2011 Duff Phelps Risk Premium Report EXCERPT

Risk Premium Report 2011
Selected Pages and Examples

This document is an excerpt of the 2011 Risk Premium Report, and includes an overview of the methodologies employed in performing the analysis required for the Size Study, Risk Study, and High-Financial-Risk Study that constitute the Duff & Phelps Risk Premium Report. The excerpt also includes a limited number of examples demonstrating how the Risk Premium Reports size premia and risk premia data can be used to estimate cost of equity capital (more examples are available in the complete Report). The excerpt does not include the size and risk premia data exhibits that are available in the full version of the Risk Premium Report.
Inside
2
Purchasing Information
3
Table of Contents (2011 Report)
5
Introduction
7
How the 2011 Report is Organized
9
Portfolio Methodology
14
Using the 2011 Report
18
The Size Study
39
The Risk Study
55
High-Financial-Risk Study
67
New in 2011: The Risk Premium Calculator (web-based)
Publication information/Disclaimer/ Purchasing information

2011 Duff & Phelps Risk Premium Report The information and data presented in the Duff & Phelps Risk Premium Report and the online Duff & Phelps Risk Premium Calculator has been obtained with the greatest of care from sources believed to be reliable, but is not guaranteed to be complete, accurate or timely. Duff & Phelps, LLC expressly disclaims any liability, including incidental or consequential damages, arising from the use of the Duff & Phelps Risk Premium Report and/or the online Duff & Phelps Risk Premium Calculator or any errors or omissions that may be contained in either the Duff & Phelps Risk Premium Report or the online Duff & Phelps Risk Premium Calculator. Copyright 2011 Duff & Phelps, LLC. All Rights Reserved. No part of this publication may be reproduced or used in any other form or by any other meansgraphic, electronic, or mechanical, including photocopying, recording, taping, or information storage and retrieval systemswithout Duff & Phelps prior, written permission. To obtain permission, please write to: Risk Premium Report, Duff & Phelps, 311 S. Wacker Dr., Suite 4200, Chicago, IL 60606. Your request should specify the data or other information you wish to use and the manner in which you wish to use it. In addition, you will need to include copies of any charts, tables, and/or figures that you have created based on that information. There is a $1500 processing fee per request. There may be additional fees depending on your proposed usage. Published by: Duff & Phelps, LLC 311 South Wacker Drive Suite 4200 Chicago, IL 60606 (312) 697-4600 www.duffandphelps.com Additional copies of the 2011 Duff & Phelps Risk Premium Report may be obtained from our Distributors: Business Valuation Resources (BVR)* www.bvresources.com/dp 1-(888)-287-8258 Morningstar global.morningstar.com/riskpremiareports 1-(888)-298-3647 ValuSource www.valusource.com 1-(800)-825-8763 * The Duff & Phelps Risk Premium Report is intended to be used as a companion publication to the web-based Duff & Phelps Risk Premium Calculator. Exclusive distribution for the Calculator is through BVR.
Duff & Phelps
| 2
Table of Contents
NOTE: The table of contents shown here is the table of contents of the full 2011 Risk Premium Report, and is included so that readers can view the contents of the full Report. It is not the table of contents for this document.
Acknowledgements Introduction
History of the Duff & Phelps Risk Premium Report Recent Changes and Additions Who Should Use the Duff & Phelps Risk Premium Report Appropriate Use of the Duff & Phelps Risk Premium Report
The Size Study 6

6 6 7 7 What is Size? Reasons for Using Alternative Measures of Size The Difference between the A Exhibits and the B Exhibits The Difference Between Risk Premia Over the Risk Free Rate and Risk Premia Over CAPM Risk Premium Over Risk Free Rate, RPm+s Risk Premium Over CAPM (Size Premium), RPs Overview of Methods Used to Estimate Cost of Equity Capital Using the Size Study Estimating Cost of Equity Capital Using the Buildup 1 Method The Basic Buildup Equation The Buildup 1 Equation Example 1a: Buildup 1 Method (using guideline portfolios) Example 1b: Buildup 1 Method (using regression equations) Unlevered Cost of Equity Capital Overview of the Current Methodology and Assumptions Used to Unlever Risk Premia in the 2011 Risk Premium Report Unlevered Risk PremiaReconciliation of A, B and C Exhibits Relevering Estimating Cost of Equity Capital Using theBuildup 1-Unlevered Method Example 2a: Buildup 1 Method-Unlevered (using guideline portfolios) Example 2b: Buildup 1-Unlevered Method (using regression equations) Estimating Cost of Equity Capital Using the CAPM Method Example 3a: CAPM Method (using guideline portfolios) Example 3b: CAPM Method (using regression equations) 17 18 18 Estimating Cost of Equity Capital Using the Buildup 2 Method Example 4a: Buildup 2 Method (using guideline portfolios) Example 4b: Buildup 2 Method (using regression equations)
19
20 20 22 23 23 24 26 29 29 30 31 32 37 37 39 39 40 40 44 48 49 52 56 57 58

First Section: Methodology Second Section: Data Exhibits
8
8 9
Data Sources Historical Time Period Used Exclusions Unseasoned Companies High-Financial-Risk Study Exclusions are Based on Past Information Portfolio Creation Size Study Portfolio Creation Risk Study Portfolio Creation Correcting for Delisting Bias Size and Risk Rankings are Based on Past Information
10
10 10 10 11 11 11 12 12 13 14 14

Using Smoothed Premia versus Using Average Premia Using the Regression Equation Method to Calculate Interpolated Risk Premia Between Guideline Portfolios Guideline Portfolio Method or Regression Equation Method? Using the Regression Equation Method to Calculate Interpolated Risk Premia for Smaller Companies Size Study or Risk Study?
15
15 16
Duff & Phelps
| 3
Table of Contents
The Risk Study

Size and Risk Reasons for Using Fundamental Measures of Risk in Addition to Measures of Size Presentation of the Results Overview of Methods Used to Estimate Cost of Equity Capital using the Risk Study Gathering Accounting Information to Calculate Fundamental Risk Measures Estimating Cost of Equity Capital Using the Buildup 3 Method Risk Premia Over Risk Free Rate, RPm+u The Buildup 3 Equation Example 5a: Buildup 3 Method (using guideline portfolios) Example 5b: Buildup 3 Method (using regression equations) Unlevered Cost of Equity Capital Estimating Cost of Equity Capital Using the Buildup 3-Unlevered Method Example 6: Buildup 3 Method-Unlevered (using guideline portfolios) The C Exhibits Provide a Link Between the Size Study and the Risk Study Is Size Correlated with Market and Fundamental Risk Measures? The C Exhibits and Company-Specific Risk Presentation of the Results
60
60 61 62 63 64 65 65 66 67 70 73 73 74 77 77 79 79
The High-Financial-Risk Study

The High-Financial-Risk H Exhibits Altman z-Score Non-Public Companies and z-Score Measurement of Historical Risk Premiums Presentation of the Results Overview of Methods Used to Estimate Cost of Equity Capital Using the High-Financial-Risk Study Example 7: Estimating Cost of Equity Capital Using the Buildup 1-High-Financial-Risk Method Example 8: Estimating Cost of Equity Capital Using the CAPM-High-Financial-Risk Method Additional information on Company-Specific Risk
80
81 82 83 84 84 85 86 89 92
Appendix A Appendix B Glossary Data Exhibits
94 101 103 105
Duff & Phelps
| 4
Introduction
History of the Duff & Phelps Risk Premium Report In 1990, Roger Grabowski began closely studying the relationship between company size and stock returns.1 Grabowskis early research focused on size as measured by market capitalization, but quickly advanced to two additional areas of inquiry: whether stock returns were predicted by measures of size other than market capitalization, and whether stock returns were predicted by fundamental risk measures based on accounting data. To investigate these questions, in 1992 Grabowski, working with a colleague 2, contracted with the Center for Research in Security Prices (CRSP) at the University of Chicago to build a database that combined stock prices, number of shares, and dividend data from the CRSP database with accounting and other data from the Standard & Poors Compustat database. What they found was that as size decreases, or risk increases (as measured by fundamental accounting data), returns tend to increase (and vice versa). Thereafter, they published a series of articles reporting their findings, culminating with a seminal 1996 article and a subsequent article in 1999 which together served as the foundation of the Duff & Phelps Risk Premium Report.3 Now in its 16th year of publication, the Duff & Phelps Risk Premium Report continues to be at the forefront in providing comprehensive valuation methodology and data. The 2011 Duff & Phelps Risk Premium Report (2011 Report) includes data available through December 31, 2010, and should be used for calendar year 2011 valuations. Roger J. Grabowski, ASA, Author Managing Director, Duff & Phelps James P. Harrington, Editor Director, Duff & Phelps
Recent Changes and Additions In 2011 we are pleased to introduce the new web-based Duff & Phelps Risk Premium Calculator.4 The Risk Premium Calculator is based on the same trusted data and analysis published in the Risk Premium Report since 1996. Calculator Features y Anytime, anywhere access at www.bvmarketdata.com/DP.RPC y Complete historical database of risk premia and size premia data (19962011) y Automatic output y Executive Summary of COE estimates, including CAPM, Buildup, and unlevered COE y Microsoft Excel output of all underlying values and calculations y Easy to use / Saves time For more information about the new Risk Premium Calculator, please see New in 2011: The Duff & Phelps Risk Premium Calculator (web-based) on page 67.
Roger Grabowski, ASA, is a managing director in the Duff & Phelps Chicago office and part of the firms Valuation Advisory Service practice, and co-author with Dr. Shannon Pratt of Cost of Capital: Applications and Examples, 4th Edition (John Wiley & Sons, 2010). David King, CFA, is National Technical Director of Valuation Services at Mesirow Financial Consulting, LLC. The research began when both he and Roger Grabowski were at Price Waterhouse, predecessor firm to PricewaterhouseCoopers. Roger J. Grabowski and David King, New Evidence on Size Effects and Equity Returns, Business Valuation Review (September 1996, revised March 2000), & Roger J. Grabowski and David King, New Evidence on Equity Returns and Company Risk, Business Valuation Review (September 1999, revised March 2000). Please note that in 2011, the Duff & Phelps Risk Premium Calculator is available exclusively through Business Valuation Resources (BVR). For more information about the Calculator, visit: www.bvresources.com/dp or call toll free 1-(888)-287-8258.
Duff & Phelps
| 5
Introduction
Who Should Use the Duff & Phelps Risk Premium Report The Risk Premium Report is designed to assist financial professionals in estimating the cost of equity capital (cost of equity, or COE) for a subject company. The risk premia calculated in the Report can be used to develop COE estimates using both the build-up method and the Capital Asset Pricing Model (CAPM). In addition to the traditional professional valuation practitioner, the Risk Premium Report, and the accompanying web-based Risk Premium Calculator (Calculator), are designed to serve the needs of: y Corporate finance officers for pricing or evaluating mergers and acquisitions, raising private or public equity, property taxation, and stakeholder disputes. y Investment bankers for pricing public offerings, mergers and acquisitions, and private equity financing. y CPAs who deal with either valuation for financial reporting or client valuations issues. y Judges and attorneys who deal with valuation issues in mergers and acquisitions, shareholder and partner disputes, damage cases, solvency cases, bankruptcy reorganizations, property taxes, rate setting, transfer pricing, and financial reporting.
Appropriate Use of the Duff & Phelps Risk Premium Report The information and data in the Risk Premium Report (and in the online Risk Premium Calculator)5 is primarily designed to be used to develop cost of equity capital (COE) estimates for large majority of companies that are fundamentally healthy, and for which a going concern assumption is appropriate. High-financial-risk (i.e. distressed) companies are excluded from the base dataset and analyzed separately. Because financial services companies are excluded from the base set of companies used to develop the analyses presented in the Report, the Report (and the online Calculator) should not be used to estimate cost of equity for financial services companies. Financial services companies include those companies in finance, insurance, or real estate (i.e. companies with an SIC Code that begins with 6).
The Duff & Phelps Risk Premium Calculator is available exclusively through Business Valuation Resources (BVR). For more information about the Calculator, visit: www.bvresources.com/dp or call toll free 1-(888)-287-8258.
Duff & Phelps
| 6

The Risk Premium Report is divided into two main sections: a methodology section, followed by a data exhibits section. First Section: Methodology The first section features a discussion of the data and methodology used to create the portfolios used to perform the analysis in the Report, as well as an overview of the Size Study, Risk Study, and High-Financial-Risk Study (with examples of how to use each of these studies to estimate cost of equity capital). Also included are Appendices and a Glossary of terms: y Portfolio Methodology: A discussion of the data and methodology used to create the portfolios used to perform the analysis in the Report. y Size Study: Analyzes the relationship between equity returns and company size, using up to eight measures of company size (i.e. size measures). y Risk Study: Analyzes the relationship between equity returns and accounting-based fundamental risk measures. y High-Financial-Risk Study: Analyzes the relationship between equity returns and high-financial-risk, as measured by the Altman z-Score. y Appendices: Definitions of Compustat data items, and a summary of changes from previous versions of the Report (over time). y Glossary: A list of important terms with accompanying definitions.
Duff & Phelps
| 7

Second Section: Data Exhibits The second section features the data exhibits in which the various risk and size premia used to estimate cost of equity capital are found. Each of the three Studies (Size Study, Risk Study, and High-FinancialRisk Study) discussed in the Methodology section have corresponding data exhibits (A, B, D, or H), as illustrated in Figure 1. Figure 1: Size Study, Risk Study, High-Financial-Risk Study and Corresponding Exhibits
Size Study Risk Study High-FinancialRisk Study
For example, the C exhibits can help to answer whether the subject company is more or less profitable (as measured by operating margin) than similarly-sized companies, or whether the subject companys earnings are more or less volatile (as measured by coefficient of variation of operating margin and coefficient of variation of ROE) than similarly-sized companies. In the former case, the less profitable the subject company is, all other things held the same, the riskier it is (and vice versa). In the latter two cases (which are measures of earnings volatility), the more volatile a companys earnings are, all other things held the same, the less predictable they are, and thus the riskier the company is (and vice versa). This is an important capability because this type of analysis can be used as an indication that a company-specific risk adjustment is needed (either upward or downward). Figure 2: The C Exhibits Size Study and Risk Study Link
A B
Size Study
Risk Study
The risk premia in the A, B, D, and H exhibits can be used to develop cost of equity capital estimates using both the buildup method and the capital asset pricing model (CAPM). In addition, a link between the Size Study and Risk Study is provided in the C exhibits. y Exhibits A-1 through A-8: The A exhibits provide risk premia over the risk free rate in terms of the total effect of market risk and size risk for 25 portfolios ranked by eight alternative measures of size (RPm+s ). y Exhibits B-1 through B-8: The B exhibits provide risk premia over CAPM (size premia) in terms of size risk for 25 portfolios ranked by eight alternative measures of size (RPs ). y Exhibits C-1 through C-8: The C exhibits provide a link between the Size Study and the Risk Study. These exhibits can be used to compare a subject companys fundamental risk characteristics to the fundamental risk characteristics of portfolios made up of similarly-sized companies.
C
y Exhibits D-1, D-2, and D-3: The D exhibits provide risk premia over the risk free rate in terms of the total effect of market risk and company-specific risk for 25 portfolios ranked by three alternative measures of fundamental risk (RPm+u ). y Exhibits H-A, H-B, and H-C: The H exhibits provide high-financialrisk premia for portfolios ranked by Altman z-Score 6. These premia may be used in both buildup and CAPM estimates of cost of equity capital if the individual analyst has determined that the subject company is high-financial-risk.7 Exhibit H-A is the high-financialrisk equivalent of the A exhibits, Exhibit H-B is the high-financial-risk equivalent of the D exhibits, and Exhibit H-C is the high-financialrisk equivalent of the C exhibits.
6 7
Altman z-Score is an accounting-data-based method designed to assess financial condition and developed originally for assessing the likelihood of bankruptcy. The decision to apply a high-financial-risk premium is ultimately dependent on the analysts professional judgment, based upon the analysts detailed knowledge of the subject company.
Duff & Phelps
| 8
Data Sources The universe of companies used to perform the analyses presented in the Risk Premium Report is comprised of those companies that are found in both the Center for Research in Security Prices (CRSP) database at the University of Chicago Booth School of Business and Standard and Poors Compustat database. Historical Time Period Used In the 2011 Risk Premium Report, risk premia and other useful statistics are developed using historical equity returns (from CRSP), and fundamental accounting data (from Compustat ) over the period 1963 through 2010. The Compustat database was established in 1963. While Compustats fundamental accounting data is available for some companies going back to the 1950s, this earlier data consists only of the back histories for companies that were added to Compustat in 1963 or later. The Reports analysis begins with 1963 data in order to avoid the obvious selection bias that would result from using the earlier data. For each year covered in the Report, financial data for the fiscal year ending no later than September of the previous year is considered. For example, when assigning a company to a portfolio to calculate returns for calendar year 1995, financial data through the latest fiscal year ending September 1994 or earlier is considered (depending on when the companys fiscal year ended).
Exclusions After identifying a universe of companies that are in both the CRSP and Compustat databases, the following types of firms are excluded: y American Depository Receipts (ADRs) y Non-operating holding companies y Financial service companies (SIC code 6) Financial service companies (those companies in finance, insurance, or real estate) are excluded because some of the financial data used in the Report is difficult to apply to companies in the financial sector (for instance, sales at a commercial bank). In addition, financial service companies tend to support a much higher ratio of debt to equity than do other industries, and so including them in with non-financial firms may an apples to oranges comparison that could lead to improperly skewed results. Moreover, companies in the financial services sector were poorly represented during the early years of the Compustat database. It should be noted that since financial service companies are excluded from the set of companies used to perform the analyses presented in the Report, these results should not be used by an analyst estimating the cost of equity capital (COE) for a financial services company. Altogether, companies are excluded (or segregated) based upon their past financial performance or trading history in the Risk Premium Report. It should be noted that alternative analyses in which no companies were excluded or segregated on the basis of past financial performance or trading history have been performed (that is, using all available non-financial companies). The results are similar, but these exclusions are maintained as a precaution against the possibility of introducing a bias in favor of the size effect (to the extent that such companies tend to have low market values).
Duff & Phelps
| 9
Unseasoned Companies The small cap universe may consist of a disproportionate number of start-up companies and recent initial public offerings. These unseasoned companies may be inherently riskier than companies with a track record of viable performance. For this reason (for each year since 1963), we screen the universe of companies to exclude companies with any of the following characteristics8: y Companies lacking 5 years of publicly traded price history y Companies with sales below $1 million in any of the previous five fiscal years y Companies with a negative 5-year-average EBITDA (earnings before interest, taxes, depreciation and amortization) for the previous five fiscal years y Companies not listed on one of the major US stock exchanges (NYSE, AMEX or NASDAQ) The set of companies remaining after this screen are seasoned companies in that they have been traded for several years, have been selling at least a minimal quantity of product, and have been able to achieve a degree of positive cash flow from operations. High-Financial-Risk Study After eliminating companies with the characteristics described previously, the remaining companies are screened again to exclude companies with any of the following characteristics9: y Companies that Standard & Poors has identified in the Compustat database as in bankruptcy or in liquidation, y Companies with a 5-year average net income available to common equity less than zero for the previous five years (either in absolute terms or as a percentage of the book value of common equity), y Companies with 5-year-average operating income (sales minus cost of goods sold plus selling, general and administrative expenses plus depreciation) less than zero for the previous five years (either in absolute terms or as a percentage of net sales),
y Companies with negative book value of equity at any one of the companys previous five fiscal year-ends, and y Companies with a debt-to-total capital ratio of more than 80%, (debt is measured in book value terms, and total capital is measured as book value of debt plus market value of equity). The companies excluded in this screen are set aside and analyzed separately in the High-Financial-Risk Study. This screen is performed in an effort to isolate the effects of highfinancial-risk. Otherwise, the results might be biased for smaller companies to the extent that highly leveraged and financially distressed companies tend to have both high returns and low market values. It is possible to imagine companies that dont have any of these characteristics, but could still be classified as high-financial-risk (i.e. distressed), and it is also possible to imagine companies which do have one or more of these characteristics but are not distressed. Nevertheless, the resulting high-financial-risk database is composed largely of companies whose financial condition is significantly inferior to the average, financially healthy public company. Exclusions are Based on Past Information The exclusion of companies is based on their past financial performance or trading history as of the time that the portfolios are formed for any given year over the 19632010 time horizon. For example, to form portfolios for 1963, company data for the previous 5 fiscal years (prior to September 1962) is considered. This procedure is repeated for each year from 1963 through the latest available year for each of the eight measures of size examined in the Size Study, and for each of the three measures of fundamental risk examined in the Risk Study. All of the previously discussed exclusions are therefore not based on any unusual foresight on the part of hypothetical investors in these portfolios, but are based on information that was already history at the time the portfolios were created.
8 9
The number of companies eliminated in this screen varies from year to year. The number of companies eliminated in this screen varies from year to year. These companies represented up to 25% of the data set in recent years, but less than 5% in 1963. Certain technical changes in methodology have resulted in a greater number of companies falling into the high-financial-risk database than in versions of this study published prior to 2000.
Duff & Phelps
| 10
Portfolio Creation After excluding unseasoned and segregating high-financial-risk companies, the result is a base set of companies that is used for the analyses performed in both the Size Study and the Risk Study. The major difference between the two studies is that the portfolios presented in the Size Study are ranked by eight alternative measures of size, from largest (Portfolio 1) to smallest (Portfolio 25), while the portfolios presented in the Risk Study are ranked by three accountingbased measures of fundamental risk, from lowest risk (Portfolio 1) to highest risk (Portfolio 25). The smallest size/highest risk portfolios tend to have the highest returns. Other than that difference, portfolio formation in the Size Study and Risk Study is a very straightforward process. This process is described in the following sections.
Size Study Portfolio Creation To perform the analysis required for the Size Study, 25 portfolios are created from companies that are similarly-sized, with Portfolio 1 made up of the largest companies and Portfolio 25 made up of the smallest companies. The equity returns for each of the 25 portfolios returns are calculated using an equal-weighted average of the companies in the portfolio, and these returns are then used to calculate risk premia (and other useful information and statistics) for each. Size is defined by the traditional size measure, market value of common equity (i.e. market capitalization), as well as seven additional size measures: 1) Market value of common equity 2) Book value of common equity 3) 5-year average net income 4) Market value of invested capital (MVIC) 5) Total Assets 6) 5-year average EBITDA10 7) Sales 8) Number of employees The first step is to determine portfolio breakpoints for the 25 portfolios. Portfolio breakpoints are the upper and lower boundaries of each portfolio, represented by the largest and smallest New York Stock Exchange (NYSE) company, respectively, in each of the 25 portfolios. For example, to determine the breakpoints for the 25 portfolios ranked by Total Assets, all of the companies in the base set that are traded on the NYSE are ranked from largest (in total assets) to smallest (in total assets), and then divided into 25 equally populated portfolios.
10
Earnings before interest, income taxes, depreciation and amortization.
Duff & Phelps
| 11
Once portfolio breakpoints are determined, companies from the NYSE Amex Equities (formerly the American Stock Exchange, or AMEX)11 universe and the NASDAQ universe are added to the appropriate portfolio, depending on their size with respect to the breakpoints.12 Since NYSE Amex Equities and NASDAQ companies are generally small relative to NYSE companies, their addition to the data set produces portfolios that are more heavily populated at the small cap end of the spectrum.13 All portfolios are rebalanced annually, so this process is completed for each year from 1963 to the most recent available year, and for each of the eight measures of size. This results in the creation of 25 portfolios for each of the eight size measures, a total of 200 (8 x 25) unique portfolios for each year from 1963 to present, each ranked from largest to smallest by each respective size measure.14 Risk Study Portfolio Creation To perform the analysis required for the Risk Study, 25 portfolios are created from companies that have similar accounting-databased fundamental risk characteristics, with Portfolio 1 made up of companies with the lowest fundamental risk, and Portfolio 25 made up of companies with the highest fundamental risk. The equity returns for each of the 25 portfolios returns are calculated using an equal-weighted average of the companies in the portfolio, and these returns are then used to calculate risk premia (and other useful information and statistics) for each.
Fundamental Risk is defined by the following three alternative measures (the first is a measure of profitability; the latter two are measures of earnings variability): 1) Operating margin 2) Coefficient of variation in operating margin 3) Coefficient of variation in return on equity As in the Size Study, the first step is to determine portfolio breakpoints for the 25 portfolios. Using Operating Margin as an example, all companies in the base set that are traded on the New York Stock Exchange (NYSE) are ranked from lowest fundamental risk (highest operating margin) to highest fundamental risk (lowest operating margin), and then divided into 25 equally populated portfolios. Once portfolio breakpoints are determined, companies from the NYSE Amex Equities universe and the NASDAQ universe are added to the appropriate portfolio, depending on their fundamental risk with respect to the breakpoints. Since all portfolios are rebalanced annually, this process is followed for each year from 1963 to the most recent available year, for each of the three measures of fundamental risk. This results in the creation of 25 portfolios for each of the three fundamental risk measures, a total of 75 (3 x 25) unique portfolios for each year from 1963 to present, each ranked from lowest risk to highest risk for each respective measure of fundamental risk.15
11
12 13
14 15
On October 1 2008, NYSE Euronext acquired the American Stock Exchange (AMEX). Post merger, the AMEX equities business was branded NYSE Alternext US. NYSE Alternext US was subsequently re-branded NYSE Amex Equities, which remains its name today. NYSE Amex Equities data is available after 1962 and NASDAQ data is available after 1972. Some readers may ask why NYSE breakpoints are used rather than ranking the entire NYSE/NYSE Amex/NASDAQ universe. The consistent use of NYSE breakpoints avoids an apples-to-oranges mixing of pre-1972 (pre-NASDAQ) ranking criteria with post-1972 ranking criteria. Otherwise, average NASDAQ companies (in recent years) would be assigned to portfolios that contain much larger average NYSE companies (in earlier years) when calculating average returns for the mid-sized portfolios over the full sample period. The only logical alternatives are either to adopt the NYSE breakpoint approach or to exclude NASDAQ companies altogether. In the 2011 Report, this represents 8 size measures x 25 portfolios x 48 years (19632010) = 9,600 unique portfolio formations to perform the analysis presented in the Size Study. In the 2011 Report, this represents 3 measures of fundamental risk x 25 portfolios x 48 years (19632010) = 3,600 unique portfolio formations to perform the analysis presented in the Risk Study.
Duff & Phelps
| 12
Correcting for Delisting Bias Previous evidence indicated that the CRSP database omits delisting returns for a large number of companies for the month in which a company is delisted from an exchange.16 Data was collected for a large number of companies that had been delisted for performance reasons (e.g. bankruptcy, or insufficient capital) and found that investors incurred an average loss of about 30% after delisting. While CRSP has improved their database by reducing the number of companies for which it omits delisting returns, we incorporate this evidence into our rate of return calculations by applying a 30% loss in the month of delisting in all cases where the delisting return is missing and for which CRSP identified the reason for delisting as performance related. As an additional precaution, this adjustment is also applied in all cases in which the reason for delisting was identified by CRSP as unknown. 17 Size and Risk Rankings are Based on Past Information The ranking of companies based on size and fundamental risk does not imply any unusual foresight on the part of hypothetical investors in these portfolios the data used is as of the beginning of each year, and thus was already history at the time the portfolios are formed.
16 17
The Delisting Bias in CRSP Data, Tyler Shumway, Journal of Finance (March 1997). This approach is consistent with updates that we have published since 1998. More recent evidence suggests that the average delisting loss is less than Shumways original estimate. For more information about CRSP and CRSP delisting returns, visit www.CRSP.com.
Duff & Phelps
| 13
Using Smoothed Premia versus Using Average Premia The difference between average risk premia and smoothed risk premia is illustrated in Graph 1a and Graph 1b. Graph 1a: Average Risk Premia for 25 Portfolios with a Best Fit Line Added
14% 12% 10% 8% 6% 4% 2% 1.0 1.5 2.0 2.5 3.0 Log of Average Size Measure
A scatter plot of risk premia smoothed in this fashion and the log of the size measures will necessarily fall on the best fit line (smoothed risk premia are represented by the blue diamonds in Graph 1b). Graph 1b: Smoothed Risk Premia
3.5
4.0
4.5
5.0
14% 12% 10% 8% 6% 4% 2% 1.0 1.5 2.0 2.5 3.0 Log of Average Size Measure
Risk Premium
Risk Premium
3.5
4.0
4.5
5.0
In Graph 1a, the square gray points represent a scatter plot of size (on the horizontal x axis), and the average risk premium (for each of 25 size-ranked portfolios, on the vertical y axis).18 Note that as size increases from left to right, the risk premium tends to decrease (and vice versa). The best fit line is the straight (smooth) line in Graph 1a. Using regression analysis, an equation for the best fit line can be calculated, and this equation can be used to estimate smoothed risk premia for the 25 portfolios based upon the average size measure of each portfolio.
18
In this example, risk premium is used generically. The same statistical techniques described in this example are used to calculate smoothed risk premia over the risk free rate (the A exhibits) and risk premia over CAPM (the B exhibits), as well as smoothed unlevered premia (the C Exhibits).
Duff & Phelps
| 14
Smoothing the premia essentially averages out the somewhat scattered nature of the raw average premia. The smoothed average risk premium is generally the most appropriate indicator for most of the portfolio groups. It should be noted, however, that at the largestsize and smallest-size ends of the range, the average historical risk premiums may tend to jump off of the smoothed line, particularly for the portfolios ranked by size measures that incorporate market capitalization (exhibits A-1 and A-4). Because the size measure is expressed in logarithms, this is equivalent to the change in risk premium given the percentage change in the size of the companies from portfolio to portfolio. Smoothed risk premia are found in the data exhibits. For example, in Figure 3 the smoothed risk premium over the risk free rate for Portfolio 24 in Exhibit A-2 is 11.16 percent.19 In this example, the 11.16 percent risk premium is calculated based upon the average book value of equity of companies in Portfolio 24 ($166 million). However, the subject companys size rarely exactly matches the average size of companies in the guideline portfolio. In the next section, how to interpolate an exact risk premium value when the subject companys size is in between guideline portfolios is explained.
Using the Regression Equation Method to Calculate Interpolated Risk Premia Between Guideline Portfolios The Risk Premium Report provides two ways for users to match their subject companys size (or risk) characteristics with the appropriate smoothed premia: the guideline portfolio method, and the regression equation method. When the subject companys size (or risk) does not exactly match the average company size (or risk) of the guideline portfolio, the regression equation method is a straightforward and easy way to interpolate between the guideline portfolios.
Figure 3: Smoothed Premia in Exhibit A-2 Companies Ranked by Book Value of Equity Historical Equity Risk Premium: Average Since 1963 Data for Year Ending December 31, 2010
Portfolio Rank by Size 1 2 3 Average Book Val. ($mils.) 39,141 12,811 8,823 Log of Average Book Val. 4.59 4.11 3.95 Number as of 2010 37 31 31 Beta (SumBeta) Since 63 0.81 0.85 0.92 Standard Deviation of Returns 15.98% 16.44% 16.60% Geometric Average Return 10.51% 10.68% 11.82% Arithmetic Average Return 11.98% 12.21% 13.42% Arithmetic Average Risk Premium 5.08% 5.30% 6.52% Smoothed Average Risk Premium 4.42% 5.80% 6.26% Average Debt/ MVIC 23.89% 28.38% 29.11%
///
24 25 166 60 2.22 1.78 110 374 1.27 1.27 25.04% 26.09% 14.79% 15.41% 18.05% 18.89% 11.15% 11.98% 11.16% 12.41% 23.61% 24.19%
19
The A exhibits include risk premia over the risk free rate which are added to a risk free rate to estimate cost of equity capital using the buildup method. Please refer to the individual examples provided for these models for more information and examples.
Duff & Phelps
| 15
For example, if the subject company book value of equity in the previous example was $114 million, one would expect the smoothed average size premium to fall somewhere between 11.16 percent (the smoothed size premium for guideline Portfolio 24) and 12.41 percent (the smoothed size premium for guideline Portfolio 25). To calculate the exact smoothed premium between guideline portfolios, use the regression equations provided in each of the exhibits (please note that there is a different equation for each of the exhibits). For example, in Figure 4 the regression equation provided for Exhibit A-2 is 20: Smoothed Premium = 17.475% 2.843% x log (Book Value) Inserting the subject companys market cap of $114 million into this equation results in an exact smoothed premium of 11.63%: Smoothed Premium = 17.475% 2.843% x log ($114 million) = 11.63% = 17.475% 2.843% x 2.06
Guideline Portfolio Method or Regression Equation Method? The major difference between the guideline portfolio and the regression equation methods is that with the guideline method, one accepts the smoothed average risk premium published in the report (calculated using the average size in each of the 25 guideline portfolios), while with the regression equation method, one can calculate an exact interpolated value between the guideline portfolios. For this reason, although the guideline portfolio is simpler and more direct, the more flexible regression equation method is the suggested method in most cases. In practice this approach generally produces results that are very similar to those of the guideline portfolio approach presented above (unless one is extrapolating to a company that is much smaller than the average size for the 25th portfolio).
Figure 4: Location of Regression Method Equation in the Data Exhibits

