# Methods

Though significant amounts of data are available for the subject firms in this study through FactSet, Value Line, Bloomberg and other data repositories, some of the necessary data variables critical to the formation of the targeted asset valuation models required estimating values based on aggregating known values, calculating average values or rates of change during a particular year, or synthesizing data points.

Data Collection1

Data collection for this study was performed by undergraduate student researchers in the Bill and Vieve Gore School of Business at Westminster College, Salt Lake City in May 2017 under the direction of Richard Haskell, PhD.  The students each participated in a May Term course, titled “The Power of Multiples”.  The course was taught by Haskell, Associate Professor of Finance, and Principal Investigator on The Valuation Project and The Power of Multiples.

The intended outcomes for the course included aiding students in better understanding the data, models and methods behind the calculation of value for publicly-traded firms, how the relationship these calculated values have to observable values, and the role of valuation multiples in these processes.  The course pursued these outcomes by engaging students through research into various valuation metrics, firm and industry analyst reports, market conditions, firm performance, and investor and firm expectations.  These inputs include the variables driving changes in a firm values, both in the markets and via valuation modeling.

The knowledge level of the student researchers ranged from uninformed and inexperienced to highly proficient in the use of valuation models, financial markets, and data repositories in which valuable market information is held.  These students were organized into six teams of three to four students each with each team assigned 30 of the study’s subject firms.  Team Leader assignments were given to students with high levels of proficiency in the relevant metrics of the study, who were experienced users of FactSet2, who held current Bloomberg3 certifications and who had proven their leadership capabilities throughout their undergraduate experiences.

Each firm in the study was assigned to two separate teams with each team assigning one individual to collect the required data.  The data was checked for accuracy by each team leader prior to being rechecked by a representative from each of the two teams assigned to the subject firm.  After the two team representatives signed off on the accuracy and completeness of the collected data, the data was then subjected to another quality and completeness inspection by one more team leader.  Any data discrepancies were reconciled by the respective team representatives or team leaders.  Discrepancies requiring further attention were adjudicated by Haskell prior to being included in valuation analyses.

The process of data collection, accuracy and completion checks, and reconciliation continued was continued by the study’s Research Associates, who then organized the data into Primary Input Data for each firm preparatory to calculating Derived Input Data and Valuation Model Outcomes reported on each of the subject Firm pages.

1 Richard Haskell, PhD (2017), Associate Professor of Finance, Bill & Vieve Gore School of Business, Westminster College, Salt Lake City, Utah, rhaskell@westminstercollege.edu, www.richardhaskell.net; Bryce Nieberger (2017), Student Research Associate, Bill and Vieve Gore School of Business, Westminster College, Salt Lake City, Utah, https://www.linkedin.com/in/bryce-nieberger-91571ab7

2 FactSet is a data repository delivering information to investment professionals and investors using publicly available analytics, service, content and technology; copyright FactSet Research Systems, Inc.; www.factset.com

3 Bloomberg connects investors, decision makers and investment professionals with real-time and historical information for a wide range of domestic and international firm; Bloomberg L.P.; http://www.valueline.com/

Primary Input Data1

Primary Input Data collected for this study includes data collected for firms differentiated by between asset and equity based operational structures.  Asset based firms primarily generate cash flows by employing operating assets in the production of the firm’s output goods and comprise the majority of firms in any marketplace.  Equity based firms, such as commercial and consumer banks, generate cash flows by employing their equity components in the production of the firm’s output goods.  The differentiated nature structures of these firms, their financial statements and utilize modified versions of the financial multiples compared to capital based companies.

Asset Based Firms

Balance Sheet

Current Assets (CA), Cash & Equivalents, Property, Plant, and Equipment (PPE), Other Long Term Assets, Current Liabilities (CA) and Book Value of Long Term Debt were identified for each of the study’s subject firms for calendar year end 2000 through 2016 from the Balance Sheet under the Financials section of FactSet2.   FactSet’s reporting of Total Current Assets was taken as Current Assets (CA), Total Current Liabilities as Current Liabilities (CL), and Long-Term Debt as the Book Value of Long Term Debt.  Values for Cash and Equivalents used the reported value for Cash & Sort-Term Investments; in some cases the reported value for Cash was added to the reported value for Short-Term Investments was to arrive at Cash and Equivalents.  The reported value for Net PPE was taken for Property, Plants and Equipment (PPE) as this value included accumulated depreciation.

