________________
[***] Information has been redacted in accordance with Legg Mason's request for confidential treatment under separate cover.
Page 9 of 17
• | Permal, International and Cross Border funds as significant contributors to consolidated results. Our collective International business contributes over one-third of consolidated revenues, and Permal is the second-largest contributor to consolidated earnings and the most sensitive to AUM fluctuation of all our affiliates. |
Quantifying our alpha factor is not formulaic, but guided by relevant market observations (market capitalization, EBITDA multiples and other indicators of value), such that we add [***] to the cost of company-wide equity.
According to data from Morningstar, as published in its Cost of Capital Yearbook 2011, our conclusions are within a reasonable range of discount rates observed for companies in the SIC 6282, Investment Advisors, which ranged from 10.6% to 15.0%.
◦ | Market and Flow Growth by Product |
Growth assumptions were developed for equity, fixed-income and liquidity product classifications, with consideration of both market and flow characteristics by asset classification. Advisors with circumstances that differentiate their expected cash flows, such as WAM, Royce, and Permal were also considered separately. These growth assumptions agree to Legg Mason's multi-year plan, and reflect our review of market and competitor research, along with discussions with Legg Mason management and our key asset managers (see Attachment 1).
Consistent with the market growth assumptions used in the multi-year plan and 20-year projections, equity products are expected to experience market growth of 6% during fiscal 2013 and 2014 and 7% thereafter. Support for these growth rates for equity products can be found from observations of past market history that equity returns average 7-10%, per Bloomberg, Barclays, and NASDAQ. As of September 30, 2011, 80% of our marketed equity composites have outpaced their 10-year benchmark returns, further supporting the use of these historical trends in our equity growth assumptions. Also consistent with market growth assumptions used in the multi-year plan and 20-year projections, fixed income products are expected to grow 3%. We note our Western Asset fixed income products have 10-year performance clearly exceeding 5%.
We have reviewed long-term growth rates of various instruments, as published by Morningstar in its SBBI Valuation Edition 2011 Yearbook. This yearbook documents historical returns on Stocks, Bonds, Bills, and Inflation from 1926-2010. We note that returns vary by year but over the period 1926-2010, geometric average total returns ranged from 3.6% on Treasury Bills, 5.9% on Corporate Bonds, 9.9% on Large Company Stocks to 12.1% on Small Company Stocks.
______________________________________________
[***] Information has been redacted in accordance with Legg Mason's request for confidential treatment under separate cover.
Page 10 of 17
Except for specific flow assertions in years 1 and 2 for certain affiliates, AUM flows have been projected to be at 2% long term annual growth, which is reasonable and is supportable by census data and other industry data that is available. Also supporting this is the fact that, prior to 2007 Legg Mason affiliates exceeded 2% in annual organic growth. We note average annual equity flow of over 4% and 5%, respectively, per census data for the past 40 years and per Lipper for the past twelve years. Continued improved performance by our affiliates is expected to generate positive flows. Management has invested in and refocused on its distribution business to sell products which should have a positive impact on asset flows combined with the recent good performance. A departure from the 2% flow assumption included higher growth rates for [***] management forecast for [***] and [***] flows for fiscal years 2013 and 2014, respectively, as seen on page 6 of Attachment 1. We note that [***] history with Legg Mason includes years of significant organic growth ([***] and [***] in fiscal years 2007 and 2008, respectively). Further, if these [***] flow assumptions were reduced for fiscal year 2013 and 2014, they would not change our conclusions.
◦ | Cash flow growth rate for goodwill testing |
The cash flow growth rate for goodwill testing is calculated using a projected pre-tax profit margin. When our AUM was approximately $1 trillion (September 2007), the consolidated pre-tax profit margin was approximately 25%. Our current Global Asset Management operating segment/reporting unit pre-tax margin is approximately [***] (which excludes corporate overhead). In order to be competitive, we believe our pre-tax margins must be [***] or better and management will pursue initiatives in addition to our recently-completed business model streamlining initiative, if necessary, to ensure we return to a margin exceeding [***]. The pre-tax margin for the combined operating segment, Global Asset Management, was increased, beginning in year 2 through 20, from [***] to [***] and thereafter approaches [***].