Companies Ranked by Book Value of Equity Historical Equity Risk Premium: Average Since 1963 Data for Year Ending December 31, 2010
Equity Risk Premium Study: Data through December 31, 2010 Data Smoothing with Regression Analysis Dependent Variable: Average Premium Independent Variable: Log of Average Book Value of Equity Arithmetic Average Risk Premium 5.08% 5.30% 6.52% 5.87% 6.31% 6.99% 7.13% 7.51% 7.91% 8.15% 8.31% 9.52% 9.80% 8.23% 10.38% 11.70% 9.24% 10.23% 8.21% 9.09% 10.96% 9.98% 10.32% 11.15% 11.98% 4.40% 9.63%
Exhibit A-2
Portfolio Average Log of Number Beta Standard Geometric Arithmetic Rank Book Val. Average as of (SumBeta) Deviation Average Average by Size ($mils.) Book Val. 2010 Since '63 of Returns Return Return 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 39,141 12,811 8,823 6,398 4,540 3,433 2,868 2,323 1,969 1,704 1,497 1,304 1,155 1,059 939 812 710 635 548 464 380 303 230 166 60 4.59 4.11 3.95 3.81 3.66 3.54 3.46 3.37 3.29 3.23 3.18 3.12 3.06 3.02 2.97 2.91 2.85 2.80 2.74 2.67 2.58 2.48 2.36 2.22 1.78 37 31 31 31 37 30 30 39 32 34 34 41 38 38 45 38 47 44 55 59 87 71 86 110 374 0.81 0.85 0.92 0.91 1.02 1.02 1.05 1.07 1.12 1.06 1.08 1.09 1.10 1.14 1.11 1.22 1.22 1.19 1.24 1.21 1.22 1.25 1.26 1.27 1.27 15.98% 16.44% 16.60% 16.58% 18.59% 18.45% 20.33% 18.97% 20.68% 19.27% 20.21% 20.59% 21.58% 20.85% 21.22% 24.33% 23.69% 22.14% 21.36% 22.02% 21.24% 23.17% 23.26% 25.04% 26.09% 10.51% 10.68% 11.82% 11.25% 11.26% 12.01% 11.85% 12.42% 12.50% 13.07% 13.01% 14.17% 14.23% 12.86% 14.91% 15.66% 13.30% 14.57% 12.66% 13.47% 15.43% 14.01% 14.43% 14.79% 15.41% 9.84% 13.72% 6.88% 11.98% 12.21% 13.42% 12.78% 13.21% 13.89% 14.04% 14.41% 14.82% 15.06% 15.22% 16.42% 16.70% 15.13% 17.29% 18.61% 16.15% 17.14% 15.12% 15.99% 17.86% 16.88% 17.22% 18.05% 18.89% 11.30% 16.53% 6.90%
Smoothed Average Average Risk Debt/ Premium MVIC 4.42% 5.80% 6.26% 6.66% 7.08% 7.42% 7.65% 7.91% 8.11% 8.29% 8.45% 8.62% 8.77% 8.88% 9.02% 9.20% 9.37% 9.51% 9.69% 9.89% 10.14% 10.42% 10.76% 11.16% 12.41%
23.89% 28.38% 29.11% 28.50% 26.89% 26.37% 24.98% 24.85% 25.15% 25.56% 26.77% 25.69% 25.53% 23.85% 23.90% 24.16% 23.28% 23.97% 23.84% 23.57% 23.47% 23.98% 24.14% 23.61% 24.19%
Constant Std Err of Y Est R Squared No. of Observations Degrees of Freedom X Coefcient(s) Std Err of Coef. t-Statistic
Regression Output:
17.475% 0.876% 81% 25 23
-2.843% 0.285% -9.98
Smoothed Premium = 17.475% - 2.843% * Log(Book Value)
Large Stocks (Ibbotson SBBI data) Small Stocks (Ibbotson SBBI data) Long-Term Treasury Income (Ibbotson SBBI data)
Duff & Phelps, LLC 201103 CRSP, Center for Research in Security Prices. University of Chicago Booth School of Business used with permission. All rights reserved. www.crsp.chicagogsb.edu
Smoothed Premium = 17.475% 2.843% *Log(Book Value)

20
The term log is the base 10 logarithm. The base 10 log of 114 is 2.06. To calculate a base 10 log in Microsoft Excel, use =log(size measure). Remember that the logarithmic relationship is base-10, and that the financial size data is in millions of dollars, such that the log of $10 million is log (10), and not log (10,000,000).
Duff & Phelps
| 16
Using the Regression Equation Method to Calculate Interpolated Risk Premia for Smaller Companies. Sometimes one needs to estimate the cost of equity capital for a company that is significantly smaller than the average company size of even the smallest of the Reports 25 portfolios. In such cases, it may be appropriate to extrapolate the risk premium to smaller sizes using the regression equation method. Table 1 summarizes the size of companies by each of the eight alternative size measures, by percentile ranking.21 For example, the 95th percentile of size for book value of equity is $123.477 million, which means that 95 percent of the companies in Portfolio 25 have book value of equity that is less than $123.477 million (alternatively, this means that 5 percent of the companies in Portfolio 25 have book value of equity that is greater than $123.477 million). Or, looking now to the 5th percentile, 5 percent of the companies in Portfolio 25 have book value of equity that is less than $9.563 million (alternatively, this means that 95 percent of the companies in Portfolio 25 have book value of equity that is greater than $9.563 million). As a general rule, extrapolating a statistical relationship far beyond the range of the data used in the statistical analysis is not recommended. However, extrapolations for companies with size characteristics that are within the range of companies comprising the 25th portfolio are within reason. In some cases the size of the subject company may be equal to or greater than the smallest size of the companies included in the 25th portfolio for one size measure (e.g., sales), but less than the smallest size of the companies included in the 25th portfolio for another size measure (e.g., 5-year average income). In such cases analysts may consider including the size measure for sales, but excluding the size measure for 5-year average net income. One should never use those size measures for which the subject companys size is equal to zero or negative.
Table 1: Size Measures of Companies that Comprise Portfolio 25, by Percentile

Market Value of Equity 5th Percentile 25th Percentile 50th Percentile 75th Percentile 95th Percentile $8.060 26.447 57.626 104.870 152.923 Book Value of Equity $9.563 26.244 56.114 91.835 123.477 5-year Average EBITDA $1.829 5.649 12.388 21.391 31.393 5-year Average Income $0.495 1.683 3.654 6.165 8.247 Market Value of Invested Capital $10.372 37.277 84.184 141.129 220.326
Total Assets 5th Percentile 25th Percentile 50th Percentile 75th Percentile 95th Percentile $16.732 48.191 110.834 171.989 252.663
Sales $17.345 49.578 105.445 176.928 241.513
Number of Employees 13 112 245 379 524
Size Study or Risk Study? Analysts should use the Size Study if it has been determined that the risks of the subject company are comparable to the average of the portfolio companies of comparable size (e.g., comparable operating margin). One can determine the relative risk characteristics by looking at Exhibits C-1 through C-8. But if the risk characteristics of the subject company are significantly greater or less than the portfolios comprised of companies of similar size, the Risk Study helps the analyst determine how much greater or less than the average expected returns should he conclude for the subject company than indicated in the Exhibits A (for build-up) or Exhibits B (for CAPM).
21
The information in Table 1 was published as Exhibit E in previous reports.
Duff & Phelps
| 17
The Size Study
The Size Study analyzes the relationship between stock returns and company size. In addition to presenting risk premia and size premia for 25 size-ranked portfolios using the traditional market capitalization measure, the Size Study also considers 7 other measures of company size, including book value of equity, 5-year average net income, market value of invested capital (MVIC), total assets, 5-year average EBITDA, sales, and number of employees.22 As demonstrated in Graph 2, the data shows a clear inverse relationship between size and historical rates of return, regardless of how size is measured.
In Graph 2, as size decreases (from left to right), the average annual return over the study time horizon (19632010) tends to increase for each of the eight size measures. For example, in the 2011 Report, the average annual return of the portfolios made up of the largest companies (Portfolio 1 for each of the eight size measures) was 12.4 percent, while the average annual return of the portfolios made up of the smallest companies (Portfolio 25 for each of the eight size measures) was 21.1 percent.
Graph 2: Average Annual Return, 8 Alternative Measures of Company Size 19632010 Average Annual Return
25% Market Value of Equity 5-year Average Net Income Total Assets Sales 20% Book Value of Equity Market Value of Invested Capital (MVIC) 5-year Average EBITDA Number of Employees 15% Average (all size measures)
10%
5% 1 5 10 Portfolio (1 = Largest, 25 = Smallest) 15 20 25
22
For a detailed discussion of portfolio creation methodology, see Portfolio Methodology on page 10.
Duff & Phelps
| 18
The Size Study
What is Size? The size of a company is one of the most important risk elements to consider when developing cost of equity estimates for use in valuing a firm. Traditionally, researchers have used market value of equity (i.e. market capitalization, or simply market cap) as a measure of size in conducting historical rate of return research. For instance, this is the basis of the small stock return series published in the Stocks, Bonds, Bills and Inflation (SBBI) Valuation Yearbook.23 The Size Study measures the relationship between equity returns and up to eight alternative measures of size (including market capitalization). As illustrated in Graph 2, as size decreases, average annual return tends to increase for each of the eight size measures. Moreover, the size effect is not just evident for the smallest companies, but is evident for all but the largest groups of companies, including companies with a market capitalization in excess of several billions of dollars.
Reasons for Using Alternative Measures of Size There are various reasons for using alternative measures of size (in addition to market value of equity). First, financial literature indicates a bias may be introduced when ranking companies by market value.24 Second, a companys market capitalization may be affected by characteristics of the company other than size. In other words, some companies might be small because they are risky (high discount rate), rather than risky because they are small (low market capitalization). One simple example could be a company with a large asset base, but a small market capitalization as a result of high leverage or depressed earnings. Another example could be a company with large sales or operating income, but a small market capitalization due to being highly leveraged. Market capitalization may be an imperfect measure of the risk of a companys operations. In addition, using alternative measures of size may have the practical benefit of removing the need to make a guesstimate of size for comparative purposes. Fundamental accounting measures (such as assets or net income) are generally readily available, while market capitalization, at least for a closely held firm, is not. This is related to another reason one might consider using alternative measures of size: potential circularity issues. When you are valuing a closely held company, you are trying to determine market capitalization. If you need to make a guesstimate of the subject companys market capitalization first in order to know which size premium to use, a circularity problem may be introduced.25 Finally, when doing analysis of any kind it is generally prudent to approach things from multiple directions if at all possible. This is good practice for several reasons, with the most important being that it has the potential of strengthening the conclusions of the analysis. For instance, we intuitively place more faith in a poll of 1,000 people than a poll of 10 people.26
23 24 25
26
Stocks, Bonds, Bills and Inflation Valuation Yearbook, Morningstar (formerly Ibbotson Associates) (2011). A Critique of Size Related Anomalies, Jonathan Berk, Review of Financial Studies, vol. 8, no. 2 (1995). For a complete discussion of the history of the size premium and criticisms of the size premium, see chapter thirteen in Cost of Capital: Applications and Examples 3rd ed. by Shannon Pratt and Roger Grabowski, Wiley (2008); or chapter fourteen in Cost of Capital: Applications and Examples 4th ed. by Shannon Pratt and Roger Grabowski, Wiley (2010). The same logic can be extended to sources of valuation data. Both the SBBI Yearbook and the Duff & Phelps Risk Premium Report are excellent sources of valuation data, and both are recommended for use in developing discount rates.
Duff & Phelps
| 19
The Size Study
While the A and B exhibits present different types of risk premia, both the A and B exhibits 25 portfolios are ranked by the same eight alternative measures of size, which are described in Table 2.27
Table 3: Statistics Reported for 25 size-ranked portfolios in the Size Studys A and B Exhibits
Exhibits A-1 through A-8 Exhibits B-1 through B-8 Average of the sorting criteria (e.g., average number of employees) for the latest year used in determining the size of the companies (i.e., the size criteria when the latest years portfolios are formed). For example, the market value in exhibit B-1 is the market value of equity at the beginning of the latest year. The other size criteria are based on what was known at the beginning of the latest year when the portfolios are formed. Beta estimate calculated using the sum beta method applied to monthly returns for 1963 through the latest year (see the 2011 SBBI Valuation Yearbook pp. 77-78 for a description of the sum beta method). Arithmetic average historical equity return since 1963.
Table 2: Eight Alternative Measures of Size