The value for Other Long Term Assets was calculated by subtracting the combined value of Total Current Assets and Net PPE from the reported Total Assets: Other Long-Term Assets = Total Assets – Current Assets + Net PPE.

Income Statement

Sales, Earnings Before Interest and Taxes (EBIT), Depreciation & Amortization, Net Income, Interest Paid, Taxes Paid, Preferred Dividend and Common Dividend were identified for each of the study’s subject firms for calendar years 2000 through 2016 through the Income Statement under the Financials section of FactSet.   FactSet’s reporting of Sales was taken as Sales, Interest Paid as Interest Expense, and Income Taxes as Taxes Paid.

Earnings Before Interest and Taxes (EBIT) and Operating Income were each taken as reported for Earnings Before Interest and Taxes (EBIT). Depreciation and Amortization was taken as the non-cash expenditure for Depreciation & Amortization. Sometimes a company didn’t report a combined line of Depreciation and Amortization, an addition of the two was used to generate the value used in this study.

Market Values of the Firm’s Capital Components

Values for Market Cap Common and Market Cap Preferred were taken found from FactSet’s Capital Structure section under the Overview heading.  In some cases, FactSet reported no preferred stock outstanding, but included an entry for Preferred Dividends Paid.  This was further investigated on www.preferredstockchannel.com and Market Cap – Preferred was taken from that resource.  Market values for each firm’s debt capital were synthesized as noted in the Synthesizing Market Value of Debt section.

Equity Based Firms

Balance Sheet Items

Cash, Investments, Total Assets, Book Value LT Debt, Total Liabilities, and Total Equity were identified for each of the study’s subject firms for calendar years 2000 through 2015 from the Balance Sheet under the Financials section of FactSet.

Interest Income, Non-Interest Income, Depreciation & Amortization, Earnings Before Interest and Taxes (EBIT), Net Income, Taxes Paid, Preferred Dividend and Common Dividend were identified for each of the study’s subject firms for calendar years 2000 through 2015 through the Income Statement under the Financials section of FactSet.   FactSet’s reporting of Interest Income was taken as Interest Income, Non-Interest Income as Non-Interest Income, and Income Taxes as Taxes Paid.

Earnings Before Interest and Taxes (EBIT) and Operating Income were each taken as reported for Earnings Before Interest and Taxes (EBIT). Depreciation and Amortization was taken as the non-cash expenditure for Depreciation & Amortization. Sometimes a company didn’t report a combined line of Depreciation and Amortization, an addition of the two was used to generate the value used in this study.

Income Statement Items

Interest Income, Non-Interest Income, Depreciation & Amortization, Earnings Before Interest and Taxes (EBIT), Net Income, Taxes Paid, Preferred Dividend and Common Dividend were identified for each of the study’s subject firms for calendar years 2000 through 2015 through the Income Statement under the Financials section of FactSet.   FactSet’s reporting of Interest Income was taken as Interest Income, Non-Interest Income as Non-Interest Income, and Income Taxes as Taxes Paid.

Earnings Before Interest and Taxes (EBIT) and Operating Income were each taken as reported for Earnings Before Interest and Taxes (EBIT). Depreciation and Amortization was taken as the non-cash expenditure for Depreciation & Amortization. Sometimes a company didn’t report a combined line of Depreciation and Amortization, an addition of the two was used to generate the value used in this study.

Market Values of the Firm’s Capital Components

Market values for each firm’s equity and debt capital components were calculated as noted for Asset Based Firms.