The template for the calculation of our Global Asset Management Segment/Reporting Unit Goodwill impairment testing reflects the annualized [***] pre-tax margin on revenues adjusted for non-cash charges such as depreciation and amortization. The growth rates described for our equity and fixed income products are also reflected in the template.
We have considered analyst reports on Legg Mason during the course of 2011. The consensus estimate for long-term earnings growth embedded in those analyst reports, typically over a period of five years, is approximately 11.0%. Also, we note that we considered the long-term earnings growth estimates for our peer group. The long-term earnings growth for the peer group ranged from 8.0% to 13.5%.
◦ | Market capitalization implications |
Consistent with valuation practices, we performed a reconciliation analysis of the sum of our equity values derived from a weighted average of the results of our discounted cash flow (DCF) and guideline transactions (GLT) valuation approaches utilized to determine the implied value of the reporting unit, to our public market capitalization as a whole.
______________________________________________
[***] Information has been redacted in accordance with Legg Mason's request for confidential treatment under separate cover.
Page 11 of 17
In performing the reconciliation, we subtracted the net present value of corporate costs and interest bearing debt (net of cash) from the DCF approach since these items were not allocated to the Reporting Unit but are theoretically contemplated in the Company's market capitalization. Then, we calculated the weighted average result from the sum of the DCF (25% weight) and GLT (75% weight) implied values for the consolidated entity. Based on this calculation, the control premium implied when comparing market capitalization to the equity value at the time we were performing our tests derived from the weighting of the DCF and GLT valuation approaches (control value) was 41%. This figure was compared to research performed on control premiums seen in historical market transactions across a variety of industries and market transactions of asset managers. Through discussion with valuation consultants and review of various studies, we find that transaction values for asset management firms relative to their respective book values do not vary significantly from relative transaction values in other industries. Our analysis noted 16 recent transactions of firms engaged in the brokerage, investment and management consulting industry with observed control premiums ranging from 8 to 83% with an average premium of 55%. We have also examined transactions in the asset management industry. The control premiums indicated in the selected transactions indicated a range of 10 to 84% in past years. We believe that the nature of Legg Mason's affiliate business model and the related central corporate infrastructure negatively impacts the current market pricing of our stock relative to other asset management firms. In a control transaction, there would be synergies and significant reductions in our corporate center would result. Our control premium also reflects a revenue multiple of approximately 2.1% which is well below our peer group. As reported in the Morningstar report, the median for our SIC Composite is 3.33%. We have also examined transactions in the banking and finance industry. Because of the wide range of premiums observed in this population, we excluded outlying values. Based on all of the control premium research performed, the nature and structure of our company, the implied premium from our equity value conclusion over market capitalization appears to fall within available market evidence of control premiums and does not appear unreasonable.
◦ | Contributory Asset Charges |
Our asset management businesses generally do not require significant levels of cash, such that no related contributory asset charges are appropriate. While affiliates may make investments in seed capital, these investments generally return more than their cost, therefore, no contributory asset charges are included for these investments.
◦ | Budget reconciliations |
In developing our growth assumptions, we considered Legg Mason's multi-year plan approved in June 2011 and 20-year projections, market and competitor research, and discussions with Legg Mason management and our key asset managers. Growth assumptions, yields and margins by affiliate used for intangible testing agree with the variables used for budget purposes. Intangible testing projections have the benefit of updated AUM data and actuals through October 2011.
Page 12 of 17
◦ | Prior estimates compared to actual |
Annually we compare the prior year impairment testing growth assumptions to actuals through the current year assessment date. Despite variances in the current year due to some unfavorable flows and market circumstances, overall, our actual growth is generally consistent with our estimates.
We believe that performance is a leading indicator of flows, and we note that our Western Asset (“WAM”) and ClearBridge products with recent flow challenges demonstrate compelling performance trends. Weighted-average three-year returns for WAM and ClearBridge products are 9.6% and 14.4%, respectively.