Exhibits A-1 and B-1 Market value of common equity (common stock price times number of common shares outstanding). Exhibits A-2 and B-2 Book value of common equity (does not add back the deferred tax balance) Exhibits A-5 and B-5 Total Assets (as reported on the balance sheet). Exhibits A-6 and B-6 5-year average earnings before interest, income taxes, depreciation and amortization (EBITDA) for the previous five fiscal years (operating income before depreciation plus non-operating income). Exhibits A-7 and B-7 Sales (net).
Average of the sorting criteria (e.g., average number of employees) for the latest year used in determining the size of the companies (i.e., the size criteria when the latest years portfolios are formed). For example, the market value in exhibit A-1 is the market value of equity at the beginning of the latest year. The other size criteria are based on what was known at the beginning of the latest year when the portfolios are formed. The number of companies in each portfolio at the beginning of the latest year.
Exhibits A-3 and B-3 5-year average net income for previous five fiscal years (net income before extraordinary items). Exhibits A-4 and B-4 Market value of invested capital (MVIC) (market value of common equity plus carrying value of preferred stock plus long-term debt (including current portion) and notes payable).
Exhibits A-8 and B-8 Number of employees (number of employees, either at year-end or yearly average, including part-time and seasonal workers and employees of consolidated subsidiaries; excludes contract workers and unconsolidated subsidiaries).
Beta calculated using the sum beta method applied to monthly returns for 1963 through the latest year (see the 2011 SBBI Valuation Yearbook pp. 77-78 for a description of the sum beta method). Standard deviation of annual historical equity returns.
Arithmetic average historical risk premium over long-term Treasuries (average return on equity in excess of long-term Treasury bonds) since 1963 (RPm+s). Indicated CAPM premium, calculated as the beta of the portfolio multiplied by the average historical market risk premium since 1963 (measured as the difference between SBBI Large Stock total returns and SBBI income returns on long-term Treasury bonds). Premium over CAPM, calculated by subtracting the Indicated CAPM Premium from the Arithmetic Risk Premium (RPs). Smoothed Premium over CAPM: the fitted premium from a regression with the historical Premium over CAPM as dependent variable and the logarithm of the average sorting criteria as independent variable (RPs)
Each of the exhibits A-1 through A-8 and B-1 through B-8 displays one line of data for each of the 25 size-ranked portfolios. The A and B exhibits include the statistics outlined in Table 3. For comparative purposes, the average returns from the SBBI series for Large Companies (essentially the S&P 500 Index), Small Companies, and Long-Term Government Bond Income Returns for the period 1963 through the latest year are also reported in each exhibit.28
Geometric average historical equity return since 1963.
Arithmetic average historical equity return since 1963.
Arithmetic average historical risk premium over long-term Treasuries (average return on equity in excess of long-term Treasury bonds) since 1963 (RPm+s).
Smoothed average historical risk premium: the fitted premium from a regression with the average historical risk premium as dependent variable and the logarithm of the average sorting criteria as independent variable. (We present the coefficients and other statistics from this regression analysis in the top right hand corner of the exhibits) (RPm+s) Average carrying value of preferred stock plus long-term debt (including current portion) plus notes payable (Debt) as a percent of MVIC since 1963.
27
28
For a detailed description of the Standard and Poors Compustat data items used in the Risk Premium Report, please see Appendix A (in the full Report). Source: Morningstar EnCorr Analyzer software.
Duff & Phelps
| 20
The Size Study
The Difference between the A Exhibits and the B Exhibits The results of the Size Study are presented in Exhibits A-1 through A-8 and Exhibits B-1 through B-8. The main difference between the A and B exhibits is how they are used: the A exhibits are used if you are using a buildup method to develop cost of equity capital estimates, and the B exhibits are used if you are using the capital asset pricing model (CAPM) to develop cost of equity capital estimates. This difference in usage is a function of the type of risk premia presented in each of the exhibits: y The A exhibits provide risk premia over the risk free rate in terms of the total effect of market risk and size risk for 25 portfolios ranked by eight alternative measures of size (RPm+s ). These premia can be added to a risk free rate (Rf ) to estimate cost of equity capital (COE) in a buildup model. y The B exhibits provide risk premia over CAPM (size premia) in terms of size risk for 25 portfolios ranked by eight alternative measures of size (RPs ). These premia are commonly known as beta-adjusted size premia, or simply size premia. These premia can be added as a size adjustment to a basic CAPM to estimate cost of equity capital (COE).29
29
The basic CAPM formula is COE = Risk Free Rate + (Beta x ERP). A modified CAPM refers to the common modification to the CAPM formula that is used to incorporate an adjustment for size: COE = Risk Free Rate + (Beta x ERP) + Size Premium. Please note that the modified CAPM as presented is after addition of a size premium and prior to the addition of any company-specific risk premiums that may be applicable.
Duff & Phelps
| 21
The Size Study
The Difference Between Risk Premia Over the Risk Free Rate and Risk Premia Over CAPM The Size Study measures the relationship between equity returns and up to eight measures of size, including market capitalization. As size decreases, returns tend to increase. The Size Study develops two primary types of risk premia, those that can be added to a risk free rate if you are using the buildup method (found in Exhibits A-1 through A-8), and premia over CAPM, which are commonly referred to as beta adjusted size premia, or simply size premia (found in Exhibits B-1 through B-8). Size premia can be added as a size adjustment if you are using the capital asset pricing model (CAPM). Risk Premium Over Risk Free Rate, RPm+s Risk premia over the risk free rate represent the difference between the historical (observed) return of equities over the return of the risk free rate. A long-run average historical risk premium is often used as an indicator of the expected risk premium of a typical equity investor. Returns are based on dividend income plus capital appreciation and represent returns after corporate taxes (but before owner-level taxes).
To estimate historical risk premiums, the average rate of return for each of the 25 size-based portfolios is calculated over the sample period, and then the average income return of long-term Treasury bonds (using SBBI data) over the same period is subtracted. The result is a clear negative relationship between size and premium over long-term bond yields (i.e. as size decreases, the return over the risk free rate increases). This difference is a measure of risk in terms of the total effect of market risk and size risk. In Figure 5, for example, an abbreviated version of Exhibit A-6 is shown. The average annual arithmetic return for Portfolio 25 is 20.79 percent over the time period 19632010, and the average annual long-term Treasury income return over this period was 6.90%. This implies actual excess returns of 13.89 percent (20.79% - 6.90%) for this portfolio. Because these premia have an embedded measure of market (i.e. beta) risk, these premia are appropriate for use in buildup methods that do not already include a measure of market risk, but are not appropriate for use in models (e.g. CAPM) that already have a measure of market risk (in the form of beta). y Risk premia over the risk free rate (RPm+s ) are presented in Exhibits A-1 through A-8. In the 2011 Report, these risk premia are calculated over the period 1963 (the year that the Compustat database was inaugurated) through December 2010.
Figure 5: Calculating Risk Premia Over the Risk Free Rate (RPm+s ) Companies Ranked by 5-Year Average EBITDA Historical Equity Risk Premium: Average Since 1963 Data for Year Ending December 31, 2010 Exhibit A-6
Portfolio Rank by Size 1 2 Average EBITDA ($mils.) 17,770 4,983 Log of Average EBITDA 4.25 3.70 Number as of 2010 33 32 Beta (SumBeta) Since 63 0.79 0.84 Standard Deviation of Returns 16.26% 16.02% Geometric Average Return 11.46% 11.14% Arithmetic Average Return 12.71% 12.39% Arithmetic Average Risk Premium 5.81% 5.48% Smoothed Average Risk Premium 4.01% 5.66% Average Debt/MVIC 23.33% 29.11%
///
25 14 1.15 417 1.31 28.71% 17.15% 9.84% 13.72% 6.88% 20.79% 11.30% 16.53% 6.90% 13.89% 4.40% 9.63% 13.28% 22.86%
Large Stocks (Ibbotson SBBI data) Small Stocks (Ibbotson SBBI data) Long-Term Treasury Income (Ibbotson SBBI data)
Duff & Phelps
| 22
The Size Study
Risk Premium Over CAPM (Size Premium), RPs Risk Premia over CAPM represent the difference between historical (observed) excess return and the excess return predicted by CAPM. The historical excess return of portfolios comprised of smaller companies is greater than the excess return predicted by the CAPM for these portfolios. The difference is a measure of risk in terms of the effect of size risk, and is commonly referred to as a beta-adjusted size premium, or simply size premium. For each portfolio, a size premium is calculated using the methodology for doing so as described in the SBBI Valuation Yearbook.30 The formula for this adjustment is: Size Premium = Portfolio Premium (Portfolio Beta x Realized Market Premium) where: Size premium: the difference in historical excess returns (i.e. what actually happened), and the excess returns predicted by CAPM. Portfolio premium: the actual return over the risk-free interest rate (i.e. excess return) earned by a given portfolio between 1963 and 2010. Portfolio beta: the beta estimated relative to the S&P 500 Index using annual returns between 1963 and 2010. Realized market premium: the average annual excess return of the S&P 500 Index between 1963 and 2010 over the long-term risk free rate. This adjustment can be thought of as simply what actually happened (the portfolio premium) minus what CAPM predicted would happen (the portfolio beta times the realized market premium).31
30 31
2011 Ibbotson SBBI Valuation Yearbook (Chicago, Morningstar, 2011), Chapter 7, Firm Size and Return, pages 85105. The basic CAPM equation is COE = Rf + ( x ERP), which can be rewritten as COE Rf = ( x ERP). COE (i.e. expected return) minus the risk free rate (Rf ) is, by definition, the expected return over the risk free rate, and therefore, so is ( x ERP).
Duff & Phelps
| 23
The Size Study
For example, an abbreviated version of Exhibit B-6 is shown in Figure 6. The average annual arithmetic return for Portfolio 25 is 20.79 percent over the time period 19632010, and the average annual long-term Treasury income return over this period was 6.90 percent. This implies actual excess returns of 13.89 percent (20.79% 6.90%) for this portfolio. Portfolio 25 has a calculated beta 32 of 1.31, and the realized market premium over the 19632010 period is 4.40 percent 33. This implies that predicted excess return according to CAPM is 5.76 percent (1.31 x 4.40%). The size premium for Portfolio 25 in Exhibit B-6 is therefore 8.13 percent, which is what actually happened (13.89%) minus what CAPM predicted (5.76%).
The risk premia over CAPM (i.e. size premia) published in the Risk Premium Report are adjusted for beta. In other words, the portion of excess return that is not attributable to beta is controlled for, or removed, leaving only the size effects contribution to excess return. These premia are appropriate for use in the capital asset pricing model (CAPM), and in buildup methods that do not otherwise already have a measure of size risk.34 y Risk premia over CAPM, or size premia (RPs ) are presented in Exhibits B-1 through B-8. In the 2011 Report, these risk premia are calculated over the period 1963 (the year that the Compustat database was inaugurated) through December 2010.
Figure 6: Calculating Size Premia (RPs) Companies Ranked by 5-Year Average EBITDA Historical Equity Risk Premium: Average Since 1963 Data for Year Ending December 31, 2010 Exhibit B-6
Portfolio Rank by Size 1 2 Average EBITDA ($mils.) 17,770 4,983 Log of Size 4.25 3.70 Beta (SumBeta) Since 63 0.79 0.84 Arithmetic Average Return 12.71% 12.39% Arithmetic Average Risk Premium 5.81% 5.48% Indicated CAPM Premium 3.45% 3.68% Premium over CAPM 2.36% 1.81% Smoothed Premium over CAPM 0.65% 1.80%
///
25 Large Stocks (Ibbotson SBBI data) Small Stocks (Ibbotson SBBI data) Long-Term Treasury Income (Ibbotson SBBI data) 14 1.15 1.31 20.79% 11.30% 16.53% 6.90% 13.89% 4.40% 9.63% 5.76% 8.13% 7.14%
32
33 34
The betas presented in the Risk Premium Report are sum betas. Smaller companies generally trade more infrequently and exhibit more of a lagged price reaction (relative to the market) than do large stocks. One of the ways of capturing this lag movement is called sum beta. See Ibbotson, Roger G., Paul D. Kaplan, and James D. Pearson. Estimates of Small-Stock Betas Are Much Too Low, Journal of Portfolio Management, Summer 1997. As derived from the average difference in the annual average returns of the S&P 500 Index and SBBI long-term government Treasury bond income returns. Source: Morningstar EnCorr Analyzer software. For example, the size premia presented in Exhibit B cannot be used in Buildup 1. The Buildup 1 method uses risk premia over the risk free rate (from Exhibit A) that already have a measure of risk in terms of the total effect of market risk and size risk, (RPm+s ). Using size premia in Buildup 1 would be double counting size risk.
Duff & Phelps
| 24
The Size Study
Overview of Methods Used to Estimate Cost of Equity Capital Using the Size Study The Size Study provides two methods of estimating COE for a subject company, Buildup 1 and CAPM, plus one method for estimating unlevered COE (the cost of equity capital assuming a firm is financed 100% with equity and 0% debt).35 Some users of the Report have inquired whether the Size Study can be used in conjunction with the industry risk premia (IRPs) published in the SBBI Valuation Edition Yearbook, so we also include an alternative method in which a rudimentary adjustment is made to an IRP and then utilized in a modified buildup model, Buildup 2, that includes a separate variable for the industry risk premium.36 These methods are summarized below in equation format, and summarized in Figure 7 in graphical building blocks format.
1) Buildup 1 COEBuildup 1 = (Risk Free Rate) + (Risk Premium in Excess of (over) the Risk Free Rate) + (Equity Risk Premium Adjustment) Example 1a: using guideline portfolios: page 31 (in the full Report). Example 1b: using regression equations: page 33 (in the full Report). 2) Buildup 1-Unlevered COEBuildup 1-Unlevered = (Risk Free Rate) + (Unlevered Risk Premium in Excess of (over) the Risk Free Rate) + (Equity Risk Premium Adjustment) Example 2a: using guideline portfolios: page 40 (in the full Report). Example 2b: using regression equations: page 44 (in the full Report). 3) Capital asset pricing model (CAPM) COECAPM = (Risk Free Rate) + (Beta x Equity Risk Premium) + (Size Premium) Example 3a: using guideline portfolios: page 29 Example 3b: using regression equations: page 32 4) Buildup 2 COEBuildup 2 = (Risk Free Rate) + (Equity Risk Premium) + (Size Premium) + (Adjusted Industry Risk Premium) Example 4a: using guideline portfolios: page 57 (in the full Report). Example 4b: using regression equations: page 58 (in the full Report).
35 36
Unlevered risk premia over the risk free rate are presented in Exhibits C-1 through C-8 (in the full Report). Duff & Phelps does not publish IRPs. A source of IRPs is Morningstars Ibbotson SBBI Valuation Yearbook, Table 3-5.
Duff & Phelps
| 25
The Size Study
Figure 7: Four Methods of Estimating Cost of Equity Capital with the Size Study 37 Buildup 1 CAPM + Smoothed Risk Premium Over CAPM (Size Premium), RPs Cost of Equity + (Beta x ERP) Basic CAPM Risk Free Rate, Rf (Use Exhibit B size premia) Buildup 2 + IRPadjusted + Smoothed Risk Premium Over CAPM (Size Premium), RPs + ERP Risk Free Rate, Rf (Use Exhibit B size premia) Cost of Equity
+ ERP Adjustment*
+ Smoothed Risk Premium Over Risk Free Rate, RPm+s Risk Free Rate, Rf (Use Exhibit A risk premia) Buildup 1-Unlevered + ERP Adjustment*
+ Smoothed Unlevered Risk Premium Over Risk Free Rate, RPm+s, unlevered Risk Free Rate, Rf (Use Exhibit C unlevered risk premia)
Cost of Equity
Cost of Equity
* ERP Adjustment: The difference between the historical (19632010) equity risk premium (ERP) and a user of the 2011 Reports own forward ERP estimate: ERP Adjustment = Users ERP Historical ERP (19632010) The ERP Adjustment is made only in the Buildup 1, Buildup1Unlevered, Buildup 1-High-Financial-Risk, Buildup 3, and Buildup 3-Unlevered methods. Please refer to the individual examples provided for these models for more information.
NOTE: This section includes an example of using the Reports size premia data to estimate cost of equity capital using the CAPM method, plus an overview of the unlevering/relevering methodology employed in the 2011 Report. Complete examples for using the Reports size premia and risk premia to estimate cost of equity capital using the Buildup 1, Buildup 1-Unlevered, and Buildup 2 methods are available in the full version of the 2011 Report.
37
The relative size of the building blocks in Figure 7 do not necessarily represent the relative size of the various inputs.
Duff & Phelps
| 26
The Size Study
As shown in Figure 8, there are up to eight alternative size measures that can be used with any of the four methods of estimating COE provided by the Size Study. It is important to note that it would not be unusual for fewer than eight of these measures to be available for any given subject company. For example, market value of equity will probably not be available for a closely-held company, nor will market value of invested capital (in which market value of equity is embedded). In cases where fewer than eight size measures are available, it is generally acceptable to use the size measures that are available. Figure 8: Subject Company Size Characteristics (used in all examples)
Appropriate Exhibit Size Measure (in $millions) Market Value of Equity Book Value of Equity 5-year Average Net Income Market Value of Invested Capital Total Assets 5-year Average EBITDA Sales Number of Employees $120 $100 $10 $180 $300 $30 $250 200 Buildup 1 A-1 A-2 A-3 A-4 A-5 A-6 A-7 A-8 Buildup 1Unlevered C-1 C-2 C-3 C-4 C-5 C-6 C-7 C-8 CAPM B-1 B-2 B-3 B-4 B-5 B-6 B-7 B-8 Buildup 2 B-1 B-2 B-3 B-4 B-5 B-6 B-7 B-8
The C exhibits, which have not been discussed in detail thus far, provide unlevered versions of the risk premia over the risk free rate found in the A Exhibits. These unlevered premia (RPm+s, unlevered ) can be used to estimate cost of equity capital assuming a firm is financed 100% with equity and 0% debt.38 The C exhibits also provide another important functionthey serve as a link between the Size Study and the Risk Study. These exhibits can be used to compare a subject companys fundamental risk characteristics to the fundamental risk characteristics of portfolios made up of similarly-sized companies. Examples of this important functionality are found in the Risk Study, which begins on page 39. In each of the following examples of using the Size Study to estimate COE, the subject company size measures summarized in Figure 8 will be used (total assets of $300 million, for instance, will be used in all examples). Also, the long-term risk free rate, ERP, and the ERP Adjustment established in the first example (Example 1a, Buildup 1 using guideline portfolios) will be used (as appropriate) for all the subsequent examples, mirroring the fact that for any given valuation engagement, the same risk free rate and ERP will generally be used in each of the models presented by the individual analyst. Please note that for any given valuation engagement these inputs may (and probably will) be different than the ones used in the examples.
Figure 8 also includes the data exhibits in which the appropriate risk premia for each of the size measures can be found. For example, for use in the Buildup 1 method, risk premia over the risk free rate (RPm+s) for Total Assets are found in Exhibit A-5. For use in the CAPM method, the appropriate premia over CAPM (RPs, or size premia) for Total Assets are found in Exhibit B-5.
38
The D exhibits also include unlevered risk premia, but these are unlevered versions of the corresponding levered risk premia found in the Risk Studys D exhibits. The unleverered premia in the C exhibits are unlevered versions of the levered risk premia found in the Size Studys A exhibits.
Duff & Phelps
| 27
The Size Study
Estimating Cost of Equity Capital Using the CAPM Method
CAPM + Smoothed Risk Premium Over CAPM (Size Premium), RPs + (Beta x ERP) Basic CAPM Risk Free Rate, Rf (Use Exhibit B size premia) The capital asset pricing model (CAPM) is the most widely used method for estimating the cost of equity capital. For example, one survey found that while many firms use multiple methods of estimating the cost of equity capital, 75% of them use the CAPM.70 Despite its criticisms, the CAPM has been one of the most widely used models for estimating the cost of equity capital for more than 30 years. The basic CAPM formula for estimating the cost of equity capital (COE) is: COECAPM = Rf + ( x ERP) where: Rf = the risk free rate as of the valuation date (typically a longterm US Treasury bond yield) = a measure of market (called systematic) risk of a stock; the sensitivity of changes in the returns (dividends plus price changes) of a stock relative to changes in the returns of a specific market benchmark or index.71 = the equity risk premium. The ERP is the rate of return added to a risk free rate to reflect the additional risk of equity instruments over risk free instruments. Cost of Equity
Research tells us that the CAPM often misprices risk for certain investments. Specifically, researchers have observed that commonly used methods of measuring risk used in the CAPM (specifically, beta) often understate the risk (and thus understate the required return) for small company stocks. Examination of market evidence shows that within the context of CAPM, beta does not fully explain the difference between small company returns and large company returns. In other words, the historical (observed) excess return of portfolios comprised of smaller companies is greater than the excess return predicted by the CAPM for these portfolios. This premium over CAPM is commonly known as a beta-adjusted size premium or simply size premium.72 It follows that the size premium is a necessary correction to the basic CAPM because risk, as measured by the betas of smaller companies (even sum betas), is systematically underestimated.73 Moreover, the size effect is not just evident for the smallest companies in the marketplace, but is evident for all but the largest groups of companies, including companies with a market capitalization in excess of $1 billion. A common practice is to incorporate this evidence by adding a size premium to the CAPM formula when valuing companies that are comparatively small. The modified CAPM formula is74: COECAPM = Rf + ( x ERP) + RPs where: RPs = the beta-adjusted size premium. It is important to note that the risk premia over CAPM (i.e. size premia) published in the Risk Premium Report are adjusted for beta.75 In other words, the portion of excess return that is not attributable to beta is controlled for, or removed, leaving only the size effects contribution to excess return. These premia are appropriate for use in the capital asset pricing model (CAPM), and in buildup methods that do not otherwise already have a measure of size risk.76
ERP
70 71 72
73
74 75 76
John R. Graham and Campbell R. Harvey, The Theory and Practice of Corporate Finance, Journal of Financial Economics (May 2001): 187243. For the purposes of this report, the market is defined as the S&P 500 Index. The S&P 500 Index is a broad-based, market-capitalization-weighted index widely regarded as being representative of the overall market. A sample of academic research articles include: Rolf Banz, The Relationship Between Return and Market Value of Common Stocks, Journal of Financial Economics (March 1981): 318; Eugene Fama and Kenneth French, The Cross Section of Expected Stock Returns, Journal of Finance (June 1992): 427486; Kent Daniel and Sheridan Titman, Evidence on the Characteristics of Cross Sectional Variation in Stock Returns, Journal of Finance (March 1997): 133. The betas presented in the Risk Premium Report are sum betas. Smaller companies generally trade more infrequently and exhibit more of a lagged price reaction (relative to the market) than do large stocks. One of the ways of capturing this lag movement is called sum beta. See Ibbotson, Roger G., Paul D. Kaplan, and James D. Pearson. Estimates of Small-Stock Betas Are Much Too Low, Journal of Portfolio Management, Summer 1997. A modified CAPM typically refers to the common modification to the CAPM formula that is used to incorporate an adjustment for size. For a detailed discussion of how premia over CAPM (size premia) are calculated, see The Difference Between Risk Premia Over the Risk Free Rate and Risk Premia Over CAPM on page 22. For example, the size premia presented in Exhibit B cannot be used in Buildup 1. The Buildup 1 method uses risk premia over the risk free rate (from Exhibit A) that already have a measure of risk in terms of the total effect of market risk and size risk, (RPm+s). Using size premia in Buildup 1 would be double counting size risk.
Duff & Phelps
| 28
The Size Study
Please note that base estimates of COE developed with the modified CAPM equation presented above are after addition of a size premium, but prior to the addition of any company-specific risk premiums (RPu) that the individual analyst may deem to be applicable.77 Company-specific risk can be added by the individual analyst to the modified CAPM in the following fashion: COECAPM = Rf + ( x ERP) + RPs + RPu The Risk Premium Report provides two ways for analysts to match their subject companys size (or risk) characteristics with the appropriate smoothed premia from the data exhibits: the guideline portfolio method and the regression equation method.78 In general, the regression equation method is preferred because this method allows for interpolation between the individual guideline portfolios, although the guideline portfolio method is less complicated, and more direct. Examples of both the guideline portfolio method and the regression equation method follow, starting with the simpler guideline portfolio method.
Example 3a: CAPM Method (using guideline portfolios) Four pieces of information are needed to estimate the cost of equity capital using the CAPM method and guideline portfolios: a risk free rate (Rf), a beta (), an equity risk premium (ERP), and a risk premium over CAPM (RPs, otherwise known as a beta-adjusted size premium). All of the information needed is summarized in Figure 27. Figure 27: Information Needed to Estimate COE Using CAPM and Guideline Portfolios
Step 1 Step 2 Step 3 Step 4 Step 5
Rf
ERP
RPs
(using regression equations)
COE
This example utilizes the risk free rate (Rf) and ERP that were established in Example 1a (see page 31 in the full Report). This mirrors the fact that for any given valuation engagement the same risk free rate and ERP will generally be used in each of the models presented by the individual analyst. For any given valuation engagement these inputs may (and probably will) be different than the ones used in the examples.
77 78
The u in RPu stands for unique risk or company-specific risk, and is also commonly referred to as unsystematic risk. See pages 1416 for a detailed explanation of the differences between the guideline portfolio method and the regression equation method.
Duff & Phelps
| 29
The Size Study
Step 1, Risk Free Rate (Rf): The risk free rate is typically a long-term US Treasury bond yield as of the valuation date. This example utilizes the assumed long-term treasury yield of 4.1 percent established in Example 1a (page 31 in the full Report). Step 2, Beta (): Beta is a measure of the sensitivity of changes in the returns (dividends plus price changes) of a stock relative to changes in returns of a specific market benchmark or index. Duff & Phelps does not currently publish company betas or peer group betas.79 Because the sum betas calculated for the 25 size-ranked portfolios in the B exhibits are betas for a particular size of company (rather than a particular industry), they would in all likelihood not be appropriate for use within a CAPM estimate of COE, where the beta should be a measure of market, (or industry) risk. For this example, a beta of 1.2 is assumed. Step 3, Equity Risk Premium (ERP): The ERP is the rate of return added to a risk free rate to reflect the additional risk of equity instruments over risk-free instruments. For this example, the Duff & Phelps Recommended ERP as of the end of 2010 (5.5%) is assumed.80, 81 Step 4, Risk Premium Over CAPM (size premium) (RPs): Match the various size measures of the subject company with the guideline portfolios composed of companies of similar size in Exhibits B-1 through B-8, and identify the corresponding smoothed average risk premium over CAPM (i.e. size premium).
The subject company in this example has a market value of equity of $120 million, and the appropriate data exhibit is Exhibit B-1 (see Figure 8 on page 27). An abbreviated version of Exhibit B-1 is shown in Figure 28. Of the 25 portfolios, the portfolio that has an average market value closest to the subject companys $120 million market value is portfolio 25 ($68 million). The corresponding smoothed average size premium is 8.02 percent (8.0 percent, rounded). Match each of the subject companys size measures in this fashion. For example, the second size measure for the subject company in this example is book value of equity of $100 million. Of the 25 guideline portfolios in Exhibit B-2 (not shown here), the portfolio that has an average book value of equity closest to the subject companys $100 million book value is portfolio 25 ($60 million). The corresponding smoothed average size premium is therefore 6.4 percent. After all of the available size measures for the subject company have been matched to the closest guideline portfolio in the appropriate exhibit and the corresponding smoothed average size premium has been identified for each, Step 4 is complete.
Figure 28: Exhibit B-1 (abbreviated) Companies Ranked by Market Value of Equity Historical Equity Risk Premium: Average Since 1963 Data for Year Ending December 31, 2010
Portfolio Rank by Size 1 2 3 Average Mkt Value ($mils.) 109,765 32,309 22,008 Beta (SumBeta) Since 63 0.84 0.95 0.93 Arithmetic Average Return 11.70% 10.57% 11.26% Arithmetic Average Risk Premium 4.80% 3.67% 4.35% Indicated CAPM Premium 3.67% 4.18% 4.09% Premium over CAPM 1.13% -0.51% 0.27% Smoothed Premium over CAPM -1.22% 0.31% 0.79%
Log of Size 5.04 4.51 4.34
///
24 25 232 68 2.36 1.83 1.25 1.29 19.26% 23.28% 12.36% 16.37% 5.50% 5.67% 6.86% 10.71% 6.48% 8.02%
79 80
81
Company betas and industry betas are available from multiple sources, including Bloomberg, MSCI, and Value Line. For more information on the equity risk premium, see Cost of Capital: Applications and Examples 4th ed., by Shannon P. Pratt and Roger J. Grabowski (John Wiley & Sons, Inc., 2010), Chapter 9, Equity Risk Premium, pages 115158. For more information on cost of capital issues, including developing risk-free rates and ERP during periods of flight to quality, please visit www.duffandphelps.com/costofcapital.
Duff & Phelps
| 30
The Size Study
Step 5, Estimate Cost of Equity (COE): With the completion of Steps 1 through 4, the information needed to estimate a base cost of equity capital using the CAPM is now completed. The risk premia over CAPM (RPs, or size premia) can now be added to the basic CAPM equation (COECAPM = Rf + ( x ERP) + RPs) to estimate an indicated cost of equity capital (COE) for the subject company, as illustrated in Figure 29.
The range of COE estimates for the hypothetical subject company in this example is 16.5 percent to 18.7 percent, with an average of 17.3 percent, and a median of 17.0 percent. The mean represents the average estimate, but the mean can be unduly influenced by very large or very small outliers. For this reason, the median estimate is generally preferred to the mean. The median estimate tends to not be as heavily influenced by very large or very small outliers, and can be considered a measure of the typical estimate in the group. Remember that the full CAPM equation is: COECAPM = Rf + ( x ERP) + RPs + RPu The base COE estimates derived in this example are therefore prior to the addition of any other company-specific risk premiums (RPu) that the individual analyst may deem appropriate.
Figure 29: CAPM COE Inputs (using guideline portfolios)