1Richard Haskell, PhD (2017), Assistant Professor of Finance, Bill & Vieve Gore School of Business, Westminster College, Salt Lake City, Utah, rhaskell@westminstercollege.edu, www.richardhaskell.net.  Bryce Nieberger (2017), Student Research Associate, Bill and Vieve Gore School of Business, Westminster College, Salt Lake City, Utah, https://www.linkedin.com/in/bryce-nieberger-91571ab7

2 FactSet is a data repository delivering information to investment professionals and investors using publicly available analytics, service, content and technology; copyright FactSet Research Systems, Inc.; www.factset.com

Weighted Average Cost of Capital1

The Weighted Average Cost of Capital (WACC) is calculated as a market oriented opportunity cost for this study through the equation $WACC\,\,=\,\,\frac{E}{V}R_{E}\,\,+\,\,\frac{P}{V}R_{P}\,\,+\,\,\frac{D}{V}R_{D}(1-T)$ in which D represents the market cap of the firm’s common stock, P represents the market cap of the firm’s preferred stock, D represents the market value of the firm’s debt capital, and V represents the firm’s total equity and debt capital such that V = E + P + D.

Both market based and book value based values for WACC are included in the study’s year-end values for 2015 and 2016.  Book-value-based WACC is somewhat of a misnomer in that the firm’s equity capital components and taken at their market values, interacted with observable market opportunity costs, while the firm’s debt capital, cash and cash equivalents are taken at book value.

Market based WACC uses only market values for the firm’s equity and debt capital components and results in a true opportunity cost of capital as of year-end 2015 and 2016.  While book value based WACC is collected for each firm through Bloomberg2, market based WACC is calculated using the following methods.

Market cap of the firm’s common share equity and preferred shares were each described in the Capital Structure under the Overview Section of FactSet3.  If FactSet reported no preferred stock but preferred dividends were paid, then further investigation was conducted on Preferred Stock Channel and Market Cap Preferred was taken from the website.

The CAPM formula, $R_{E} = R_{F} + (R_{M}-R_{F})\beta$,  was used to calculate RE in the WACC equation.   RM was calculated through the use of FactSet’s Industry 10-year Performance Percentage by charting the industry’s performance compared to the S&P 500 from 12/31/2005 to 12/31/2015.  The industry’s performance was then annualized by dividing by 10.  Beta was taken from the Snapshot under the Overview section in FactSet.  The 10 year US treasury yield as of 12/31/2015 is used as RF .

The values for Market Value of Long-Term Debt were synthesized as described in Synthesizinng Market Value of Debt section. The Average Tax Rate was calculated by using firm specific values from the supplemental part of the Income Statement in FactSet and was calculated as  $\frac{Taxes\,\,Paid}{Taxes\,\,Paid\,\,-\,\,Net\,\,Income}$.

1Richard Haskell, PhD (2017), Assistant Professor of Finance, Bill & Vieve Gore School of Business, Westminster College, Salt Lake City, Utah, rhaskell@westminstercollege.edu, www.richardhaskell.net.  Bryce Nieberger (2017), Student Research Associate, Bill and Vieve Gore School of Business, Westminster College, Salt Lake City, Utah, https://www.linkedin.com/in/bryce-nieberger-91571ab7

2 Bloomberg connects investors, decision makers and investment professionals with real-time and historical information for a wide range of domestic and international firm; Bloomberg L.P.; http://www.valueline.com/

3 FactSet is a data repository delivering information to investment professionals and investors using publicly available analytics, service, content and technology; copyright FactSet Research Systems, Inc.; www.factset.com

Synthesizing Market Value of Debt1

The values for Market Value of Long-Term Debt were identified by calculating the number of term loans and bonds issued by the subject firm as 12/31/2015, normalizing each long-term debt instrument into \$1,000 face value units with the total being based on the current amount outstanding.  The average bond duration and average bond rating for the firm’s outstanding bond portfolio were used to determine the value of Yield to Maturity (YTM), or current yield in this study.  The data for this was collected from the DCS Detail under the Credit Analysis section of FactSet2.