◦ | Comparison of Inputs to December 31, 2011 |
December 2011 AUM activity, interest rates, market, and other relevant factors are relatively static compared to October 2011, such that updating the analyses to December 31, 2011 inputs is not necessary.
All significant assumptions were vetted with management, external valuation specialists, and our external auditors.
With respect to our disclosure referenced in your comment relating to the change in discounted projected cash flows of 51% for our reporting unit, that disclosure was calculated only for the reporting unit and goodwill in isolation. This disclosure was intended only to provide an indication of the magnitude of changes in the assumptions underlying the discounted cash flow analysis that would be required to trigger further analysis of a potential impairment of goodwill. However, as noted in other disclosures (page 51 of our Form 10-K for the fiscal year ended March 31, 2012), certain changes in assumptions in cash flows and discount rates could result in impairments to certain indefinite-life intangible assets. Many of the assumptions and growth rates used in the indefinite-life intangible analyses are also a component of the goodwill analysis and assumptions. Changes to our assumptions underlying the goodwill analysis would also be reflected in the impairment analyses for other intangible assets. Declines in cash flows would most likely result in impairments to indefinite-life assets prior to goodwill impairment. In such an event, the carrying value of our reporting unit would change as a result of the impairment of the other intangible asset(s) and thus the test to assess goodwill impairment would also be affected. Upon further consideration, we believe our goodwill related disclosure could be more meaningful with acknowledgment of the impact that changes in assumptions would have on indefinite-life intangible assets, including a reference to the more specific disclosures relating to the potential indefinite-life intangible asset impairment.
Please note that our fiscal year 2009 Form 10-K included more detailed disclosure about control premiums because we had experienced a material impairment of goodwill during that fiscal year. In future filings, we will expand our disclosure to include an explanation of the control premium estimated for goodwill impairment tests to the extent our total stockholders' equity exceeds our market capitalization. In addition, in future filings, we will also clearly disclose the fact that corporate debt and corporate overhead are not part of our reporting unit, and therefore, are also reconciling items in comparing the assessed value of our reporting unit to our consolidated
Page 13 of 17
market capitalization. We will also expand our disclosure of our use of a market valuation approach to corroborate the results of our discounted projected cash flow analysis of our reporting unit.
Beginning with our September 30, 2012 Form 10-Q filing, we will revise our existing Critical Accounting Policies and Estimates disclosure, the full text of which is noted below, with the updates in italics.
Goodwill
Goodwill is evaluated at the reporting unit level and is considered for impairment when the carrying amount of the reporting unit exceeds the implied fair value of the reporting unit. In estimating the implied fair value of the reporting unit, we use valuation techniques based on discounted projected cash flows, similar to techniques employed in analyzing the purchase price of an acquisition target. In December 2010, we announced a realignment of our executive management team, which during fiscal 2012, resulted in the combination of our Americas and International divisions into one operating segment, Global Asset Management. Internal management reporting has been modified consistent with this realignment such that discrete financial information regularly received by the chief operating decision maker, our Chief Executive Officer, is at the consolidated Global Asset Management business level. As a result, the former Americas and International operating segments are no longer our reporting units, and subsequently, goodwill is recorded and evaluated at one Global Asset Management reporting unit level. See Note 17 of Notes to Consolidated Financial Statements for additional information related to business segments. Our Global Asset Management reporting unit consists of the operating businesses of our asset management affiliates and our centralized global distribution operations and excludes corporate debt and our holding company corporate costs and overhead, including costs associated with executive management, finance, human resources, legal and compliance, internal audit and other central corporate functions.
Goodwill principally originated from the acquisitions of Citigroup Asset Management, Permal and Royce. The value of the reporting unit is based on projected net cash flows of assets managed in our mutual funds, closed-end funds and other proprietary funds, in addition to separate account assets of our managers.