Step 1 Size Measure (in $millions) Market Value of Equity Book Value of Equity 5-year Average Net Income Market Value of Invested Capital Total Assets 5-year Average EBITDA Sales Number of Employees $120 $100 $10 $180 $300 $30 $250 200 Appropriate Exhibit B-1 B-2 B-3 B-4 B-5 B-6 B-7 B-8 Guideline Portfolio 25 25 24 25 24 24 24 25 Risk Free Rate, Rf 4.1% 4.1% 4.1% 4.1% 4.1% 4.1% 4.1% 4.1% + + + + + + + + Step 2 Beta (1.2 (1.2 (1.2 (1.2 (1.2 (1.2 (1.2 (1.2 x x x x x x x x Step 3 Step 4 Smoothed Premium Over CAPM (size premium), RPs + + + + + + + + 5.8% 8.0% = = = = = = = = 16.5% Step 5
ERP 5.5%) 5.5%) 5.5%) 5.5%) 5.5%) 5.5%) 5.5%) 5.5%)
COE 18.7%
Mean (average) values Median (typical) values
4.1% 4.1%
+ +
(1.2 (1.2
x x
5.5%) 5.5%)
+ +
6.6% 6.3%
= =
17.3% 17.0%
NOTE: Some values intentionally blurred.
Duff & Phelps
| 31
The Size Study
Example 3b: CAPM Method (using regression equations) When the subject company size measures do not exactly match the respective average company size of the guideline portfolios, the data exhibits provide a straightforward way to interpolate an exact risk premium over CAPM between guideline portfolios using the regression equation method. The only difference between estimating cost of equity capital (COE) using the CAPM method using guideline portfolios (as in the previous example) and estimating COE using the CAPM method using regression equations is how the risk premia over CAPM (RPs, or size premia) are identified in Step 4.
In the previous example, the smoothed average risk premia over CAPM published in the Report for the appropriate guideline portfolios were used to estimate COE.82 In this example, however, the regression equations found in each of the data exhibits will be used to calculate custom interpolated size premia, based upon the specific size measures of the subject company. This example utilizes the long-term risk free rate (Rf) and ERP established in a previous example (the Size Studys Buildup 1 method using guideline portfolios; see page 31 in the full Report), and the Beta (), established for the previous example (Example 3a on page 29). This mirrors the fact that for any given valuation engagement the same inputs will generally be used in each of the models presented by the individual analyst. Please note that for any given valuation engagement these inputs may (and probably will) be different than the ones used in the examples. The only missing ingredients needed to estimate COE are the premia over CAPM, or size premia (RPs), as summarized in Figure 30.
Figure 27: Information Needed to Estimate COE Using CAPM and Regression Equations
Step 1 Step 2 Step 3 Step 4 Step 5
Rf
ERP
RPs
COE
Figure 30: NeededSmoothed Premia Over CAPM (RPs , or Size Premia) Calculated Using Regression Equations
Step 1 Size Measure (in $millions) Market Value of Equity Book Value of Equity 5-year Average Net Income Market Value of Invested Capital Total Assets 5-year Average EBITDA Sales Number of Employees $120 $100 $10 $180 $300 $30 $250 200 Appropriate Exhibit B-1 B-2 B-3 B-4 B-5 B-6 B-7 B-8 Risk Free Rate, Rf 4.1% 4.1% 4.1% 4.1% 4.1% 4.1% 4.1% 4.1% + + + + + + + + Step 2 Beta (1.2 (1.2 (1.2 (1.2 (1.2 (1.2 (1.2 (1.2 x x x x x x x x Step 3 Step 4 Smoothed Premium Over CAPM (size premium), RPs + + + + + + + + ? ? ? ? ? ? ? ? = = = = = = = = Step 5
ERP 5.5%) 5.5%) 5.5%) 5.5%) 5.5%) 5.5%) 5.5%) 5.5%)
COE
4.1% 4.1%
+ +
(1.2 (1.2
x x
5.5%) 5.5%)
+ +
= =
82
The smoothed risk premia published in the Risk Premium Report are based upon the average size (or risk) measure in each of the respective guideline portfolios.
Duff & Phelps
| 32
The Size Study
Step 1, Risk Free Rate (Rf): The risk free rate is typically a long-term US Treasury bond yield as of the valuation date. This example utilizes the assumed long-term treasury yield of 4.1 percent used in Example 1a (page 31 in the full Report). Step 2, Beta (): Beta is a measure of the sensitivity of a stocks price relative to movements of a specific market benchmark or index. For this example, the beta of 1.2 that was assumed in Example 3a (page 29) is assumed. Step 3, Equity Risk Premium (ERP): The ERP is the rate of return added to a risk-free rate to reflect the additional risk of equity instruments over risk-free instruments. For this example, the Duff & Phelps Recommended ERP as of the end of 2010 (5.5%) is assumed.83, 84 Step 4, Risk Premium Over CAPM (RPs): The hypothetical subject company in this example has a market value of equity of $120 million, and the appropriate Size Study data exhibit to use is Exhibit B-185. In this case one would expect that the smoothed average premium over CAPM, or size premium, would fall somewhere between 6.48 percent (the smoothed size premium for Portfolio 24) and 8.02 percent (the smoothed size premium for Portfolio 25), as illustrated in Figure 31.
An easy way to calculate a custom interpolated risk premium over CAPM (RPs, or size premia) in between Portfolio 24 and Portfolio 25 is by using the regression equations provided for this purpose in each of the data exhibits. The regression equations are located in the same spot in each of the exhibits (see Figure 4 on page 16).86
Figure 31: Exhibit B-1 (abbreviated) Companies Ranked by Market Value of Equity Historical Equity Risk Premium: Average Since 1963 Data for Year Ending December 31, 2010
Portfolio Rank by Size 1 2 3 Average Mkt Value ($mils.) 109,765 32,309 22,008 Beta (SumBeta) Since 63 0.84 0.95 0.93 Arithmetic Average Return 11.70% 10.57% 11.26% Arithmetic Average Risk Premium 4.80% 3.67% 4.35% Indicated CAPM Premium 3.67% 4.18% 4.09% Smoothed Premium over CAPM (RPs) -1.22% 0.31% 0.79%
Log of Size 5.04 4.51 4.34
Premium over CAPM 1.13% -0.51% 0.27%
///
24 Subject Company 25 232 120 68 1.83 1.29 23.28% 16.37% 5.67% 10.71% 2.36 1.25 19.26% 12.36% 5.50% 6.86% 6.48% ? 8.02%
83
84 85 86
For more information on the equity risk premium, see Cost of Capital: Applications and Examples 4th ed., by Shannon P. Pratt and Roger J. Grabowski (John Wiley & Sons, Inc., 2010), Chapter 9, Equity Risk Premium, pages 115158. For more information on cost of capital issues, including developing risk-free rates and ERP during periods of flight to quality, please visit www.duffandphelps.com/costofcapital. The same eight size measures (for a hypothetical subject company) are used in all examples of estimating COE using the Size Study, as outlined in Figure 8 on page 27. In addition to regression equations for interpolating risk premia between guideline portfolios in the Size Studys A and B exhibits, the Risk Studys D exhibits also provide regression equations for easy interpolation of risk premia between guideline portfolios, as do the C exhibits (for unlevered risk premia).
Duff & Phelps
| 33
The Size Study
The regression equation provided in Exhibit B-1, which includes 25 portfolios ranked by market value 87, is: Smoothed Premium = 13.285% 2.879% * Log(Market Value) To calculate an interpolated smoothed risk premium over CAPM (RPs, size premia) for the subject companys $120 million market value, substitute the market value into the regression equation as follows88: Smoothed Premium= 13.285% 2.879% * Log(120) 7.3% = 13.285% 2.879% * 2.08
Continue interpolating smoothed risk premium over CAPM for each size measure available for the subject company using the regression equations from the data exhibits. For example, the second size measure for the subject company is book value of equity of $100 million. The equation found on Exhibit B-2 is: Smoothed Premium= 9.849% 1.957% * Log(Book Value) The interpolated smoothed risk premium over CAPM is therefore 5.9 percent (9.849% - 1.957% * 2). After interpolating smoothed size premia for all of the subject companys available size measures, Step 4 is complete, as shown in Figure 32.
Figure 32: Calculation of Risk Premia Over CAPM (RPs) Using Regression Equations
Step 4 Subject Company Size Measures (in $millions) 120 100 10 180 300 30 250 200 Smoothed Risk Premium Over CAPM (size premium), RPs = = = = = = = = 7.3%
Appropriate Exhibit B-1 B-2 B-3 B-4 B-5 B-6 B-7 B-8
Size Measure Market Value of Equity Book Value of Equity 5-year Average Net Income Market Value of Invested Capital Total Assets 5-year Average EBITDA Sales Number of Employees
Appropriate Regression Equation Smoothed Premium = 13.285% 2.879% * Log(Market Value ) Smoothed Premium = Smoothed Premium = Smoothed Premium = Smoothed Premium = Smoothed Premium = Smoothed Premium = Smoothed Premium = * Log(Book Value) * Log(Net Income) * Log(MVIC) * Log(Assets) * Log(E BITDA) * Log(Sales) * Log(Employees)
87 88
Figure 8 on page 27 lists the appropriate B exhibits in which the size premia for each of the eight size measures can be found. Please note that the logarithmic relationship is base-10, and that the financial size data is in millions of dollars, such that the log of $10 million is log(10), not log(10,000,000). The formula to calculate a values base-10 logarithm in Microsoft Excel is =log(value). The * used in the regression equation is the symbol used in Microsoft Excel to denote the multiplication symbol, x. The * format is used to denote multiplication in the regression equations in the data exhibits.
Duff & Phelps
| 34
The Size Study
Step 5, Estimate Cost of Equity (COE): With the completion of Steps 1 through 4, the information needed to estimate a base cost of equity capital using the CAPM (using regression portfolios) is now completed. The risk premiums over CAPM (RPs, or size premia) can now be added to the basic CAPM equation (COECAPM = Rf + ( x ERP) + RPs) to estimate an indicated cost of equity capital (COE) for the subject company, as illustrated in Figure 33. The range of COE estimates for the hypothetical subject company in this example is 16.6 percent to 18.0 percent, with an average of 17.2 percent, and a median of 17.2 percent. The mean estimate is the simple average of the COE estimates, but the mean can be unduly influenced by very large or very small outliers. For this reason, the median COE estimate is generally preferred to the mean. The median tends to not be as heavily influenced by very large or very small outliers, and can be considered a measure of the typical COE estimate in the group.
Remember that the full CAPM equation is: COECAPM = Rf + ( x ERP) + RPs + RPu The base COE estimates derived in this example are therefore prior to the addition of any company-specific risk premiums (RPu) that the individual analyst may deem appropriate.
Figure 33: CAPM COE Inputs (using regression equations)