The date was set to Dec ’15 although some companies had to be set to Nov ’15 or Jan ’16 based on their fiscal year, only Fiscal Year and Quarter data is available on FactSet.  The listed Amount Outstanding and Reported Coupon Rate was taken as Amount Outstanding and Coupon Rate.  Clicking on the Description of the debt creates a pop-up box that the Maturity Date and Bond Rating was collected as the Maturity Date and Bond Rating respectfully. A buy date was set for 31 Dec 2015.  Time to maturity was found by $\frac{Maturity\,\, Date\,\, -\,\, Buy\,\, Date}{365}$, Excel calculated the numerator in days and a division by 365 transferred it into years.

There was some troubleshooting necessary with Debt Data Collection.  When clinking on the Description of the debt was not available, the maturity date provided in the list was used as the first of the month reported.  When a “Range” would appear in the Coupon Rate or Maturity Date, hovering the mouse curser would display the range.  The higher percentage was used for Coupon Rate.  The higher year was used in Maturity Date as DEC 1, Year.

When there was no coupon rate available the Maturity Date was compared to the dates around it to determine if there was another debt that had a similar Maturity Date. If there was, then that Coupon Rate was used as the missing Coupon Rate.  If there was no Coupon Rate and no Maturity Date, then the overall Bond Rating was used to determine which of the 2-year rates would be used.  For example, if a company had an overall Rating of “AA” then the “AA” Coupon Rate of 1.29% was used and the Time to Maturity was set to 2 years.

 US Corporate and Treasury Bond Yields Year End 20153 Year End 20164 Bond Rating 2 YR 5 YR 10 YR 20 YR 30 YR 2 YR 5 YR 10 YR 20 YR 30 YR A 1.49% 2.47% 3.35% 4.35% 4.33% 1.71% 2.49% 3.17% 4.10% 4.22% AA 1.29% 2.21% 3.03% 4.32% 4.26% 1.52% 2.25% 3.15% 3.89% 4.01% AAA 0.92% 1.90% 2.92% 3.83% 4.13% 1.35% 2.32% 3.12% 3.85% 3.96% BB 3.38% 4.84% 5.70% 6.17% 6.35% 2.91% 4.45% 5.24% 8.02% 7.05% BBB 1.77% 2.94% 4.62% 5.17% 4.92% 1.99% 3.50% 3.99% 5.10% 4.81% US Treasury 0.61% 1.36% 2.04% 2.48% 2.87% 1.47% 1.93% 2.45% 2.79% 3.06%

Using the rates above, the YTM was calculated as $Rate\,\,Below\,\,+\,\,(Rate\,\,Above\,\,-\,\,Rate\,\,Below)\,\, x\,\,\frac{Average\,\,Length\,\,of\,\,Years\,\,-\,\,Year\,\,Below}{Difference\,\, of\,\, years\,\, between\,\, Rate\,\, Above\,\, and\,\, Rate\,\, Below}$.

This equation was set-up to calculate an aggregated YTM by taking the published rates and adding the difference between them with a straight-line addition for the years; i.e. $\frac{10\,\,year\,\,rate\,\,-\,\,5\,\,year\,\,rate}{10\,\,-\,\,5}=\frac{3.35\%\,\,-2.47\%}{5}\,\,=\,\,0.18\% per\,\,year$.

In order to add the correct amount,  $\frac{Average\,\,length\,\,of\,\,years\,\,-\,\,Year\,\,below}{Difference\,\, of\,\, years\,\, between\,\, Rate\,\, Above\,\, and\,\, Rate\,\, Below}$is used to determine the percentage of the difference of  to be added to the Rate Below.  For example, Costco has an average length of years for bonds of 6.15 years and an average rating of A.

The Rate Above would be the 10 year “A” rate of 3.35% and the Rate Below would be the 5 year “A” rate of 2.47%.  The difference of years would be .  Costco’s YTM calculation would be $YTM\,\,=2.47\%\,\,+\,\,\big[(3.35\%\,\,-\,\,2.47\%) \,\,x\,\,\frac{6.15\,\,years\,\,-\,\,5\,\,years}{5\,\,years}\big]\,\,=\,\,2.26\%$.