Significant assumptions used in assessing the implied fair value of the reporting unit under the discounted cash flow method include the projected cash flows generated by the reporting unit, including profit margins, expected cash flow growth rates, and the discount rate used to determine the present value of the cash flows. Cash flow growth rates consider estimates of both AUM flows and market expectations by asset class (equity, fixed income and liquidity) and by investment manager based upon, among other things, historical experience and expectations of future market performance from internal and external sources. The impact of both net client flows and market performance on cash flows are projected for the near-term (generally the first five years) based on a year-by-year assessment that considers current market conditions,
Page 14 of 17
our experience, our internal financial projections, relevant publicly available statistics and projections, and discussions with our own market experts. Actual cash flows in any one period may vary from the projected cash flows without resulting in an impairment charge because a variance in any one period must be considered in conjunction with other assumptions that impact projected cash flows.
Discount rates are based on appropriately weighted estimated costs of debt and capital using a market participant perspective. We estimate the cost of debt based on published debt rates. We estimate the cost of capital based on the Capital Asset Pricing Model, which considers the risk-free interest rate, market risk and size premiums, peer-group betas and unsystematic risk. The discount rates are also calibrated based on an assessment of relevant market values. For our annual impairment test as of December 31, 2011, the projected cash flows are discounted at 14.0% to determine their present value.
To confirm the reasonableness of our discounted cash flow analysis, the results of this analysis for the reporting unit (after deducting corporate debt and costs) are compared to a market-based valuation of our equity value, which applies an average of peer transaction EBITDA multiples paid in change of control transactions for peer companies to our EBITDA. We further assess the accuracy of the results from these valuation methods by comparing their appropriately weighted average to our market capitalization to determine an implied control premium.
As of December 31, 2011, the implied fair value of the reporting unit significantly exceeds the carrying value. However, considering relevant prices of our common shares, our market capitalization, along with a reasonable control premium, exceeds the aggregate carrying value of our consolidated net assets by a small margin.
For discounted projected cash flow analysis, projected cash flows, on an aggregate basis across all asset classes, are assumed to have an average annual growth rate of approximately 8%. Cash flow growth is based on separate factors for equity, fixed income, and liquidity products. Equity product growth projections are based on long-term growth experience and current market conditions. Fixed income product growth projections are based on the past experience of our primary fixed income manager and current market influences relevant to their business, available historical experience and market statistics, and estimates of future expectations. We believe our growth assumptions are reasonable given our consideration of multiple inputs, including internal and external sources described above, although our assumptions are subject to change based on fluctuations in our actual results and market conditions. However, decreases in cash flow from actual results or projections would also impact the impairment testing for other intangible assets. As noted above, if our cash flows over the long-term deviate greater than 5%, certain of our indefinite-life intangibles could be impaired by a material amount. We would likely have material impairments to our other indefinite-life intangibles prior to a goodwill impairment.
Page 15 of 17
As noted above, our market capitalization, along with a reasonable control premium, exceeds the aggregate carrying value of our consolidated net assets by a small margin. In calculating our market capitalization for these purposes, market volatility can have a significant impact on our capitalization, and if appropriate, we may consider the average market prices of our stock to determine market capitalization, but never for more than several months within the test date. The control premium arises from the fact that in an acquisition, there is typically a premium paid over current market prices of publicly traded companies that relates to the ability to control the operations of an acquired company. Recent market evidence regarding this control factor suggests premiums of 23% to 74% as realistic and common, and Legg Mason believes such premiums to be a reasonable estimation for our equity value. Our market evidence is from a published source for the year ended December 31, 2011 and includes 10 transactions in the brokerage, investment management and consulting industry group and 18 transactions from the banking and finance industry group. Based on our analysis and consideration, we believe our implied control premium is reasonable. We exclude consideration of transactions with unique circumstances and find that transaction values for asset management firms relative to their respective book values do not vary significantly from relative transaction values in other industries.
Under separate cover, with our request for both confidential treatment and return to us upon completion of your review, pursuant to Exchange Act Rule 12b-4, are additional specifics of our response relating to control premium, related objective market data, and our December 31, 2011 calculations of estimated cash flows of our Global Asset Management business reporting unit.