Step 1 Size Measure (in $millions) Market Value of Equity Book Value of Equity 5-year Average Net Income Market Value of Invested Capital Total Assets 5-year Average EBITDA Sales Number of Employees $120 $100 $10 $180 $300 $30 $250 200 Appropriate Exhibit B-1 B-2 B-3 B-4 B-5 B-6 B-7 B-8 Risk Free Rate, Rf 4.1% 4.1% 4.1% 4.1% 4.1% 4.1% 4.1% 4.1% + + + + + + + + Step 2 Beta (1.2 (1.2 (1.2 (1.2 (1.2 (1.2 (1.2 (1.2 x x x x x x x x Step 3 Step 4 Smoothed Premium Over CAPM (size premium), RPs + + + + + + + + 7.3% 5.9% = = = = = = = = Step 5
ERP 5.5%) 5.5%) 5.5%) 5.5%) 5.5%) 5.5%) 5.5%) 5.5%)
COE 18.0% 16.6%
4.1% 4.1%
+ +
(1.2 (1.2
x x
5.5%) 5.5%)
+ +
6.5% 6.5%
= =
17.2% 17.2%
Duff & Phelps
| 35
The Size Study
Unlevered Cost of Equity Capital In the 2011 Report, the methodology and assumptions for unlevering risk premiums reported in Exhibits C-1 through C-8 were updated.55 Unlevered premia are used to estimate cost of equity capital assuming a firm is financed 100% with equity and 0% debt. Generally, as the percentage of leverage (debt) in a companys capital structure increases, the cost of equity capital increases. The unlevered realized risk premiums displayed in exhibits C-1 through C-8 are also informative in that they generally indicate that the market views smaller companies operations to be riskier than the operations of larger companies (i.e., unlevered risk premiums increase as size decreases).
Overview of the Current Methodology and Assumptions Used to Unlever Risk Premia in the 2011 Risk Premium Report The average (levered) risk premia presented in Exhibits A-1 through A-8 are unlevered as follows56: RPunlevered = RPlevered [(Wd / We ) x ( u d ) x RPm] where: RPunlevered = Unlevered realized risk premium over the risk free rate RPlevered = Levered realized risk premium over the risk free rate u d RPm = Unlevered equity beta57 = Debt beta, assumed equal to 0.1 = General equity risk premium (ERP) estimate for the market, represented by the average historical risk premium since 1963 = Percent of debt capital in capital structure = Percent of equity capital in capital structure
Wd We
The average debt to equity (Wd / We ) ratio of the portfolio is based on the average debt to MVIC for the portfolio since 1963. A debt beta (d) of 0.1 is assumed, which is the average estimated debt beta for the companies included in portfolios 1 through 25 over the years 1963 through 2010 after excluding high-financial-risk companies (highfinancial-risk companies are excluded from the base set of companies used in the analysis performed in the Size Study). A debt beta greater than zero indicates debt capital is bearing risk of variability of operating net cash flow in that interest payments and principal repayments may not be made when owed, inferring that tax deductions on the interest expense may not be realized in the period in which the interest is paid.58 Preferred capital is included with debt capital in measuring the effect of leverage on the risk of equity capital, which is consistent with recent research.59
55
56
57 58 59
Also updated were Exhibits C-1 through C-8 for the 2010 Duff & Phelps Risk Premium Report, applying the same (updated) methodology and assumptions. The updated 2010 Exhibits C-1 through C-8 can downloaded here www.duffandphelps.com/CostofCapital. Derived from R.S. Harris and J. J. Pringle, Risk-Adjusted Discount Rates Extensions from the Average Risk Case, Journal of Financial Research (Fall 1985) 237-244. Also see: Arzac, Enrique R., and Lawrence R. Glosten. A Reconsideration of Tax Shield Valuation. European Financial Management (2005): 453-461. For a more complete discussion see chapter eleven in Cost of Capital: Applications and Examples 4th ed. by Shannon Pratt and Roger Grabowski, Wiley (2010). Unlevered betas are often called asset betas because they represent the risk of the operations of the business with the risk of financial leverage removed. For a more complete discussion see Chapter 11 in Cost of Capital: Applications and Examples 4th ed. by Shannon Pratt and Roger Grabowski, Wiley (2010). C.S. Agnes Cheng, C.Z. Liu, K. Newberry, and K.J. Reichelt, Should Preferred Stock be Classified as a Liability? Evidence from Implied Cost of Equity Capital, working paper (September 2007).
Duff & Phelps
| 36
The Size Study
An example of unlevering the average risk premia from the A exhibits is demonstrated using the information found in Figure 17a, 17b, and 17c (these are abbreviated versions of exhibits A-2, B-2, and C-2, respectively). The unlevered average risk premium of Portfolio 25 in Exhibit C-2 (Figure 17c) is 10.74 percent, calculated using the following information from Figure 17a, Figure 17b, and Figure 17c: y The arithmetic average risk premium of Portfolio 25 in Exhibit A-2 (see Figure 17a) is 11.98 percent. y The debt to market value of equity (Wd / We) of Portfolio 25 in Exhibit C-2 (see Figure 17c) is 31.9 percent. y The unlevered sum beta (u) of Portfolio 25 in Exhibit C-2 (see Figure 17c) is 0.99. y The debt beta (d) is an assumed 0.1, as discussed previously. y The market premium (RPm) used to perform the analysis in the 2011 Report is the historical ERP from 19632010, 4.40%.60 To unlever the average (levered) risk premium in Exhibit A-2 (11.98%), substitute these values into the unlevering equation presented earlier: RPunlevered = RPlevered [(Wd / We) x ( u d ) x RPm] 10.74% = 11.98% ((31.9% x (0.99 0.1) x 4.40%)) Figure 17a: Exhibit A-2 (abbreviated) Companies Ranked by Book Value of Equity Historical Equity Risk Premium: Average Since 1963 Data for Year Ending December 31, 2010 EXHIBIT A-2
Portfolio Rank by Size 1 2 Average Book Val. ($mils.) 39,141 12,811 Beta (SumBeta) Since 63 0.81 0.85 Arithmetic Average Risk Premium 5.08% 5.30%
Figure 17b: Exhibit B-2 (abbreviated) Companies Ranked by Book Value of Equity Historical Equity Risk Premium: Average Since 1963 Data for Year Ending December 31, 2010 EXHIBIT B-2
Portfolio Rank by Size 1 2 Average Book Val. ($mils.) 39,141 12,811 Premium over CAPM 1.52% 1.57%
///
25 60 6.41%
Figure 17c: Exhibit C-2 (abbreviated) Companies Ranked by Book Value of Equity: Comparative Risk Characteristics Data for Year Ending December 31, 2010 EXHIBIT C-2
Average Book Val. ($mils.) 39,141 12,811 Average Debt to Market Value of Equity 31.39% 39.63% Average Unlevered Risk Premium 4.33% 4.37% Average Unlevered Beta 0.64 0.64
Portfolio Rank by Size 1 2
///
25 60 31.9% 10.74% 0.99
///
25 60 1.27 11.98%
60
Derived as the average annual difference between SBBI Large Stock total returns (essentially the S&P 500 index) and SBBI income returns on long-term Treasury bonds over the time period 19632010. Source: Morningstar EnCorr Analyzer software.
Duff & Phelps
| 37
The Size Study
Unlevered Risk PremiaReconciliation of the A, B and C Exhibits Reconciliation of the levered and unlevered betas for use in CAPM (found in Exhibits B-2 and C-2, respectively) now reconcile with the levered and unlevered arithmetic average risk premia for the build-up (found in Exhibits A-2 and C-2, respectively), as demonstrated below using the values from the previous example: Levered risk premium = Levered beta x Historical market risk premium + Premium over CAPM (i.e. size premum) 11.98% = 1.27 x 4.40% + 6.41% Unlevered risk premium = Levered beta x Historical market risk premium + Premium over CAPM (i.e. size premum) 10.74% = 0.99 x 4.40% + 6.41% Relevering What if the debt-to-market-value-of-equity ratio (Wd /We ) of the subject company is different than the average (Wd /We ) of the companies making up Portfolio 25 (31.9%)? It may be possible to adjust the (levered) risk premiums over the risk free rate (RPm+s) from Exhibits A-1 through A-8 for differences in financial leverage between the subject company and the given guideline portfolio.61 Again, the average (levered) risk premia presented in Exhibits A-1 through A-8 are unlevered as follows: RPunlevered = RPlevered [(Wd /We ) x ( u d ) x RPm]
The unlevered risk premia in the C exhibits, which assume a firm is financed 100% with equity and 0% debt, are calculated by unlevering the average risk premia in the A exhibits. In the example, the unlevered risk premium over the risk free rate (RPm+s, unlevered) for Portfolio 25 in Exhibit C-2 (10.74%) was calculated by unlevering the average risk premium over the risk free rate (RPm+s) for Portfolio 25 in Exhibit A-2 (11.98%). This calculation was performed assuming the 31.9 percent average debt-to-market-value-of-equity ratio (Wd/We) of the companies making up Portfolio 25.62 The percentage of debt in the capital structure went from 31.9 percent to 0 percent, and the unlevered risk premia is lower than the levered risk premium. This formula can be rearranged to relever: RPlevered = RPunlevered + [(Wd /We ) x ( u d ) x RPm] If the subject company has a Wd /We ratio that is less (say 20%) than the average Wd /We of the guideline portfolio (31.9%), the unlevered risk premium may be relevered at the subject companys lower ratio: 11.52% = 10.74% + [(20%) x (0.99 0.1) x 4.4%] The subject company has less debt relative to equity than the average company in the guideline portfolio (20% versus 31.9%), and the relevered risk premium is lower than the average levered risk premium of the guideline portfolio (11.52% versus 11.98%). Generally, as the percentage of leverage (debt) in a companys capital structure decreases, risk to equity investors decreases (and vice versa).
61
62
If one relevers at a debt to equity (Wd/We) ratio different than the average of Wd/We of the given portfolio, other assumptions may not hold. For example, a debt beta of 0.1 is assumed in the unlevering calculations performed in the Report. If one relevers at a Wd/We ratio that is significantly higher than the average Wd/We ratio of the given guideline portfolio, a higher debt beta than 0.1 may be expected, all things held the same. As found in Exhibit C-2. It is important to understand that each of the A, B, and C exhibits is sorted by different size criteria. For instance, the base set of companies used to perform the analyses in the Size Study is sorted by book value of equity, and then used to calculate the different data and information presented in the A-2, B-2, and C-2 exhibits. Citing the present unlevering/relevering example, the average debt-to-marketvalue-of-equity ratio (Wd/We) of the smallest companies (Portfolio 25) as sorted by book value of equity is found in Exhibit C-2, while the average debt-to-market-value-of-equity ratio (Wd/We) of the smallest companies (Portfolio 25) as sorted by total assets is found in Exhibit C-5.
Duff & Phelps
| 38
The Risk Study
The Risk Study is an extension of the Size Study. The main difference between the Risk Study and the Size Study is that while the Size Study analyzes the relationship between size and return, the Risk Study analyzes the relationship between fundamental risk measures (based on accounting data) and return. These are called fundamental measures of company risk to distinguish these risk measures from a stock market-based measure of equity risk such as beta. A variety of academic studies have examined the relationship between financial statement data and various aspects of business risk.95 Research has shown that measures of earnings volatility can be useful in explaining credit ratings, predicting bankruptcy, and explaining the CAPM beta. As in the Size Study, 25 portfolios are created, but instead of being ranked by eight alternative measures of size as is done in the Size Study, the Risk Study portfolios are ranked by three fundamental risk measures: five-year average operating income margin, the coefficient of variation in operating income margin, and the coefficient of variation in return on book equity.96, 97 The first statistic measures profitability and the other two statistics measure volatility of earnings. All three measures use average financial data for the five years preceding the formation of annual portfolios. Size and Risk Traditionally, valuation professionals have used company size as a factor in determining discount rates for smaller companies. Small companies are believed to have higher required rates of return than large companies because small companies are inherently riskier. The historical data (as published in the Duff & Phelps Risk Premium Report, as well as in the SBBI), verify that small companies have, in fact, earned higher rates of return over long-run periods.
It has been pointed out in the financial literature that researchers may be mixing a size effect with a risk effect when measuring company size by market value,98 but market value is not just a function of size; it is also a function of the discount rate. In other words, some companies might be small because they are risky, rather than risky because they are small. The Risk Study goes beyond size and investigates the relationship between equity returns and fundamental risk measures. Does the evidence support the claim that smaller companies inherently have greater risk? The Risk Study analyzes this question, and demonstrates that as company size decreases, measures of risk calculated from financial statement data do, as a matter of fact, tend to increase.99 The data clearly shows that as fundamental risk increases in the form of lower profitability or greater variability of earnings, the return over the risk free rate tends to increase. These relationships are summarized in Figure 38.
Figure 38: Operating Margin (i.e. profitability) and Variability of Earnings versus Risk.
Operating Margin
Risk
Variability of Earnings
Risk
95 96 97 98 99
A survey of the academic research can be found in The Analysis and Use of Financial Statements, 3rd edition, White et al., Wiley (2003), chapter 18. Coefficient of variation is defined here as the standard deviation divided by the mean. For a detailed discussion of portfolio creation methodology, see Portfolio Methodology on page 9. A Critique of Size Related Anomalies, Jonathan Berk, Review of Financial Studies, vol. 8, no. 2 (1995). A similar point was made by Barry Goodman in a presentation at the October 1997 American Society of Appraisers Advanced Business Valuation Conference in San Francisco.
Duff & Phelps
| 39
The Risk Study
Previously, it was demonstrated in the Size Study that there is a clear inverse relationship between size and historical rates of return (as size decreases, returns tend to increase; see Graph 2 on page 18). In the Risk Study, the data show a clear direct relationship between accounting-data-based fundamental risk measures and historical rates of return (as fundamental risk increases, returns tend to increase). In Graph 3, as fundamental risk increases (from left to right), the average annual return over the study time horizon (19632010) tends to increase for each of the three fundamental risk measures. For example, in the 2011 Report, the average annual return of the portfolios made up of companies with the lowest risk as measured by each of the three fundamental risk measures was 13.2 percent, while the average annual return of the portfolios made up of companies with the highest risk as measured by each of the three fundamental risk measures was 20.6 percent.
Reasons for Using Fundamental Measures of Risk in Addition to Measures of Size First, certain measures of size (such as market value of equity) may be imperfect measures of the risk of a companys operations in some situations. For example, a company with a large and stable operating margin may have a small and unstable market value of equity if it is highly leveraged. In this case the risk of the underlying operations is low while the risk to equity is high. Second, while small size may indicate greater risk, some small companies may maintain near economic monopolies by holding a geographic niche or market niche such that their true riskiness is less than what would be indicated by their size.
Graph 3: Average Annual Return, Three Measures of Fundamental Risk 19632010 Average Annual Return
25% Operating Margin Coefcient of Variation of Operating Income 20% Coefcient of Variation of Return on Equity Average (all fundamental risk measures)
15%
10%
5% 1 5 10 Fundamental Risk (increasing from left to right) 15 20 25
Duff & Phelps
| 40
The Risk Study
Alternatively, while larger size (as measured by sales, for example) may indicate less risk, some companies may be riskier than the average of companies with similar sales. For example, assume the subject company was expecting to emerge from reorganization following bankruptcy. The risk premium appropriate for this company may be more accurately imputed from the pro-forma operating profit (after removing non-recurring expenses incurred during the bankruptcy) than from its size as measured by sales. In other words, the subject company may be riskier than companies with similar sales volumes. Use of fundamental accounting measures of risk allows for direct assessment of the riskiness of the subject company. For example, if the appropriate equity risk premium for the subject company when measuring risk by one or more fundamental risk measures is different than the equity risk premium based on size measures, this difference may be an indication of the company-specific risk of the subject company.100 Presentation of the Results The Risk Studys D exhibits present 25 portfolios ranked by three fundamental risk factors (based on accounting data). These fundamental risk factors are described in Table 4.101 Table 4: Three Measures of Fundamental Risk in the Risk Studys D Exhibits
Exhibits D-1 Operating Margin: The mean operating income for the prior five years divided by the mean sales for the prior five years. Operating income is defined as sales minus cost of goods sold plus selling, general, and administrative expenses plus depreciation. Note that this composite ratio is usually very close to a simple average of the annual ratios of operating income to sales, except in extreme cases generally involving companies with high growth rates. Exhibit D-2 Coefficient of Variation of Operating Margin: The standard deviation of operating margin over the prior five years divided by the average operating margin for the same years. Note that for calculating this coefficient, average operating margin is a simple average of the annual ratios of operating income to sales rather than the composite ratio used in exhibit D-1. Exhibit D-3 Coefficient of Variation of Return on Book Value of Equity: The standard deviation of return on book equity for the prior five years divided by the mean return on book equity for the same years. Return on book equity is defined as net income before extraordinary items minus preferred dividends divided by book value of common equity.
Each of the Risk Studys exhibits D-1 through D-3 displays one line of data for each of the 25 fundamental-risk-ranked portfolios. The D exhibits include the twelve statistics outlined in Table 5. For comparative purposes, the average returns from the SBBI series for large companies (essentially the S&P 500 Index), small companies, and long-term government bond income returns for the period 1963 through the latest year are also reported on each exhibit.102 Table 5: Statistics Reported for 25 fundamental-risk-ranked portfolios in the Risk Studys D Exhibits
The average of the sorting criteria for the latest year (e.g., the average operating margin for the latest five years before 2010). In the 2011 Report, the latest year is 2010. Note that the reported average risk statistics in exhibits D-1, D-2, and D-3 are not the same numbers as reported in exhibits C-1 through C-8. In exhibits C-1 through C-8, the reported statistics are calculated for portfolios of companies grouped according to size and are averages since 1963. In exhibits D-1, D-2, and D-3, the reported statistics are calculated for portfolios grouped according to risk, independent of the size of the companies, and are not averages since 1963 Log (base-10) of the average of the sorting criteria. The number of companies in each portfolio in the latest year. In the 2011 Report, the "latest year" is 2010. Geometric average historical equity return since 1963.
Arithmetic average historical equity return since 1963. Arithmetic average historical risk premium over long-term Treasuries (average return on equity in excess of long-term Treasury bonds) since 1963. (RPm+u) Unlevered arithmetic average historical risk premium over long-term Treasuries (average return on equity in excess of long-term Treasury bonds) since 1963. (RPm+u, unlevered) Smoothed average historical risk premium over long-term Treasuries (average return on equity in excess of long-term Treasury bonds) since 1963: the fitted premium from a regression with the historical risk premium over long-term Treasuries as dependent variable and the logarithm of the average sorting criteria as independent variable. (RPm+u) Average Debt as a percent of the MVIC since 1963.
Beta calculated using the sum beta method applied to monthly returns for 1963 through the latest year (see the 2011 SBBI Valuation Yearbook pp. 77-78 for a description of the sum beta method). Unlevered beta calculated using the "sum beta" method applied to monthly returns for 1963 through the latest year.
Standard deviation of annual historical equity returns.
100
101 102
Valuing a Business, 4th ed., Pratt et al, McGraw-Hill (2000), p 181. Company-specific risk factors can include concentration of customer base, key person dependence, key supplier dependence, or any number of other factors that are unique to the subject company. For a detailed description of the Standard and Poors Compustat data items used in the Duff & Phelps Risk Premium Report, please see Appendix A (in the full Report). Source: Morningstar EnCorr Analyzer software.
Duff & Phelps
| 41
The Risk Study
Overview of Methods Used to Estimate Cost of Equity Capital using the Risk Study The Risk Study provides one method of estimating COE for a subject company, Buildup 3, plus one method for estimating unlevered COE (the cost of equity capital assuming a firm is financed 100% with equity and 0% debt).
These methods are summarized below in equation format, and summarized in Figure 39 in graphical building blocks format. 1) Buildup 3 COEBuildup 3 = (Risk Free Rate) + (Risk Premium Over Risk Free Rate) + (Equity Risk Premium Adjustment) Example 5a: using Guideline portfolios: page 46 Example 5b: using regression equations: page 49 2) Buildup 3-Unlevered COEBuildup 3-Unlevered = (Risk Free Rate) + (Unlevered Risk Premium Over Risk Free Rate) + (Equity Risk Premium Adjustment) Example 6: using Guideline portfolios: page 74 (in the full Report)
Figure 39: Two Methods of Estimating Cost of Equity Capital with the Risk Study103 Buildup 3 Buildup 3-Unlevered
+ ERP Adjustment* Cost of Equity
+ ERP Adjustment* Cost of Equity
+ Smoothed Risk Premium Over Risk Free Rate, RPm+u Risk Free Rate, Rf (Use Exhibit D risk premia)
+ Smoothed Unlevered Risk Premium Over Risk Free Rate, RPm+u, unlevered Risk Free Rate, Rf (Use Exhibit D risk premia)
*ERP Adjustment: The difference between the historical (19632010) equity risk premium (ERP) and a user of the Duff & Phelps Reports own forward ERP estimate: ERP Adjustment = Users ERP Historical (19632010) ERP The ERP Adjustment is made only in the Buildup 1, Buildup1Unlevered, Buildup 1-High-Financial-Risk, Buildup 3, and Buildup 3-Unlevered methods. Please refer to the individual examples provided for these models for more information.
NOTE: This section includes an example of using the Reports risk premia data to estimate cost of equity capital using the Buildup 3 method, plus an overview of the C data exhibits, which provide a link between the Size Study and the Risk Study. The C exhibits can be used to gauge company-specific risk adjustments. A complete example for using the Reports risk premia to estimate cost of equity capital using the Buildup 3-Unlevered method is available in the full version of the 2011 Report.
103
The relative sizes of the building blocks in Figure 39 do not necessarily represent the relative size of the various inputs. Also note that the names given to the models in the Risk Premium Report (e.g. Buildup 1, Buildup 2, Buildup 3, etc) are naming conventions used within the Report to make referring to the different methods easier.
Duff & Phelps
| 42
The Risk Study
The three risk measures outlined in Table 4 (page 41) can be used with either of the two methods of estimating COE provided by the Risk Study. It is important to note that the subject company information necessary to calculate all of these measures may not be available. In these cases, it is generally acceptable to use the fundamental risk measures that are available. It is recommended, however, that Report users calculate available risk measures for the subject company using at least the three most recent years of data, and the five most recent years of data for best results. Gathering Accounting Information to Calculate Fundamental Risk Measures The first step in using the Risk Study to estimate cost of equity capital (COE) is to gather the accounting-based information for the subject company needed to calculate the three fundamental risk measures analyzed in the Risk Study. y To calculate operating margin and coefficient of variation of operating margin, net sales and operating income are needed. y To calculate coefficient of variation of ROE, book value and net income before extraordinary items are needed. The accounting information for the last 5 years needed to calculate the three fundamental risk measures for a hypothetical subject company is summarized in Figure 40a and Figure 40b.
Figure 40a: Subject Company Operating Margin and Coefficient of Variation of Operating Margin (used in all examples)
2010 Net Sales Operating Income Operating Margin Standard Deviation of Operating Margin Average Operating Margin Coefficient of Variation of Operating Margin $900 $150 16.7% = $150/$900 2.3% 2009 $800 $120 15.0% = $120/$800 2008 $850 $130 15.3% = $130/$850 2007 $750 $80 10.7% = $80/$750 2006 $900 $140 15.6% = $140/$900
14.6%
15.8% = 2.3%/14.6%
Figure 40b: Subject Company Coefficient of Variation of ROE (used in all examples)
2010 Book Value Net Income before extraordinary items Return on Book Equity (ROE) Standard Deviation of ROE Average ROE Coefficient of Variation of ROE $820 $110 2009 $710 $80 2008 $630 $90 2007 $540 $40 2006 $500 $100
13.4% = $110/$820 4.6%
11.3% = $80/$710
14.3% = $90/$630
7.4% = $40/$540
20.0% = $100/$500
13.3% 34.7% = 4.6%/13.3%
Duff & Phelps
| 43
The Risk Study
The hypothetical subject company has an average operating margin of 14.6 percent, a coefficient of variation of operating margin of 15.8 percent, and a coefficient of variation of ROE of 34.7 percent, as summarized in Figure 41.104 Figure 41: Subject Company Fundamental Risk Characteristics (used in all Examples)
Appropriate Exhibit Risk Measure Average Operating Margin Coefficient of Variation of Operating Margin Coefficient of Variation of ROE 14.6% 15.8% 34.7% Buildup 3 D-1 D-2 D-3 Buildup 3-Unlevered D-1 D-2 D-3
Estimating Cost of Equity Capital Using the Buildup 3 Method Buildup 3 + ERP Adjustment + Smoothed Risk Premium Over Risk Free Rate, RPm+u Risk Free Rate, Rf (Use Exhibit D risk premia) Cost of Equity
Figure 41 also includes the data exhibits in which the appropriate risk premia for each of the size measures can be found. For example, for use in the Buildup 3 method, risk premia over the risk free rate (RPm+u) for coefficient of variation of operating margin are found in Exhibit D-2. For use in the Buildup 3-Unlevered method, unlevered risk premia over the risk free rate (RPm+u, unlevered) for coefficient of variation of operating margin are also found in Exhibit D-2. In each of the following examples of using the Risk Study to estimate COE, the subject company risk measures summarized in Figure 41 will be used (average operating margin of 14.6 percent, for instance, will be used in all examples).
The buildup method is an additive model commonly used for calculating the required rate of return on equity. As the name implies, successive building blocks are summed, each representing the additional risk inherent to investing in alternative assets. An example of this is the extra return (i.e. premium), that investors demand for investing in stocks versus investing in a riskless security.105 Risk Premia Over Risk Free Rate, RPm+u The risk premia developed in the Risk Study (RPm+u) take the form of risk premia over the risk free rate, but are slightly different from the risk premia over the risk free rate (RPm+s) that are developed in the Size Study, which are a measure of risk in terms of the total effect of market risk and size risk.106 Because operating efficiencies (or lack thereof) of the subject company are being captured by the use of accounting-based risk measures, the difference in the average rate of return for each risk-based portfolio over the sample period and the income return earned of long-term Treasury bonds (using SBBI data) is a measure of risk in terms of the total effect of market risk, and company-specific risk (RPm+u).107 The result is a clear direct relationship between fundamental risk and premium over long-term bond yields. As fundamental risk increases, the return over the risk free rate (i.e. excess return) tends to increase. The RPm+u risk premia can be added to the risk free rate (Rf) to estimate cost of equity capital (COE) using the Buildup 3 method.
104
105 106 107
Coefficient of variation is defined here as the standard deviation divided by the mean. For example (using a Microsoft Excel formula), the coefficient of variation of operating margin of the hypothetical subject company used in all examples = STDEV(16.7,15.0,15.3,10.7,15.6)/AVERAGE(16.7,15.0,15.3,10.7,15.6). Throughout this report the risk free asset is represented by the yield on a 20-year constant maturity Treasury bond. For a detailed discussion of how premia over the risk free rate are calculated, see The Difference Between Risk Premia Over the Risk Free Rate and Risk Premia Over CAPM on page 22. Because these premia have an embedded measure of market (i.e. beta) risk, these premia are appropriate for use in buildup methods that do not otherwise include a measure of market risk, but are not appropriate for use in models (e.g. CAPM) that already have a measure of market (beta) risk. Risk Study risk premia over the risk free rate (RPm+u) are published in Exhibits D-1, D-2, and D-3 of the Risk Premium Report.
Duff & Phelps
| 44
The Risk Study
The Buildup 3 Equation As an alternative to the basic buildup equation (see page 29 in the full Report), one can use the Risk Study to develop a risk premium for the subject company for which RPm (the market premium) and RPu (the company-specific risk premium) are combined into a single premium, RPm+u. The basic buildup equation therefore becomes: E(Ri) = Rf + RPm+u where: E(Ri) Rf RPm+u = Expected rate of return on security i (this is cost of equity capital, or COE) = risk free rate as of the valuation date (typically a long-term US Treasury bond yield) = risk premium for the subject company for which RPm (the market premium) and RPu (the company-specific risk premium) are combined into a single premium.
Adding the ERP Adjustment to the basic buildup formula produces the full equation for the Buildup 3 method: COEBuildup 3 = Rf + RPm+u + ERP Adjustment The Buildup 3 method is a straightforward way of estimating cost of equity capital (COE) using the historical risk premiums over the long-term risk-free rate (RPm+u) presented in exhibits D-1 through D-3. It is important to understand that because the risk premia presented in the D exhibits have an embedded measure of market (i.e. beta) risk, they are appropriate only for use in buildup methods that do not otherwise include a measure of market risk; these premia are not appropriate for use in models (e.g. CAPM) that already have a measure of market (beta) risk.112 As noted previously, the Risk Premium Report provides two ways for analysts to match their subject companys size (or risk) characteristics with the appropriate smoothed premia from the data exhibits: the guideline portfolio method and the regression equation method.113 In general, the regression equation method is preferred because this method allows for interpolation between the individual guideline portfolios, although the guideline portfolio method is less complicated, and more direct. Examples of both the guideline portfolio method and the regression equation method follow, starting with the simpler guideline portfolio method.
One final important modification of the basic buildup formula is needed: the equity risk premium (ERP) adjustment. The equity risk premium adjustment is made to reconcile the historical data presented in the Risk Premium Report with the forward-looking ERP chosen by the individual analyst as of valuation date.108 The ERP Adjustment is simply the difference between the users own forward-looking ERP and the historical 19632010 ERP (4.40%).109 For example, many users of the Report use the Duff & Phelps Recommended ERP, which was 5.5 percent at the end of 2010).110, 111 In this case, the ERP Adjustment would be 1.1 percent (5.5%4.4%).
108
109
110
111
112 113
The ERP Adjustment is necessary in the Size Studys Buildup 1 method and Buildup 1-Unlevered method, and in the Risk Studys Buildup 3 method and Buildup 3-Unlevered method. See page 30 in the full Report for more a detailed discussion of the equity risk premium adjustment. Calculated as the annual SBBI Large Company Stock (essentially the S&P 500 Index) return minus the average annual long-term SBBI government bond income return over the time horizon 19632010. Source: Morningstar EnCorr Analyzer software. For more information on the equity risk premium, see Cost of Capital: Applications and Examples 4th ed., by Shannon P. Pratt and Roger J. Grabowski (John Wiley & Sons, Inc., 2010), Chapter 9, Equity Risk Premium, pages 115158. See Roger J. Grabowski, Developing the Cost of Equity Capital: Risk-Free Rate and ERP During Periods of Flight to Quality. This paper will appear in the Business Valuation Review and can also be downloaded at Duff & Phelps Cost of Capital site at www.duffandphelps.com/CostofCapital. Please refer to page 28 for examples illustrating how to use size premia in conjunction with CAPM to estimate COE. See pages 1416 for a detailed explanation of the differences between the guideline portfolio method and the regression equation method.
Duff & Phelps
| 45
The Risk Study
Example 5a: Buildup 3 Method (using guideline portfolios) Three pieces of information are needed to estimate the cost of equity capital using the Buildup 3 method using guideline portfolios: a risk free rate (Rf), a risk premium over the risk free rate (RPm+u), and an ERP Adjustment (if necessary). All of the information needed is summarized in Figure 42. Figure 42: Information Needed to Estimate COE Using Buildup 3 and Guideline Portfolios
Step 1 Step 2 Step 3 Step 4
Step 1, Risk Free Rate (Rf): The risk free rate is typically a long-term US Treasury bond yield as of the valuation date. This example utilizes the assumed long-term treasury yield of 4.1 percent established in Example 1a (on page 31 in the full Report). Step 2, Risk Premium Over Risk Free Rate (RPm+u): Match the various fundamental risk measures of the subject company with the guideline portfolios composed of companies of similar fundamental risk measures in Exhibits D-1 through D-3, and identify the corresponding smoothed average risk premium. The subject company in this example has an average operating margin of 14.6 percent, and the appropriate data exhibit is Exhibit D-1 (see Figure 41 on page 44). An abbreviated version of Exhibit D-1 is shown in Figure 44. Of the 25 portfolios, the portfolio that has an average operating margin closest to the subject companys 14.6 percent is Portfolio 9 (14.37%). The corresponding smoothed average risk premium (RPm+u) is 8.72 percent (8.7%, rounded).
Rf
RPm+u
(using guideline portfolios)
ERP Adj.
COE
This example utilizes the long-term risk free rate (Rf) and the ERP Adjustment established in a previous example (the Size Studys Buildup 1 method using guideline portfolios; see page 31 in the full Report). This mirrors the fact that for any given valuation engagement the same risk free rate and ERP will generally be used in each of the models presented by the individual analyst. Please note that for any given valuation engagement these inputs may (and probably will) be different than the ones used in the examples. The only missing ingredients needed to estimate COE are the risk premia over the risk free rate (RPm+u), as summarized in Figure 43.
Figure 43: NeededSmoothed Risk Premia Over the Risk Free Rate (RPm+u) Using Guideline Portfolios
Step 1 Step 2 Smoothed Premium Over Risk Free Rate, (RPm+u) + + + ? ? ? + + + Step 3 Step 4
Risk Measure Operating Margin Coefficient of Variation of Operating Margin Coefficient of Variation of ROE 14.6% 15.8% 34.7%
Appropriate Exhibit D-1 D-2 D-3
Guideline Portfolio ? ? ?
Risk Free Rate, Rf 4.1% 4.1% 4.1%
ERP Adjustment 1.1% 1.1% 1.1% = = =
COE
4.1% 4.1%
+ +
+ +
1.1% 1.1%
= =
Duff & Phelps
| 46
The Risk Study
Match all of the subject companys risk measures in this fashion. For example, the subject company in this example has a coefficient of variation of operating margin of 15.8 percent. Of the 25 guideline portfolios in Exhibit D-2 (not shown here), the portfolio that has a coefficient of variation of operating margin closest to the subject companys 15.8 percent coefficient of variation of operating margin is Portfolio 13 (16.2%). The corresponding smoothed average risk premium is 9.5 percent. In the case of the third risk measure, the subject company has a coefficient of variation of ROE of 34.7 percent. Of the 25 guideline portfolios in Exhibit D-3 (not shown here), the portfolio that has a coefficient of variation of ROE closest to the subject companys 34.7 percent coefficient of variation of ROE is Portfolio 14 (33.7%). The corresponding smoothed average risk premium is 9.5 percent.
At this point, all of the available risk measures for the subject company have been matched to the closest guideline portfolio in the appropriate exhibit, and the corresponding smoothed average risk premium has been identified for each, and Step 2 is complete. Step 3, Equity Risk Premium (ERP) Adjustment: The ERP Adjustment is needed to account for any difference in the users own ERP estimate and the historical (19632010) ERP. This example utilizes the ERP Adjustment (1.1%) established in Example 1a (page 31 in the full Report).
Figure 44: Exhibit D-1 (abbreviated) Companies Ranked by Operating Margin Historical Equity Risk Premium: Average Since 1963 Data for Year Ending December 31, 2010
Arithmetic Average Unlevered Risk Premium 5.32%
Portfolio Rank 1 2 3 4 5 6 7 8 9 10
Average Operating Margin 38.99% 29.61% 25.17% 22.02% 19.69% 17.89% 16.85% 15.79% 14.37% 13.43%
Log of Average Op Margin -0.41 -0.53 -0.60 -0.66 -0.71 -0.75 -0.77 -0.80 -0.84 -0.87
Number as of 2010 57 58 51 58 52 53 47 52 57 52
Beta (SumBeta) Since 63 0.87 0.82 0.85 0.94 0.98 1.06 1.10 1.11 1.16 1.16
Standard Deviation of Returns 17.00% 17.51% 17.43% 16.98% 18.47% 17.88% 19.22% 20.07% 20.22% 20.80%
Arithmetic Average Return 13.08% 11.37% 13.50% 13.09% 14.61% 14.31% 15.06% 14.84% 16.59% 15.78%
Arithmetic Average Risk Premium 6.17%
Smoothed Average Risk Premium 4.98%
Average Debt/MVIC 25.51% 28.26% 26.83% 23.23% 19.85% 17.37% 17.81% 18.39%
9.69%
8.78%
8.72%
19.37% 20.21%
///
24 25 3.78% 2.18% -1.42 -1.66 81 113 1.29 1.29 26.33% 28.90% 20.13% 20.62% 30.84% 30.57%
Note: Some values intentionally blurred.
Duff & Phelps
| 47
The Risk Study
Step 4, Estimate Cost of Equity (COE): With the completion of Steps 1 through 3, the information needed to estimate a base cost of equity capital using the Buildup 3 method (using guideline portfolios) is now completed. The risk premiums over the risk free rate (RPm+u) can be added to the risk-free rate (Rf) and the ERP Adjustment to estimate an indicated cost of equity capital (COE) for the subject company, as illustrated in Figure 45.
The range of COE estimates for the hypothetical subject company in this example is 13.9 percent to 14.7 percent, with an average of 14.4 percent, and a median of 14.7 percent. The mean represents the average estimate, but the mean can be unduly influenced by very large or very small outliers. For this reason, the median estimate is generally preferred to the mean. The median estimate tends to not be as heavily influenced by very large or very small outliers, and can be considered a measure of the typical estimate in the group. Use of fundamental accounting measures of risk allows for direct assessment of the riskiness of the subject company. For example, if the appropriate equity risk premium for the subject company when measuring risk by one or more fundamental risk measures is different than the equity risk premium based on size measures, this difference may be an indication of the company-specific risk of the subject company.114
Figure 45: Buildup 3 COE Inputs (using guideline portfolios)