There were a few companies that had an average length of years above 30, such as 38.99 for Ford Motor Co.  For these companies the 30 year rate for the corresponding average rating was used as the YTM.

1 Richard Haskell, PhD (2017), Associate Professor of Finance, Bill & Vieve Gore School of Business, Westminster College, Salt Lake City, Utah, rhaskell@westminstercollege.edu, www.richardhaskell.net.  Bryce Nieberger (2017), Student Research Associate, Bill and Vieve Gore School of Business, Westminster College, Salt Lake City, Utah, https://www.linkedin.com/in/bryce-nieberger-91571ab7

2FactSet is a data repository delivering information to investment professionals and investors using publicly available analytics, service, content and technology; copyright FactSet Research Systems, Inc.; www.factset.com

3Sources: BondsOnline.com; http://www.bondsonline.com/Corporate_Bond_Yield_Index.php?fa=yield_curve&fileDate=12%2F31%2F2015&sector1=C_AAA&sector2=C_AA&sector3=C_A&sector4=&fa2=Export#

412/31/2016 Corporate Bond yields calculated as the average yield to maturity for a sampling of 3-5 corporate bonds among the S&P 100 and DOW 20 Industrials for each of the ratings and maturities identified.  Individual bond yields collected through Bond Central (https://boltonglobal.com/tag/bond-central/) and The Muni Center (https://www2.themunicenter.com/) by Jonathan McKenzie, CFA Director, Bill & Vieve Gore School of Business

Analyst Consensus Estimates1

Analyst Consensus Estimates were used to forecast future cash flows (CFi) to be used in the valuation models for the subject firms in this study.  Data was captured for each firm in the study by examining each firm’s Estimates data as reported in FactSet2 as of calendar year end 2015 and 2016.  For 2015 the estimated data includes cash flows for 2016 through no later than 2021 (if available); for 2016 the estimated data includes cash flows for 2017 through no later than 2022 (if available).  FactSet’s reporting of Estimates includes industry analyst consensus estimates for each subject firm and is an average of the estimates reported by those analysts covering the respective firms.

The captured data includes information on each of the reported variables in this section of the study, including Operating Income (EBIT), Free Cash Flow (FCF), Current Assets (CA), Current Liabilities (CL),  Total Assets (TA), Interest Expense and Tax Expense in the case of asset based firms and Net Income (NI), Cash Flow From Equity (CFE), EBIT, Total Equity (TE) and Total Assets in the case of equity  based firms.  From these reported data, values for Net Operating Profit Less Adjusted Taxes (NOPLAT) were calculated as EBIT x (1-T), in which T is the average US federal corporate income tax rate potentially payable for the estimated level of EBIT based on the 2015 US Corporate Federal Income Tax Table.

There are a few items worth note with respect to the manner in which these data were captured and employed:

• FactSet’s reporting of Analyst Consensus Estimates for the study variables ranged from as few as three years to as many as six years.
• A maximum of five years of estimates are employed in this study, forming the explicit forecast period in the valuation models.
• When a sixth year estimate is available it is NOT used as the first year of the continuation period. Rather the value used for the first year of the continuation period is estimated by expanding the last year of the explicit forecast period data by one (1) plus the appropriate long-run growth rate.

1 Richard Haskell, PhD (2017), Associate Professor of Finance, Bill & Vieve Gore School of Business, Westminster College, Salt Lake City, Utah, rhaskell@westminstercollege.edu, www.richardhaskell.net.  Bryce Nieberger (2017), Student Research Associate, Bill and Vieve Gore School of Business, Westminster College, Salt Lake City, Utah, https://www.linkedin.com/in/bryce-nieberger-91571ab7

2FactSet is a data repository delivering information to investment professionals and investors using publicly available analytics, service, content and technology; copyright FactSet Research Systems, Inc.; www.factset.com