9. Commitments and Contingencies, page 81
Comment 3:
We note your response to comment 14 in our letter dated July 18, 2012, and the disclosures you provided in Note 8 to your first quarter of fiscal year 2013 Form 10-Q. Specifically, we note your statements: “The ultimate resolution of other matters…cannot be currently determined. In the opinion of management and after consultation with legal counsel, due to the preliminary nature of these matters, Legg Mason is currently unable to estimate the amount or range of reasonably possible losses from these matters…” There is a concern that your reference to not being able to determine the ultimate resolution of the other matters is the sole basis for your inability to provide the amount or range of reasonably possible loss in excess of accrual. While we understand that there are uncertainties associated with estimating the amount or range of reasonably possible loss in excess of accrual, ASC 450-20-50 does not require the estimate to be precise. As such, we continue to request that you revise your disclosure in future filings to disclose an estimate of the reasonably possible loss or range of loss, or, if true, state that the estimate is immaterial in lieu of providing quantified amounts. If you continue to conclude that you cannot estimate the reasonably possible additional loss or range of loss, please ensure that it is clear this inability is not due to the lack of knowing the ultimate outcome. Also, please supplementally (1) explain to us the procedures you undertake on a quarterly basis to attempt to develop a range of reasonably possible loss for disclosure and (2) for each material matter, what
Page 16 of 17
specific factors are causing the inability to estimate and when you expect those factors to be alleviated. We recognize that there are a number of uncertainties and potential outcomes associated with loss contingencies. Nonetheless, an effort should be made to develop estimates for purposes of disclosure, including determining which of the potential outcomes are reasonably possible and what the reasonably possible range of losses would be for those reasonably possible outcomes. Please note that you may provide your quantified disclosures on an aggregated basis. Please include your proposed disclosures in your response.
Response:
We continue to conclude that we cannot estimate the reasonably possible additional loss or range of loss associated with the other matters referenced in your comment. The inability to provide a reasonably possible amount or range of losses is not because there is uncertainty as to the ultimate outcome of a matter, but because liability and damage issues have not developed to the point where we can conclude that there is both a reasonable possibility of a loss and a meaningful amount or range of possible losses. There are numerous aspects to these customer complaints, legal actions, inquiries, proceedings and investigations that prevent Legg Mason from estimating a related amount or range of reasonably possible losses. These aspects include, among other things, the very nature of the matters; that significant relevant facts are not known, are uncertain or are in dispute; and that damages are not specified, are uncertain, unsupportable or unexplained. In addition, for legal actions, discovery may not yet have started, may not be complete or may not be conclusive and meaningful settlement discussions may not have occurred. Further, for regulatory matters, investigations may run their course without any clear indication of wrongdoing or fault until their conclusion.
In management's opinion, an adequate accrual has been made as of June 30, 2012, to provide for any reasonably possible losses that may arise from matters for which the company could reasonably estimate an amount.
In future filings, we will include such explanations in our disclosures relating to matters where a related amount or range of reasonably possible losses cannot be estimated.
Under separate cover, with our request for both confidential treatment and return to us upon completion of your review, pursuant to Exchange Act Rule 12b-4, are additional specifics of our response relating to the procedures we undertake on a quarterly basis to attempt to develop a range of reasonably possible loss for disclosure and the specific factors relating to significant matters that are causing the inability to estimate and when those factors may be alleviated.
In connection with our response to your comment letter, we acknowledge that:
• | We are responsible for the adequacy and accuracy of the disclosure in the filing; |
• | Staff comments or changes to disclosure in response to staff comments do not foreclose the Commission from taking any action with respect to the filing; and |
Page 17 of 17
• | We 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. |
If you have any additional questions or would like any additional clarification, please contact Brian Eakes (410-454-2965) or me (410-454-2935).
Sincerely,
/s/ Peter H. Nachtwey
Peter H. Nachtwey
Chief Financial Officer
Attachments (provided under separate cover with our request for confidential treatment and return to us upon completion of your review pursuant to Exchange Act Rule 12b-4)
Unredacted Version of Revised Comment Letter Response
1 - Comment 2, Projected Discounted Cash Flow Analysis, Assumptions and Control Premium Assessment Support
2 - Comment 3, Procedures and Specific Factors
cc: | Tracey Smith, U.S. Securities and Exchange Commission, Division of Corporation Finance |