Step 1 Step 2 Smoothed Premium Over Risk Free Rate, (RPm+u) + + + 8.7% 9.5% 9.5% + + + Step 3 Step 4
Guideline Portfolio 9 13 14
ERP Adjustment 1.1% 1.1% 1.1% = = =
COE 13.9% 14.7% 14.7%
4.1% 4.1%
+ +
9.2% 9.5%
+ +
1.1% 1.1%
= =
14.4% 14.7%
114
Valuing a Business, 4th ed., Pratt et al, McGraw-Hill (2000), p 181. Company-specific risk factors can include concentration of customer base, key person dependence, key supplier dependence, or any number of other factors that are unique to the subject company.
Duff & Phelps
| 48
The Risk Study
Example 5b: Buildup 3 Method (using regression equations) When the subject company risk measures do not exactly match the respective average risk measure of the guideline portfolios, the data exhibits provide a straightforward way to interpolate an exact risk premium over the risk free rate between guideline portfolios using the regression equation method. The only difference between estimating cost of equity capital (COE) using the Buildup 3 method using guideline portfolios (as in the previous example) and COE using the Buildup 3 method using regression equations is how the risk premia over the risk free rate (RPm+u) are identified in Step 2. Figure 46: Steps Needed to Estimate COE Using Buildup 3 and Regression Equations
In the previous example, the smoothed average risk premia published in the report for the appropriate guideline portfolios were used to estimate COE.115 In this example, however, the regression equations found in each of the data exhibits will be used to calculate custom interpolated risk premia, based upon the specific risk measures of the subject company. Please note that this example utilizes the long-term risk free rate (Rf) and the ERP Adjustment established in a previous example (the Size Studys Buildup 1 method using guideline portfolios; see page 31 in the full Report). This mirrors the fact that for any given valuation engagement the same risk free rate and ERP will generally be used in each of the models presented by the individual analyst. Please note that for any given valuation engagement these inputs may (and probably will) be different than the ones used in the examples. The only missing ingredients needed to estimate COE are the risk premia over the risk free rate (RPm+u), as summarized in Figure 47.
Rf
RPm+u
ERP Adj.
COE
Figure 47: Buildup 3 COE Inputs (using regression equations)

Step 1 Step 2 Smoothed Premium Over Risk Free Rate, (RPm+u) + + + ? ? ? + + + Step 3 Step 4
ERP Adjustment 1.1% 1.1% 1.1% = = =
COE
4.1% 4.1%
+ +
+ +
1.1% 1.1%
= =
115
The smoothed risk premia published in the Risk Premium Report are based upon the average size (or risk) measure of each of the respective guideline portfolios.
Duff & Phelps
| 49
The Risk Study
Step 1, Risk Free Rate (Rf): The risk free rate is typically a long-term US Treasury bond yield as of the valuation date. This example utilizes the assumed long-term treasury yield of 4.1 percent established in Example 1a (page 31 in the full Report). Step 2, Risk Premium Over the Risk Free Rate (RPm+u): The hypothetical subject company in this example has an average operating margin of 14.6 percent, and the appropriate data exhibit is Exhibit D-1 (see Figure 41)116. In this case one would expect that the smoothed average risk premium over the risk free rate (RPm+u) would fall somewhere between 8.36 percent (the smoothed risk premium over the risk free rate for Portfolio 8) and 8.72 percent (the smoothed risk premium over the risk free rate for Portfolio 9), as illustrated in Figure 48:
An easy way to calculate a custom interpolated risk premium over the risk free rate (RPm+u) in between Portfolio 8 and Portfolio 9 is by using the regression equations provided for this purpose in each of the data exhibits. The regression equations are located in the same spot in each of the data exhibits (see Figure 4 on page 16).117 The regression equation provided in Exhibit D-1, which includes 25 portfolios ranked by operating margin118, is: Smoothed Premium = 1.453% 8.622% * Log(Operating Margin) To calculate an interpolated risk premium for the subject company, substitute the subject companys 14.6 percent operating margin into the regression equation as follows119: Smoothed Premium = 1.453% 8.622% * Log(14.6%) 8.6% = 1.453% 8.622% * (-0.84)
Figure 48: Exhibit D-1 (abbreviated) Companies Ranked by Operating Margin Historical Equity Risk Premium: Average Since 1963 Data for Year Ending December 31, 2010
Average Operating Margin 38.99% 29.61% 25.17% 22.02% 19.69% 17.89% 16.85% 15.79% 14.60% 14.37% 13.43% -0.84 -0.87 57 52 1.16 1.16 20.22% 20.80% 16.59% 15.78% 9.69% 8.78% Log of Average Op Margin -0.41 -0.53 -0.60 -0.66 -0.71 -0.75 -0.77 -0.80 Number as of 2010 57 58 51 58 52 53 47 52 Beta (SumBeta) Since 63 0.87 0.82 0.85 0.94 0.98 1.06 1.10 1.11 Standard Deviation of Returns 17.00% 17.51% 17.43% 16.98% 18.47% 17.88% 19.22% 20.07% Arithmetic Average Return 13.08% 11.37% 13.50% 13.09% 14.61% 14.31% 15.06% 14.84% 7.94% 7.12% 8.36% ? 8.72% 19.37% 20.21% Arithmetic Average Risk Premium 6.17% Arithmetic Average Unlevered Risk Premium 5.32% Smoothed Average Risk Premium 4.98%
Portfolio Rank 1 2 3 4 5 6 7 8 Subject Company 9 10
Average Debt/MVIC 25.51% 28.26% 26.83% 23.23% 19.85% 17.37% 17.81% 18.39%
///
24 25 3.78% 2.18% -1.42 -1.66 81 113 1.29 1.29 26.33% 28.90% 20.13% 20.62% 30.84% 30.57%
Note: Some values intentionally blurred.

116 117
118 119
The same three risk measures (for a hypothetical subject company) are used in all examples of estimating COE using the Risk Study, as outlined in Figure 41 on page 44. In addition to regression equations for interpolating risk premia between guideline portfolios in the Risk Studys D exhibits, the Size Studys A and B exhibits also provide regression equations for easy interpolation of risk premia between guideline portfolios, as do the C exhibits (for unlevered A exhibit risk premia). Please note that each exhibit has a different regression equation. The logarithmic relationship is base-10, and that the risk data is in percent, such that the log of 10 percent is log(10%), and not log(10). The formula to calculate a values base-10 logarithm in Microsoft Excel is =log(value). Also note that the * used in the regression equation is the symbol used in Microsoft Excel to denote the multiplication symbol, x. The * format is used to denote multiplication in the regression equations in the data exhibits.
Duff & Phelps
| 50
The Risk Study
Interpolate smoothed risk premium for each fundamental risk measure available for the subject company using the regression equations from the data exhibits. For example, the subject company in this example has a coefficient of variation of operating margin of 15.8 percent. The regression equation provided in Exhibit D-2 is: Smoothed Premium = 13.505% + 5.064% * Log(CV Op. Margin) The interpolated smoothed risk premium is therefore 9.4 percent (13.505% + 5.064% * (-0.80)). In the case of the third risk measure, the subject company has a coefficient of variation of ROE of 34.7 percent. The regression equation provided in Exhibit D-3 is: Smoothed Premium = 10.632% + 2.441% * Log(CV ROE) The interpolated smoothed risk premium is therefore 9.5 percent (10.632% + 2.441% * (-0.46)). After interpolating smoothed risk premia (RPm+u) for the subject companys available risk measures, Step 2 is complete. Step 3, Equity Risk Premium (ERP) Adjustment: The ERP Adjustment is needed to account for any difference in the analysts own ERP estimate and the historical (19632010) ERP. This example utilizes the ERP Adjustment (1.1%) established in Example 1a (page 31 in the full Report).
Step 4, Estimate Cost of Equity (COE): With the completion of Steps 1 through 3, the information needed to estimate a base cost of equity capital using the Buildup 3 method (using regression equations) is now completed. The risk premiums over the risk free rate (RPm+u) can be added to the risk-free rate (Rf) and the ERP Adjustment to estimate an indicated cost of equity capital (COE) for the subject company, as illustrated in Figure 49: The range of COE estimates for the hypothetical subject company in this example is 13.8 percent to 14.6 percent, with an average of 14.4 percent, and a median of 14.6 percent. The mean represents the average estimate, but the mean can be unduly influenced by very large or very small outliers. For this reason, the median estimate is generally preferred to the mean. The median estimate tends to not be as heavily influenced by very large or very small outliers, and can be considered a measure of the typical estimate in the group. Use of fundamental accounting measures of risk allows for direct assessment of the riskiness of the subject company. For example, if the appropriate equity risk premium for the subject company when measuring risk by one or more fundamental risk measures is different than the equity risk premium based on size measures, this difference may be an indication of the company-specific risk of the subject company.120
Figure 49: Buildup 3 COE Inputs (using regression equations)

Step 1 Step 2 Smoothed Premium Over Risk Free Rate, (RPm+u) + + + 8.6% 9.4% 9.5% + + + Step 3 Step 4
ERP Adjustment 1.1% 1.1% 1.1% = = =
COE 13.8% 14.6% 14.7%
4.1% 4.1%
+ +
9.2% 9.4%
+ +
1.1% 1.1%
= =
14.4% 14.6%
120
Valuing a Business, 4th ed., Pratt et al, McGraw-Hill (2000), p 181. Company-specific risk factors can include concentration of customer base, key person dependence, key supplier dependence, or any number of other factors that are unique to the subject company.
Duff & Phelps
| 51
The Risk Study
The C Exhibits Provide a Link Between the Size Study and the Risk Study The C exhibits provide an important functionthey serve as a link between the Size Study and the Risk Study. These exhibits can be used to compare a subject companys fundamental risk characteristics to the fundamental risk characteristics of portfolios made up of similarly-sized companies. Specifically, the C exhibits can be used to compare the subject companys fundamental risk characteristics to the fundamental risk characteristics of portfolios made up of companies of similar size.
Size Study Risk Study
Is Size Correlated with Market and Fundamental Risk Measures? To calculate the statistics included in Exhibits C-1 through C-8, the fundamental risk characteristics are calculated for the same sizeranked portfolios that are created in the Size Study.128 For example, Exhibit A-1 is comprised of 25 portfolios ranked by market value of equity. To calculate the fundamental risk characteristics found in Exhibit C-1, the three fundamental risk measures used to rank the portfolios in the Risk Study (five-year operating income margin, the coefficient of variation in operating income margin, and the coefficient of variation in return on book equity) are calculated for each of the 25 (market-value-of-equity-ranked) portfolios in Exhibit A-1. These calculations are then made in the same fashion for each of the 25 size-ranked portfolios created for exhibits A-2 through A-8 (e.g. for each of the 25 portfolios ranked by book value of equity in Exhibit A-2, the three fundamental risk measures are calculated; then for each of the 25 portfolios ranked by 5-year average net income in Exhibit A-3, the three fundamental risk measures are calculated, etc.). It is important to understand that the 25 portfolios used to calculate the fundamental risk statistics included in the C exhibits are different from the 25 portfolios used to calculate the fundamental risk statistics included in the D exhibits. In the former case, the portfolios are ranked by each of eight alternative measures of size, and then the fundamental risk characteristics of each portfolio are calculated. In the latter case, the large base set of companies that the analyses of the Report begins with are ranked by each of the three fundamental risk measures to form 25 risk-ranked portfolios, and then the average risk characteristics of each portfolio are calculated. For example, if 10 companies were ranked by size, the order (from largest to smallest) may be quite different from the same 10 companies ranked by operating margin (from highest to lowest).
C
Exhibits C-1 through C-8 also provide unlevered versions of the risk premia over the risk free rate found in the A exhibits (RPm+s). These unlevered premia (RPm+s, unlevered) are used in Examples 2a and 2b (see page 40 and 44, respectively, in the full Report) to estimate cost of equity capital assuming a firm is financed 100% with equity and 0% debt.126, 127
126
127 128
The D exhibits also include unlevered risk premia, but these are unlevered versions of the corresponding levered risk premia found in the D exhibits. The unleverered premia in the C exhibits are unlevered versions of the corresponding levered risk premia found in the A exhibits. The unlevered risk premia over the risk free rate found in the D exhibits (RPm+u, unlevered) are used in example 6 (see page 74 in the full Report) to estimate cost of equity capital using Risk Study inputs. Exhibits A-1 through A-8 and B-1 through B-8 use the same respective size-ranked portfolios, but calculate different statistics for each exhibit. For example, the 25 portfolios ranked by book value of equity are used in Exhibit A-2 and Exhibit B-2, but risk premia over the risk free rate (RPm+s) for use in a buildup method are calculated for Exhibit A-2, while risk premia over CAPM (RPs, or size premia) for use in CAPM and Buildup 2 are calculated for Exhibit B-2.
Duff & Phelps
| 52
The Risk Study
The data suggests that size is correlated with market measures. For example, as size measures decrease in Graph 4 (from left to right), the beta (both levered and unlevered) of the portfolios increase (as expected).129 Graph 4: Average Levered and Unlevered Sum Beta (all eight size measures) 19632010 Sum Beta
1.4% 1.2% 1 .0% 0.8% 0.6% 0.4% 0.2% 0.0% 0 5 10 15 Portfolio (1 = Largest, 25 = Smallest) 20 25
premium resulting from the size effect), the Report characterizes the Risk Studys risk premia over the risk free rate (RPm+u) as being a measure of risk in terms of the total effect of market risk and companyspecific risk (also known as unsystematic risk). Graph 5a: Average Operating Margin (all eight size measures) 19632010 Average Operating Margin (all eight size measures)
14% 12% 10% 08% 06% 04% 02% 0% 0 5 10 15 Portfolio (1 = Largest, 25 = Smallest) 20 25
Average Operating Margin (all eight size measures)
Average Levered Sum Beta (all eight size measures) Average Unlevered Sum Beta (all eight size measures)
Average CV(Operating Margin), Average CV(ROE)
The data also suggests that this correlation extends to the three fundamental measures of risk. For example, in Graph 5a, as size measures decrease (from left to right), operating margin of the portfolios decreases (indicating increased risk), and in Graph 5b, as size measures decrease (from left to right), average coefficient of operating margin and average coefficient of variation of ROE of the portfolios increase (indicating increased risk). While the correlation between fundamental measures of risk and size clearly demonstrated in Graph 5a and Graph 5b implies that there may be an embedded size effect component in the Risk Studys RPm+u premia, the magnitude of this embedded size effect is difficult to quantify. In any case, the size effect embedded in the Risk Studys RPm+u premia are in all likelihood not equivalent to the size effect embedded in the Size Studys RPm+s premia, which are a measure of risk in terms of the total effect of market risk and size risk. To avoid confusion between the two premia, and because the operating efficiencies (or lack thereof) of the subject company being captured by the use of accounting-based risk measures may offset the risk premium resulting from the size effect (or increase the risk
129 130
Graph 5b: Average Coefficient of Operating Margin and Average Coefficient of Variation of ROE (all eight size measures)
60% 50% 40% 30% 20% 10% 0% 0 5 10 15 Portfolio (1 = Largest, 25 = Smallest) 20 25
Average CV(ROE) (all eight size measures) Average CV(Operating Margin) (all eight size measures)
Generally, the three fundamental measures of risk display increasing risk as size decreases, as the historical unlevered risk premium increases and as the unlevered beta increases.130, 131
131
In the research on size as reported in this report, we have determined that, in the context of the CAPM, the higher betas of the small companies explain some but not all of the higher average historical equity returns in these portfolios. Were one to calculate the respective correlations, those statistics would relate average portfolio statistics (e.g. average size vs. average risk) rather than correlation statistics across individual companies. At the individual company level, the correlations are much lower. There are two notable exceptions to this pattern: Exhibit C-7 indicates that there is little differentiation in operating margin as size (as measured by sales) changes, and Exhibit C-8 indicates that there is little differentiation in operating margin as size (as measured by number of employees) changes. In both cases, however, the coefficient of variation of operating margin and the coefficient in variation of return on book equity indicate increasing risk as size (as measured by sales and number or employees) decreases, as in the other exhibits.
Duff & Phelps
| 53
The Risk Study
The C Exhibits and Company-Specific Risk The use of a portfolios average historical rate of return to calculate a discount rate is based (in part) upon the implicit assumption that the risks of the subject company are quantitatively similar to the risks of the average company in the subject portfolio. If the risks of the subject company differ materially from the average company in the subject portfolio, then an appropriate discount rate may be lower (or higher) than a return derived from the average premium for a given portfolio. The data reported in exhibits C-1 through C-8 (where risk statistics are reported for each size category) may be helpful in making such a determination. For example, assume that the size of the subject company based on the 5-year average net income equals $20 million, placing it in the 23rd portfolio of exhibit A-3. The corresponding smoothed average risk premium over the risk free rate (RPm+s) from exhibit A-3 is 11.5 percent. The average operating margin for companies in the 23rd size-ranked portfolio equals 8.5% (see exhibit C-3, portfolio 23). If the subject companys operating margin is, say, 15.8%, it may be less risky than companies of similar size. We can examine the relative operating margins to estimate an appropriate company-specific risk premium (positive or negative) that adjusts for the differences in the subject company and the typical company with a 5-year average net income equal to $20 million. In exhibit D-1, for example, companies with average operating margins of approximately 15.8% make up the 8th (operating-margin-ranked) portfolio. The corresponding smoothed average risk premium over the risk free rate (RPm+u) from exhibit D-1 is 8.4 percent, which is significantly less than the 11.5 percent RPm+s premia from exhibit A-3. One can use the C exhibits to test each of the size measures to compare the fundamental risk characteristics of the subject company to the fundamental risk characteristics of portfolios made up of companies of similar size. If two or more (out of three) of the tests indicate that the subject company is less (or more) fundamentally risky than portfolios made up of companies of similar size, this can be a powerful argument in defending the downward (or upward) direction of a company-specific risk adjustment. Comparison of the subject companys fundamental risk characteristics to the corresponding risk premia found in the D exhibits can be helpful in estimating the degree of company-specific risk premium (positive or negative) that adjusts for the differences in the fundamental risk of the subject company and the average fundamental risk of companies comprising the portfolios in the D exhibits.
Presentation of the Results The C exhibits 25 portfolios are ranked by the same eight alternative measures of size as the A and B exhibits, as described in Table 2 (see page 20). Each of the exhibits C-1 through C-8 displays one line of data for each of the 25 size-ranked portfolios. The C exhibits include the statistics outlined in Table 3. In addition to information repeated from the A exhibits, the C exhibits report the additional datapoints for each of the 25 portfolios described in Table 6. Table 6: Statistics Reported for 25 Size-Ranked Portfolios in the C Exhibits (and not otherwise reported in the A Exhibits)
Average debt to market value of equity. Operating Margin: The mean operating income for the prior five years divided by the mean sales for the prior five years. Operating income is defined as sales minus cost of goods sold plus selling, general, and administrative expenses plus depreciation. Coefficient of Variation of Operating Margin: The standard deviation of operating margin over the prior five years divided by the average operating margin for the same years. Coefficient of Variation of Return on Book Value of Equity: The standard deviation of return on book equity for the prior five years divided by the mean return on book equity for the same years. Return on book equity is defined as net income before extraordinary items minus preferred dividends divided by book value of common equity.
Arithmetic average historical unlevered risk premium over long-term Treasuries (average return on equity in excess of long-term Treasury bonds) since 1963. (RPm+s, unlevered) Smoothed average historical unlevered risk premium: the fitted premium from a regression with the average historical unlevered risk premium as dependent variable and the logarithm of the average sorting criteria as independent variable (RPm+s, unlevered) (The coefficients and constants from this regression analysis are in the top right hand corner of the exhibits) Average unlevered beta calculated using the sum beta method applied to monthly returns for 1963 through the latest year (see the 2011 SBBI Valuation Yearbook pp. 77-78 for a description of the sum beta method).
Duff & Phelps
| 54
The High-FinancialRisk Study

The information and data in the Duff & Phelps Risk Premium Report (and in the online Duff & Phelps Risk Premium Calculator)132 is primarily designed to be used to develop cost of equity capital (COE) estimates for the large majority of companies that are fundamentally healthy, and for which a going concern assumption is appropriate. A set of high-financial-risk companies is set aside and analyzed separately in the High-Financial-Risk Study. The companies analyzed in the High-Financial-Risk Study are identified in a two-step process. First, companies that are losing money, have high leverage, or are in bankruptcy are identified and eliminated from the base set of companies used in the Size Study and Risk Study.133, 134 It is possible to imagine companies that dont have any of these characteristics but could still be classified as high-financial-risk (i.e. distressed), and it is also possible to imagine companies which do have one or more of these characteristics but are not distressed. For this reason, these companies are further scrutinized in a second test where they are ranked by the appropriate Altman z-Score (for manufacturing companies or for service companies).135, 136 Those companies identified as being in the safe zone (as defined by their z-Score) failed the first test, but passed the second test (z-Score), and are set aside and not used in any further analysis due to the inconclusive results. The remaining companies failed both the first test and the second test, and are placed in either the gray or distressed zone (as defined by their z-Score). The resulting base set of high-financial-risk companies is composed largely of companies whose financial condition is significantly inferior to the average, financially healthy public company. The results of the High-Financial-Risk Study are presented in the H exhibits. The H exhibits provide risk premia that may be used in both buildup and CAPM estimates of cost of equity capital if the individual analyst has determined that the subject company is high-financial-risk.137 In cases in which the individual analyst has determined that the subject company is high-financial-risk, the high-financial-risk premia reported in the H exhibits should be used instead of the returns reported in the Size Study, and not added to those returns.
132
133 134
135
136
137
The Duff & Phelps Risk Premium Calculator is available exclusively through Business Valuation Resources (BVR). For more information about the Calculator, visit: www.bvresources.com/dp or call toll free 1-(888)-287-8258. For a detailed discussion of how the high-financial-risk portfolios are created, see High-Financial-Risk Study in the portfolio methodology section on page 11. The number of companies eliminated in this screen varies from year to year. These companies represented up to 25% of the data set in recent years, but less than 5% in 1963. Certain technical changes in methodology have resulted in a greater number of companies falling into the high-financial-risk database than in versions of this study published prior to 2000. Altman z-Score is an accounting-data-based method designed to assess financial condition and developed originally for assessing the likelihood of bankruptcy. E. I. Altman, Financial Ratios, Discriminant Analysis and the Prediction of Corporate Bankruptcy, The Journal of Finance, Vol. 23, No. 4 (Sep., 1968), pp. 589-609; Predicting Financial Distress of Companies: Revisiting the s-Score and Zeta Models, July 2000; Revisiting Credit Scoring Models in a Basel 2 Environment, May 2002. Service industry companies are those SIC codes: 7200, 7300, 7500, 7600, 8000, 8100, 8200, 8300, 8400, 8700. Manufacturing are all other SIC codes, with the exception of SICs beginning with 6 (financial institutions) or 9 (government). SIC 6 and SIC 9 are not included in the Reports analysis. The decision to apply a high-financial-risk premium is ultimately dependent on the analysts professional judgment, based upon the analysts detailed knowledge of the subject company.
Duff & Phelps
| 55

The High-Financial-Risk H Exhibits Exhibit H-A is the high-financial-risk equivalent of the A exhibits. High-financial-risk premia over the risk free rate for use in a buildup method are found in the H-A exhibits. These premia can be added to the risk free rate to estimate the cost of equity capital for a company that has been judged by the analyst to be high-financial-risk. Exhibit H-B is the high-financial-risk equivalent of the B exhibits. High-financial-risk premia over CAPM (i.e. size premia) for use with the CAPM method are found in the H-B exhibits. These premia can be used in the CAPM to estimate the cost of equity capital for a company that has been judged by the analyst to be high-financial-risk. Exhibit H-C is the high-financial-risk equivalent of the C exhibits. The H-C exhibits can be used to compare the subject companys fundamental risk characteristics to the fundamental risk characteristics of portfolios made up of companies with similar z-Scores. Figure 54: The A, B, and C Exhibits and Corresponding High-Financial-Risk Exhibits
A Exhibit H-A Exhibit
B Exhibit
H-B Exhibit
C Exhibit
H-C Exhibit
Why isnt there an H-D exhibit? In the Risk Studys D exhibits, in addition to operating margin, two other measures of risk are examined (coefficient of variation in operating margin and coefficient of variation in return on equity). Because the denominators of these ratios are often negative for companies in the high-financial-risk portfolio as a result of either negative earnings or negative book value of equity, developing comparable high-financial-risk premia for these frequently results in meaningless statistics.
Duff & Phelps
| 56

Altman z-Score Altmans z-Score was originally designed as a measure to predict the risk of failure up to two years prior to distress for a sample of manufacturing companies using financial data prepared according to the standards of the day. The accuracy of predicting the risk of failure diminished substantially as the lead time increased. The z-Score resulted from a statistical analysis of company data using the statistical technique of multiple discriminant analysis. Altman has since offered improvements on the original z-Score, but the original z-Score is still frequently calculated as a convenient metric that captures within a single statistic a number of disparate financial ratios measuring liquidity, profitability, leverage, and asset turnover.138 Z-Score ratios are not strictly comparable across industries or across time (for instance, one would expect large differences in asset turnover among an industrial company or a retailer), and as such, are not used here as a predictor of bankruptcy per se, but as mechanism for ranking the high-financial-risk companies by their relative levels of distress. The following z-Score model for publicly-traded manufacturing companies (i.e. excluding service industry companies) is used in preparing the analyses presented in the H-A, H-B, and H-C exhibits: z = 1.2 x1 + 1.4 x2 + 3.3 x3 + 0.6 x4 + 0.999 x5 where: z = Overall index x1 = Net working capital / total assets x2 = Retained earnings / total assets x3 = Earnings before interest and income taxes / total assets x4 = Market value of common equity / book value of total liabilities x5 = Sales / total assets The companies are then sorted by z-Score into three portfolios: y z > 2.99 = safe zone y 1.80 < z < 2.99 = gray zone y z < 1.80 = distress zone Companies in the safe zone (z-Score greater than 2.99) are set aside and not used in any further analysis. Companies in the gray zone (z-Score between 1.80 and 2.99) and companies in the distressed zone (z-Score less than 1.80) are used to form the portfolios from which the statistics presented in H-A, H-B, and H-C exhibits are calculated. Portfolios are rebalanced annually (i.e. the companies are re-ranked and sorted at the beginning of each year). Portfolio rates of return were calculated using an equal-weighted average of the companies in the portfolio.
138
In applying any of the z-Score equations cited here, express the ratios in terms of their decimal equivalents (e.g., x1 = working capital / total assets = 0.083).
Duff & Phelps
| 57

The following z-Score model for publicly-traded service industry high-financial-risk companies is used in preparing the analyses presented in the H-A, H-B, and H-C exhibits: z = 6.56 x1 + 3.26 x2 + 6.72 x3 + 1.05 x4 where: z = Overall index x1 = Net working capital / total assets x2 = Retained earnings / total assets x3 = Earnings before interest and income taxes / total assets x4 = Book value of common equity / book value of total liabilities The companies are then sorted by z-Score into three portfolios. y z > 2.60 = safe zone y 1.10 < z < 2.60 = gray zone y z < 1.10 = distress zone Companies in the safe zone (z-Score greater than 2.60) are set aside and not used in any further analysis. Companies in the gray zone (z-Score between 1.10 and 2.59) and companies in the distressed zone (z-Score less than 1.10) are used to form the portfolios from which the statistics presented in H-A, H-B, and H-C exhibits are calculated. Portfolios are rebalanced annually (i.e. the companies are re-ranked and sorted at the beginning of each year). Portfolio rates of return were calculated using an equal-weighted average of the companies in the portfolio. Again, in both cases (manufacturing and service), we are not using the z-Score or z-Score as a predictor of bankruptcy. Rather, companies are ranked in the High-Financial-Risk Study based on their relative levels of distress, using z-Score and z-Score as proxies for distress. Non-Public Companies and z-Score The traditional z-Score was developed using data for publicly traded companies, and one of the statistics utilizes stock price. This creates problems for application of the data to non-public companies. Altman developed a similar model using only the financial statement data for non-public companies. If the subject company is not publicly traded and not in the service industry, then the analyst can calculate a z-Score for non-public companies (the z-Score) to compare with the data in the accompanying exhibits: z = 0.717 x1 + 0.847 x2 + 3.107 x3 + 0.420 x4 + 0.998 x5 where: z = Overall index x1 = Working capital / total assets x2 = Retained earnings / total assets x3 = Earnings before interest and income taxes / total assets x4 = Book value of common equity / book value of total liabilities x5 = Sales / total assets The z-Scores zones of discrimination loosely approximate the boundaries used to seperate the z-Score and z-Score ranked companies into portfolios, and are as follows: y z > 2.90 = safe zone y 1.23 < z < 2.90 = gray zone y z < 1.23 = distress zone While the H-A, H-B, and H-C exhibits are sorted by using the publically-traded company equations (z-Score for manufacturing companies and z-Score for service companies) and are not strictly comparable to the z-Score for non-public companies, the returns reported in these exhibits can be useful in developing cost of equity estimates based on the relative zones of discrimination.
Duff & Phelps
| 58

Measurement of Historical Risk Premiums The high-financial-risk Studys H exhibits report average historical risk premiums for the period 1963 (the year that the Compustat database was inaugurated) through 2010. A long-run average historical risk premium is often used as an indicator of the expected risk premium of a typical equity investor. Returns are based on dividend income plus capital appreciation and represents returns after corporate taxes (but before owner level taxes). To estimate historical risk premiums, an average rate of return is first calculated for each portfolio over the sample period. Portfolios with fewer than six companies in any given year are excluded in the averages. Lastly, the average income return earned on long-term Treasury bonds is subtracted from the portfolios returns over the same period (using SBBI data) to arrive at an average historical risk premium for investments in equity. Presentation of the Results Each of the exhibits H-A, H-B, and H-C displays one line of data for each of the the z-Score- and z-Score-ranked portfolios. These exhibits include the statistics outlined in Table 7. For comparative purposes, the average returns from the SBBI series for large companies (essentially the S&P 500 Index), small companies, and long-term government bond income returns for the period 1963 through the latest year are also reported on each exhibit.139
Table 7: Statistics Reported for the z-Score- and z-Score-ranked High-Financial-Risk Studys H-A, H-B, and H-C Exhibits
Exhibit H-A Beta calculated using the sum beta method applied to monthly returns for 1963 through the latest year (see the 2011 SBBI Valuation Yearbook pp. 77-78 for a description of the sum beta method). Standard deviation of annual historical equity returns. Exhibit H-B Beta calculated using the sum beta method applied to monthly returns for 1963 through the latest year (see the 2011 SBBI Valuation Yearbook pp. 77-78 for a description of the sum beta method). Arithmetic average historical equity return since 1963. Exhibit H-C Arithmetic average historical risk premium over long-term Treasuries (average return on equity in excess of long-term Treasury bonds) since 1963 (RPm+s, high-financial-risk). Average carrying value of preferred stock plus long-term debt (including current portion) plus notes payable (Debt) as a percent of MVIC since 1963. Average debt to market value of equity.
Geometric average historical equity return since 1963.
Arithmetic average historical risk premium over long-term Treasuries (average return on equity in excess of long-term Treasury bonds) since 1963 (RPm+s, high-financial-risk). Indicated CAPM premium, calculated as the beta of the portfolio multiplied by the average historical market risk premium since 1963 (measured as the difference between SBBI Large Stock total returns and SBBI income returns on long-term Treasury bonds). Premium over CAPM, calculated by subtracting the Indicated CAPM Premium from the Arithmetic Risk Premium (RPs, high-financial-risk).
Arithmetic average historical equity return since 1963.
Beta calculated using the sum beta method applied to monthly returns for 1963 through the latest year (see the 2011 SBBI Valuation Yearbook pp. 77-78 for a description of the sum beta method).
Arithmetic average historical risk premium over long-term Treasuries (average return on equity in excess of long-term Treasury bonds) since 1963 (RPm+s, high-financial-risk). Average carrying value of preferred stock plus long-term debt (including current portion) plus notes payable (Debt) as a percent of MVIC since 1963.
Operating Margin: The mean operating income for the prior five years divided by the mean sales for the prior five years. Operating income is defined as sales minus cost of goods sold plus selling, general, and administrative expenses plus depreciation.
139
Source: Morningstar EnCorr Analyzer software.
Duff & Phelps
| 59

Overview of Methods Used to Estimate Cost of Equity Capital Using the High-Financial-Risk Study The High-Financial-Risk Study provides two methods of estimating COE for a subject company that has been determined to be high-financial-risk: Buildup 1-High-Financial-Risk and CAPM-High-Financial-Risk. These methods are summarized in equation format, and summarized in Figure 55 in graphical building blocks format. 1) Buildup 1-High-Financial-Risk COEBuildup 1-High-Financial-Risk = (Risk Free Rate) + (High Financial Risk Premium Over Risk Free Rate) + (Equity Risk Premium Adjustment) Example 7: page 61 2) Capital asset pricing model (CAPM)-High-Financial-Risk COECAPM-High-Financial-Risk = (Risk Free Rate) + (Beta x Equity Risk Premium) + (High-Financial-Risk Size Premium) Example 8: page 89 (in the full Report)
Figure 55: Two Methods of Estimating Cost of Equity Capital with the High-Financial-Risk Study140 Buildup 1-High-Financial-Risk CAPM-High-Financial-Risk + High-Financial-Risk Premium Over CAPM (Size Premium), RPs, high-financial-risk Cost of Equity + (Beta x ERP) Basic CAPM Risk Free Rate, Rf (Use Exhibit H-B size premia) Cost of Equity
+ ERP Adjustment* + High-Financial-Risk Premium Over Risk Free Rate, RPm+s, high-financial-risk Risk Free Rate, Rf (Use Exhibit H-A risk premia)
* ERP Adjustment: The difference between the historical (19632010) equity risk premium (ERP) and a user of the Duff & Phelps Reports own forward ERP estimate: ERP Adjustment = Users ERP Historical (19632010) ERP The ERP Adjustment is made only in the Buildup 1, Buildup1Unlevered, Buildup 1-High-Financial-Risk, Buildup 3, and Buildup 3-Unlevered methods. Please refer to the individual examples provided for these models for more information.
NOTE: This section includes an example of using the Reports risk premia data to estimate cost of equity capital using the Buildup 1-High-Financial-Risk method. A complete example for using the Reports risk premia to estimate cost of equity capital using the CAPM-High-Financial-Risk method is available in the full version of the 2011 Report.
140
The relative sizes of the building blocks in Figure 55 do not necessarily represent the relative size of the various inputs. Also note that the names given to the models in the Risk Premium Report (e.g. Buildup 1, Buildup 2, Buildup 3, etc.) are naming conventions used within the Report to make referring to the different methods easier.
Duff & Phelps
| 60

In this section, the information in Figure 56 will be used to estimate cost of equity capital for a hypothetical non-service (i.e. manufacturing) subject company. Figure 56: Subject Company Characteristics (used in all examples)
(in $millions) Market value of equity Book value of equity Total assets Most recent year EBIT $80 $100 $300 -$5 Sales Current assets Current liabilities Retained earnings (in $millions) $250 $75 $50 $75
Example 7: Estimating Cost of Equity Capital Using the Buildup 1-High-Financial-Risk Method
Buildup 1-High-Financial-Risk
+ ERP Adjustment Cost of Equity
+ High-Financial-Risk Premium Over Risk Free Rate, RPm+s, high-financial-risk
Risk Free Rate, Rf The z-Score equation for a publicly-traded, non-service (i.e. manufacturing) subject company is: z= 1.2 x1 + 1.4 x2 + 3.3 x3 + 0.6 x4 + 0.999 x5 The inputs (x1, x2, x3, x4, and x5) needed for the z-Score equation are calculated as shown in Figure 57: Substituting these inputs into the z-Score equation yields a z-Score of 1.47: z = 1.2(0.0833) + 1.4(0.2500) + 3.3(-0.0167) + 0.6(0.4000) + 0.999(0.8333) 1.47 = 0.1000 + 0.3500 + (-0.0550) + 0.2400 + 0.8325 The buildup method is an additive model commonly used for calculating the required rate of return on equity. As the name implies, successive building blocks are summed, each representing the additional risk inherent to investing in alternative assets. An example of this is the extra return (i.e. premium), that investors demand for investing in stocks versus investing in a riskless security.141,142 This example utilizes the long-term risk free rate (Rf) and the ERP Adjustment established in a previous example (the Size Studys Buildup 1 method using guideline portfolios; see page 31 in the full Report). This mirrors the fact that for any given valuation engagement the same risk free rate and ERP will generally be used in each of the models presented by the individual analyst. Please note that for any given valuation engagement these inputs may (and probably will) be different than the ones used in the examples.
(Use Exhibit H-A risk premia)
Figure 57: z-Score Inputs Calculation

x1 x2 x3 x4 x5 = = = = = Net working capital / total assets Retained earnings / total assets Earnings before interest and taxes / total assets Market value of common equity / book value of total liabilities Sales / total assets = ($75 current assets - $50 current liabilities) / ($300 total assets) = ($75 retained earnings) / ($300 total assets) = (-$5 EBIT) / ($300 total assets) = ($80 market value of equity) / ($300 total assets - $100 book value of equity) = ($250 sales) / ($300 total assets) = 0.0833 = 0.2500 = -0.0167 = 0.4000 = 0.8333
141 142
Throughout this report the risk free asset (Rf) is represented by the yield on a 20-year constant maturity Treasury Bond. For a detailed discussion of the buildup model, see Estimating Cost of Equity Capital Using the Buildup 1 Method on page 29 (in the full Report).
Duff & Phelps
| 61

As in the Buildup 1 method, the Buildup 1-High-Financial-Risk method requires three pieces of information to estimate the cost of equity capital: a risk free rate (Rf), a high-financial-risk premium over the risk free rate (RPm+s, high-financial-risk), and an ERP Adjustment (if necessary). All of the information needed is summarized in Figure 58. Figure 58: Information Needed to Estimate COE Using Buildup 1-High-Financial-Risk
Determination of the high-financial-risk premium in Exhibit H-A for Step 2 is a three-step process (Steps 2a, 2b, and 2c): Step 2a: Determine whether the characteristics of the subject company better match the characteristics of the companies included in Exhibits A-1 through A-8 (the 25 portfolios) or the characteristics of the high-financial-risk portfolios of companies as described above. The most straightforward way of doing this is to answer the following five questions about the subject company:145 y Is the subject company in bankruptcy or in liquidation? y Is the subject companys 5-year average net income available to common equity less than zero for the previous five years? y Is the subject companys 5-year-average operating income less than zero for the previous five years? y Has the subject company had a negative book value of equity at any one of the companys previous five fiscal year-ends? y Does the subject company have a debt-to-total capital ratio of more than 80%? It is possible to imagine companies that dont have any of these characteristics, but could still be classified as high-financial-risk (i.e. distressed), and it is also possible to imagine companies which do have one or more of these characteristics but are not distressed. If you answered Yes to one or more of the five questions, it may suggest that the subject companys characteristics are more like the companies that make up the high-financial-risk portfolios rather than like the healthy companies that make up the standard 25 portfolios, but not necessarily so. For example, a company may have a debt to total capital ratio greater than 80%, but this does not automatically imply that the company is in distress. A decision by the individual analyst that a company should be treated as high-financial-risk should be based on a detailed evaluation of the companys current financial condition and circumstances, and will generally involve more than a review of historical financial statistics and ratios. The decision to apply a high-financial-risk premium is ultimately dependent on the individual analysts professional judgment and detailed knowledge of the subject company.146
Rf
RPm+s, highfinancial-risk
ERP Adj.
COE
The only difference between estimating cost of equity capital (COE) using the Buildup 1 method and estimating COE using the Buildup 1-High-Financial-Risk method is that the risk premium over the risk free rate used in the latter method is a high-financial-risk premium (RPm+s, high-financial-risk), while the risk premia over the risk free rate used in the former are not.143 Step 1 and Step 3: Because the risk free rate in Step 1 (Rf , 4.1%) and the ERP Adjustment in Step 3 (1.1%) established in a previous example are being used in this example144, the only missing ingredient needed to estimate COE is the high-financial-risk premium over the risk free rate (RPm+s, high-financial-risk): COEBuildup 1-High-Financial-Risk = Rf + RPm+s,high-financial-risk + ERP Adjustment = COEBuildup 1-High-Financial-Risk = 4.1% + RPm+s,high-financial-risk + 1.1%
143
144 145
146
The risk premia over the risk free rate used in the Buildup 1 method are found in the A exhibits. A, B, C, and D risk premia are designed to be used to develop cost of equity capital (COE) estimates for the large majority of companies that are fundamentally healthy; the H exhibits are designed to be used to estimate COE for companies that the individual analyst has determined to be high-financial-risk. See the Size Studys Buildup 1 method using guideline portfolios on page 31 (in the full Report). These five questions mirror the five criteria by which high-financial-risk companies are identified in (and eliminated from) the universe of US companies to form the base set of companies used in the Size Study and Risk Study. If the analyst determines that the subject company is not high-financial-risk, the returns reported in the exhibits in the Risk Premium Report for the 25 portfolios (the A, B, C, and D exhibits) may be more appropriate for the subject company than the returns reported in the H exhibits.
Duff & Phelps
| 62

Step 2b: If the individual analyst determines that the subject companys characteristics better match the characteristics of the companies comprising the high-financial-risk portfolios, calculate the z-Score of the subject company using the appropriate z-Score equation:147 y z-Score is for publicly-traded, non-service, (i.e. manufacturing) companies148 y z-Score is for publicly-traded, service companies y z-Score is non-public, non-service companies. Step 2c: Lastly, if the z-Score149 of the subject company indicates that it is in the gray zone or distress zone, match the z-Score of the subject company with the zone composed of companies with similar z-Scores in Exhibits H-A, and identify the corresponding average high-financial-risk premium over the risk free rate (RPm+s, high-financial-risk). For this example, the subject company is a manufacturing company with a z-Score of 1.47, placing it in the distressed portfolio (z-Scores <1.8; see Figure 59). The corresponding high-financial-risk arithmetic average risk premium is 17.26 percent (17.3% rounded). Step 4: Estimate a high-financial-risk cost of equity for the subject company by adding the average high-financial-risk premium over the risk free rate identified in Step 3 (RPm+s, high-financial-risk) to the risk free rate Rf and the ERP Adjustment (if appropriate). COEBuildup 1-High-Financial-Risk = Rf + RPm+s, high-financial-risk + ERP Adjustment = 22.5% = 4.1% + 17.3% + 1.1% The high-financial-risk COE estimate for the hypothetical subject company in this example is 22.5 percent.
Figure 59: Buildup 1-High-Financial-Risk COE Input Exhibit H-A, High-Financial-Risk Premia Over the Risk Free Rate Companies Ranked by z-Score Historical Equity Risk Premium: Average Since 1963 High-Financial-Risk Company Data for Year Ending December 31, 2010
Beta (SumBeta) Since 63 Standard Deviation of Returns Arithmetic Average Risk Premium
Portfolio Rank Manufacturing (z-Score) 1.8 to 2.99 (gray zone) < 1.8 (distress zone) Service Industry (z-Score) 1.1 to 2.59 (gray zone) < 1.1 (distress zone)
Geometric Average Return
Arithmetic Average Return
Average Debt/MVIC
1.58 1.65
37.20% 40.01%
15.05% 16.66%
21.50% 24.16%
17.26%
47.04% 58.28%
1.63 1.71
44.33% 46.48%
14.08% 19.93%
28.00% 35.40%
42.31% 50.33%
147 148
149
In all examples here, the z-Score for publicly-traded, non-service (i.e. manufacturing) companies is used. While the H-A, H-B, and H-C exhibits are ranked by z-Score and z-Score and are not strictly comparable to the z-Score for non-public companies, the returns reported in these exhibits can be useful in developing cost of equity estimates based on the relative zones of discrimination. Or, as appropriate, z-Score or z-Score.
Duff & Phelps
| 63

Additional information on Company-Specific Risk The Risk Study provides analysis that correlates historical equity returns (and historical risk premiums) directly with three measures of company-specific risk derived from accounting information (five-year operating income margin, the coefficient of variation in operating income margin, and the coefficient of variation in return on book equity). These may also be called fundamental measures of company risk to distinguish them from stock market-based measures of equity risk (e.g. beta). The Risk Study demonstrates that as company size decreases, measures of risk calculated from financial statement data do, as a matter of fact, tend to increase. In the High-Financial-Risk Study, one measure of accountingdata-based fundamental risk (five-year operating income margin) was examined for portfolios formed by ranking public companies by z-Score (manufacturing companies) and z-Score (service companies).155, 156 The H-C exhibits can be used to compare a subject companys operating margin to the operating margins of portfolios made up of companies with similar z-Scores (or z-Scores). For example, in the previous examples (Example 7 and Example 8), the subject company was a manufacturing company with a z-Score of 1.47, placing it in the distressed zone in exhibits H-A and H-B. The average operating margin (2.5%) of the companies comprising the portfolio used to calculate the statistics for manufacturing companies in the distress zone in exhibits H-A and H-B is published in Exhibit H-C (see Figure 62). If the hypothetical subject company in these examples has a higher operating margin of, say 7 percent, it may be less risky than companies with similar z-Scores. This may suggest that a downward companyspecific risk adjustment is justified.
Figure 62: Exhibit H-C Companies Ranked by Market Value of Equity: Comparative Risk Characteristics High-Financial-Risk Company Data for Year Ending December 31, 2010
Arithmetic Average Risk Premium Average Debt to Market Value of Equity Beta (SumBeta) Since 63
Average Debt toMVIC
Average Operating Margin
14.6% 17.3%
47.04% 58.28%
88.8% 139.7%
1.58 1.65
2.5%
21.1% 28.5%
42.31% 50.33%
73.3% 101.3%
1.63 1.71
155
156
Because the denominators of the other two ratios (coefficient of variation in operating income margin, and coefficient of variation in return on book equity) are often negative for companies in the high-financial-risk portfolios (as a result of either negative earnings or negative book value of equity), developing comparable high-financial-risk premia for them frequently results in meaningless statistics. Operating margin is defined here as the mean operating income for the prior five years divided by the mean sales for the prior five years. Operating income is defined as sales minus cost of goods sold plus selling, general, and administrative expenses plus depreciation.
Duff & Phelps
| 64

Table 8 provides additional summary statistics for the H exhibits z-Score- and z-Score-ranked portfolios.157 For example, portfolios made up of manufacturing companies with an average z-Score less than 1.8 had an average book value of equity of $ 93.788 million and an average 5-year average net income of -$16.540 million.
Table 8: Companies Ranked by Sorting Criteria High-Financial-Risk Company Data for Year Ending December 31, 2010 Portfolio Details ($mils.)
Portfolio Average Number Number as of 2010 Market Value of Equity Book Value of Equity 5-Year Average Net Income Market Value of Invested Capital Total Assets 5-Year Average EBITDA Number of Employees
Sales
123 277
168.933 205.947
130.179 93.788
(5.690) (16.540)
338.415 609.226
487.036 613.120
56.662 61.662
752.698 474.954
2,600 1,606
29 65
210.376 118.710
152.083 35.311
(1.326) (11.006)
324.831 189.694
361.914 184.182
26.870 16.294
249.228 198.233
1,189 1,064
157
The information in Table 8 was published as Exhibit H-E in previous Reports.
Duff & Phelps
| 65
Data Exhibits
NOTE: The data exhibits are not included in this document (the data exhibits are available in the complete 2011 Report).
Duff & Phelps
| 66
New in 2011: The Duff & Phelps Risk Premium Calculator (web-based)
In 2011 we introduce the web-based Duff & Phelps Risk Premium Calculator. The Calculator automatically estimates levered and unlevered cost of equity capital (COE) for your subject company dependent on its size and risk characteristics (for any valuation date from January 1, 1996 to present), using both the capital asset pricing model (CAPM) and buildup models. The Calculator is easy to use, saves time, and automatically provides full summary output in both Microsoft Word and Microsoft Excel format. In addition, the Calculator automatically looks up the long-term risk free rate for your valuation date1, automatically makes the important (but often overlooked) ERP Adjustment to your subject companys COE estimates, and automatically adjusts an SBBI industry risk premium (IRP) so that it can be used in a Buildup model using Risk Premium Report size premia.2 Calculator Features y Anytime, anywhere access at www.bvmarketdata.com/DP.RPC y Complete historical database of risk premia and size premia data (19962011) y Automatic output y Executive Summary of COE estimates, including CAPM, Buildup, and unlevered COE y Microsoft Excel output of all underlying values and calculations y Easy to use / Saves time The Calculator employs the methodology and data published in the Duff & Phelps Risk Premium Report, which has provided financial and valuation professionals defensible cost of capital data and methodology since 19963 Calculator Tour Duff & Phelps designed the Calculator with two simple goals: the user experience had to be as easy and smooth as possible, and the Calculator had to maintain the same analytical horsepower, data, and methodology under the hood as is found in the Risk Premium Report. There are three simple steps needed to calculate cost of equity capital (COE) using the Calculator.
Step 1
Step 2
Step 3
y Log in
y Enter Subject Company Inputs y Size Characteristics y Risk Characteristics
y Receive Output y Executive Summary y Excel Summary
1 2 3
20-year constant maturity Treasury bond yield as of your valuation data. Source: The Board of Governors of the Federal Reserve System. Duff & Phelps does not publish IRPs. A source of IRPs is Morningstars Ibbotson SBBI Valuation Yearbook, (Chicago, Morningstar), Chapter 3, The Buildup Method, Table 3-5. For detailed information about the Size Study, Risk Study, and High-Financial-Risk Study included in the Risk Premium Report (and now available in the Risk Premium Calculator), please see pages 18, 39, and 55, respectively.
Duff & Phelps
| 67
Step 1: Log in at www.bvmarketdata.com/DP.RPC Image 1 Logging in
Step 2a: Enter your subject companys name, and the valuation date. Image 2 Subject Company Name and Valuation Date
Duff & Phelps
| 68
Step 2b: An optional set of questions and inputs is provided if the individual analyst has determined that the subject company is high-financial-risk.4 Image 3 Optional High-Financial-Risk Information
The five questions in this step mirror the five criteria by which highfinancial-risk companies are identified in (and eliminated from) the universe of US companies to form the base set of companies used in the Size Study and Risk Study. If you answer Yes to one or more of the five questions, it may suggest that the subject companys characteristics are more like the companies that make up the high risk portfolios rather than like the healthy companies that make up the standard 25 portfolios, but not necessarily so. For example, a company may have a debt to total capital ratio greater than 80%, but this does not automatically imply than the company is in distress.
The information and data in the Duff & Phelps Risk Premium Calculator is primarily designed to be used to develop cost of equity capital (COE) estimates for the large majority of companies that are fundamentally healthy, and for which a going concern assumption is appropriate. A set of high-financial-risk companies is set aside and analyzed separately in the High-Financial-Risk Study. The decision to apply a highfinancial-risk premium is ultimately dependent on the analysts professional judgment, based upon the analysts detailed knowledge of the subject company. Please note that High-Financial-Risk Study output is available for calendar year 2010 valuation dates (and later) only.
Duff & Phelps
| 69
Step 2c: The next step is entering your subject companys size characteristics and risk characteristics. Note that the appropriate long-term risk free rate (4.13%) for the valuation date is automatically looked up and entered in the Risk Free Rate field for your convenience.5 If you want to use a different risk free rate, just type over the value that the Calculator automatically entered in this field. Image 4 Basic Inputs Screen (not filled out)
Also note that the Calculator provides information and tips which appear if you hover your mouse cursor over one of the information icons . These helpful tips provide quick assistance if you need the definition of an input, or the source of an input.
20-year constant maturity Treasury bond yield as of your valuation data. Source: The Board of Governors of the Federal Reserve System. The risk free rate field can be overtyped (edited) by the analyst.
Duff & Phelps
| 70
Fill in your subject companys size characteristics and risk characteristics, as shown in Image 5. Image 5 Basic Inputs Screen (filled out)
Under General Inputs, enter the equity risk premium (ERP) you want used in all cost of equity capital (COE) calculations. For example, many users of the Risk Premium Report use the Duff & Phelps Recommended ERP, which was 5.5 percent at the end of 2010.6,7,8 Also under General Inputs, enter a beta if you would like COE estimated using the capital asset pricing (CAPM) model, and an industry risk premium (IRP) from the SBBI Yearbook if you would like COE estimated using a buildup model that utilizes an IRP to account for market risk. Only one (of the eight total) Size Study inputs is required, but enter as many of the eight values as possible for best results.
If you wish to receive cost of equity capital estimates derived using the Risk Study, the three most recent years of information are required (for best results, enter the most recent five years of information). Please note that the Calculator automatically makes the important (but often overlooked) ERP Adjustment to your subject companys COE estimates, and automatically adjusts an SBBI industry risk premium (IRP) so that it can be used in a Buildup model using Risk Premium Report size premia.
For more information on the equity risk premium, see Cost of Capital: Applications and Examples 4th ed., by Shannon P. Pratt and Roger J. Grabowski (John Wiley & Sons, Inc., 2010), Chapter 9, Equity Risk Premium, pages 115158. See Roger J. Grabowski, Developing the Cost of Equity Capital: Risk-Free Rate and ERP During Periods of Flight to Quality. This paper will appear in the Business Valuation Review and can also be downloaded at Duff & Phelps Cost of Capital site at www.duffandphelps.com/CostofCapital If no ERP is entered, the historical ERP as calculated over the time horizon 1963 to the (year of the valuation date -1) is used. For example, for a calendar year 2011 valuation date, if no ERP is entered by the analyst in General Inputs the ERP as calculated from 19632010 (4.4%) would be used in all calculations; for a calendar year 2010 valuation date, if no ERP is entered by the analyst in General Inputs, the ERP as calculated from 19632009 (4.3%) would be used in all calculations, etc.
Duff & Phelps
| 71
Prior to calculating COE estimates for your subject company, the Calculator displays a summary of all of your inputs as shown in Image 6. At this point you can review your inputs, and change them (if necessary). By clicking the Confirm button, you are agreeing that all of your inputs are as you intend, and the Calculator then calculates cost of equity capital (COE) estimates for your subject company. Image 6 Confirm / Change Inputs
Duff & Phelps
| 72
After the Calculator calculates estimates of the subject companys cost of equity capital (COE), an abbreviated online results preview is displayed, as shown in Image 7. Image 7 Cost of Equity Capital (COE) Estimates (online results preview)
Click the DOCX and XLSX links for instant download of Executive Summary and Support and Detail documents.
Your complete (as opposed to online results preview) COE estimate report includes an Executive Summary in Microsoft Word format and a Support and Detail Microsoft Excel workbook, which can be instantly downloaded by clicking on the XLSX and DOCX links at the top of the online results preview page, as indicated in Image 7.
Duff & Phelps
| 73
Your complete COE estimate report includes: Executive Summary (in Microsoft Word format) The Executive Summary is a high-level overview of data sourcing information, key inputs used in calculations, and cost of equity capital (COE) estimates for all models employed (with your subject companys information plugged into each models equation)9, plus a concluded range of COE estimates for your subject company (using both the Size Study and Risk Study).10 Because the Executive Summary is in Microsoft Word format, you can edit it and format it to suit your individual needs. For example, inserting your own disclaimer information or adding your company logo is easy. Support and Detail summary of all inputs and calculations (in Microsoft Excel format) The Support and Detail workbook includes a summary of your subject companys size and fundamental risk characteristics (and all other inputs), and complete documentation of calculations and inputs for each of the models used to estimate cost of equity capital (COE) for your subject company. The Support and Detail workbook also includes the data exhibits11 for each of the guideline portfolios that match your subject company (by size and/or fundamental risk). This important information includes a complete listing of size premia and risk premia (both levered and unlevered), average arithmetic and geometric returns, sum betas, average debt to MVIC, average debt to market value, average operating margin, average coefficient of variation of operating margin, average coefficient of variation of ROE, z-Score, and more. An additional (and very important) capability of the Calculator that is documented in the Support and Detail workbook is that the Calculator automatically maps your subject companys size measures from the Size Study to portfolios of companies sorted by the three fundamental risk measures analyzed in the Risk Study, and then analyzes whether an upward or downward company-specific risk adjustment is indicated for each of the three fundamental risk factors. Why is this important? If two or more of the indicators are saying the same thing (upward adjustment or downward adjustment), it is a very powerful argument in defending a company-specific risk adjustment. An additional (and very important) capability of the Calculator that is documented in the Support and Detail workbook is that the Calculator automatically maps your subject companys size measures from the Size Study to portfolios of companies sorted by the three fundamental risk measures analyzed in the Risk Study, and then analyzes whether an upward or downward company-specific risk adjustment is indicated for each of the three fundamental risk factors. Why is this important? If two or more of the indicators are saying the same thing (upward adjustment or downward adjustment), it is a very powerful argument in defending a company-specific risk adjustment. Because the Support and Detail workbook is in Microsoft Excel format, you can edit it and format it to suit your individual needs. The workbook also includes a table of content tab and section divider tabs, so that when printed it is an organized, polished document ready for insertion into your valuation engagement report as a detailed support, sourcing, and documentation section designed to accompany the Executive Summary. For free samples of complete Executive Summary and Support and Detail outputs, or for more information about the Calculator, please visit: www.BVResources.com/dp
9 10
11
Please note that the number of models employed is dependent on the completeness (or lack thereof) of subject company inputs entered by Calculator users. The Duff & Phelps Risk Premium Calculator is based upon the Duff & Phelps Risk Premium Report. The Risk Premium Report includes a Size Study (which analyzes the relationship between equity returns and company size using up to eight measures of company size), a Risk Study (which analyzes the relationship between equity returns and accounting-based fundamental risk measures), and a High-Financial-Risk Study (which analyzes the relationship between equity returns and high-financial-risk, as measured by the Altman z-Score). Exhibits A-1 through A-8 (used in the Buildup method); B-1 through B-8 (used in the CAPM method); C-1 through C-8 (used to compare your subject companys risk characteristics to portfolios comprised of companies of the same size as the subject company); D-1 through D-8 (used to estimate COE based upon fundamental risk factors); H-A, H-B, and H-C (high-financial-risk premia and size premia), and z-Score calculations. Please note that the data and information included in the Support and Detail workbook is dependent on the completeness (or lack thereof) of subject company inputs entered by Calculator users.
Duff & Phelps
| 74
Product Purchasing Information You can purchase the Duff and Phelps Risk Premium Calculator exclusively through Business Valuation Resources (BVR) at: www.bvresources.com/dp 1-(888)-287-8258. All purchases of the Duff & Phelps Risk Premium Calculator include a copy of the Duff & Phelps Risk Premium Report. Calculator Option 1 Includes 16 years of size premia and risk premia data (1996 2011): 1-year subscription includes a copy of the 2011 Duff & Phelps Risk Premium Report and unlimited access to Duff & Phelps Risk Premium Calculator data from 1996-2011. Estimate cost of equity capital for any valuation date from January 1, 1996 to present. $759 Calculator Option 2 Includes 2 years of size premia and risk premia data (2010 2011): 1-year subscription includes a copy of the 2011 Duff & Phelps Risk Premium Report and unlimited access to the most recent two years of Calculator data (currently 2010-2011). Estimate cost of equity capital for any valuation date from January 1, 2010 to present. $399 Calculator Option 3 Single Year Duff & Phelps Risk Premium Report: Includes 1-time use of Risk Premium Calculator. $275
Duff & Phelps
| 75
For more information, visit: www.duffandphelps.com
Report Contributors Roger J. Grabowski, ASA, Author Managing Director James P. Harrington, Editor Director
About Duff & Phelps As a leading global provider of financial advisory and investment banking services, Duff & Phelps balances analytical skills, deep market insight and independence to help clients make sound decisions. The firm provides expertise in the areas of valuation, transactions, financial restructuring, alternative assets, disputes and taxation, with more than 1,000 employees serving clients from offices in North America, Europe and Asia.
Investment banking services in the United States are provided by Duff & Phelps Securities, LLC. Investment banking services in the United Kingdom and Germany are provided by Duff & Phelps Securities Ltd. Duff & Phelps Securities Ltd. is authorized and regulated by the Financial Services Authority. Investment banking services in France are provided by Duff & Phelps SAS. (NYSE: DUF)
Copyright 2011 Duff & Phelps Corporation. All rights reserved. DP111197

2011 Duff Phelps Risk Premium Report EXCERPT

Uploaded by

Copyright:

Available Formats

2011 Duff Phelps Risk Premium Report EXCERPT

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2011 Duff Phelps Risk Premium Report EXCERPT

Uploaded by

Copyright:

Available Formats

Risk Premium Report 2011

Selected Pages and Examples

Risk Premium Report 2011

Publication information/Disclaimer/ Purchasing information

Duff & Phelps

Risk Premium Report 2011

The Size Study 6

How the 2011 Report is Organized

Using the 2011 Report

Duff & Phelps

Risk Premium Report 2011

The Risk Study

The High-Financial-Risk Study

Appendix A Appendix B Glossary Data Exhibits

94 101 103 105

Duff & Phelps

Risk Premium Report 2011

Duff & Phelps

Risk Premium Report 2011

Duff & Phelps

Risk Premium Report 2011

How the 2011 Report is Organized

Duff & Phelps

Risk Premium Report 2011

How the 2011 Report is Organized

Duff & Phelps

Risk Premium Report 2011

Duff & Phelps

Risk Premium Report 2011

Duff & Phelps

Risk Premium Report 2011

Earnings before interest, income taxes, depreciation and amortization.

Duff & Phelps

Risk Premium Report 2011

Duff & Phelps

Risk Premium Report 2011

Duff & Phelps

Risk Premium Report 2011

Using the 2011 Report

Duff & Phelps

Risk Premium Report 2011

Using the 2011 Report

Duff & Phelps

Risk Premium Report 2011

Using the 2011 Report

Figure 4: Location of Regression Method Equation in the Data Exhibits

17.475% 0.876% 81% 25 23

-2.843% 0.285% -9.98

Smoothed Premium = 17.475% - 2.843% * Log(Book Value)

Smoothed Premium = 17.475% 2.843% *Log(Book Value)

Duff & Phelps

Risk Premium Report 2011

Using the 2011 Report

Table 1: Size Measures of Companies that Comprise Portfolio 25, by Percentile

Sales $17.345 49.578 105.445 176.928 241.513

Number of Employees 13 112 245 379 524

The information in Table 1 was published as Exhibit E in previous reports.

Duff & Phelps

Risk Premium Report 2011

The Size Study