2.1.4 Continuing Value

Considering that it is

impossible to forecast a company’s free cash flow forever, a CV formula that captures

the company’s value after the explicit forecast period is used (Rosenbaum and

Pearl, 2013). A careful estimate of CV is fundamental for any valuation because

it generally accounts for a large percentage of a company’s total value.

Damodaran (2012)

suggests two methods to calculate CV: through an estimation of liquidation

value or a going concern approach. The first assumes the company will cease

operations at a specific point in the future and sell its accumulated assets.

Nonetheless, it is more frequent to assume that the company will operate

forever, which is the case of the going concern approach. As such, an

assumption must be created regarding the company’s long-term sustainable growth

rate. Koller et al. (2015) suggest that the best estimate is “the expected

long-term rate of consumption growth for the industry’s products, plus

inflation”. Rosenbaum and Pearl (2013) and Damodaran (2012) suggest the

projected nominal GDP growth for the geography the company operates in. Koller

et al. (2015) suggest the key value driver formula:

(9)

where

= NOPLAT

in the first year after the explicit forecast period

g = long-run growth in NOPLAT

RONIC = expected rate of return on new invested capital

The

authors defend that this formula is “superior to alternative methodologies

because it is based on cash flows and links them directly to growth and ROIC”. The

authors also note that the company’s terminal year financial data represents a

steady-state level with stable revenue growth and operating margins since

perpetuity-based formulas rely on parameters that never change.

Finally,

Damodaran (2012) and Rosenbaum and Pearl (2013) bring up the idea of a multiple

approach. This multiple is often based on current trading multiples for

comparable firms. However, this approach disregards the fact that this multiple

will change over the years (Koller et al., 2015). One is implicitly assuming

that the future multiple will be worth the same as today’s multiple,

overlooking changes in growth and return prospects. Damodaran (2012) even

states that using multiples to estimate CV “results in a dangerous mix of

relative and DCF valuation”.

2.1.5 Cost of Capital

To value a company using

enterprise-DCF, FCFF must be discounted at the WACC. The WACC represents the

returns that all investors in a firm, equity holders and debt holders, expect

to earn for investing their capital in a specific business as an alternative of

others with identical risk (Koller et al., 2015):

(10)

where D

= market value of debt

E = market value of

equity

=

pretax cost of debt

=

cost of equity

=

marginal income tax rate

Hereupon, capital

structure enters valuation through the WACC since FCFF should be independent of

capital structure and nonoperating items. By removing the interest tax shield

from FCFF, an accurate picture of historical performance is obtained by solely

focusing on operations (Koller et al., 2015). The after-tax cost of debt is used to include the tax shield in

the WACC. Analysts refer to this as the tax deductibility of debt provided that

it reduces the amount of taxes paid, therefore impacting the cost of debt and

decreasing it by the value of the marginal tax rate.

2.1.5.1 Cost of Equity

The cost of equity is

defined as “the rate of return investors require on an equity investment in a

firm” (Damodaran, 2012). In this dissertation, the capital asset pricing model

(CAPM) is used to estimate ATVI’s cost of equity.

2.1.5.1.1

CAPM

The CAPM is simple and

widely accepted in the financial industry to compute a company’s cost equity (Brealey

et al., 2017 and Copeland et al., 2014). The cost of equity is determined by

“estimating the expect return on the market portfolio, adjusted for the risk of

the company being valued” (Koller et al., 2015). Company-specific risk is

adjusted through the use of beta, which is a measure of how the company’s share

price moves in conjunction with the overall market. The remaining risk, known

as idiosyncratic risk, can be diversified away by holding several securities.

The CAPM postulates that the sum of the

risk-free rate (

) and the security’s beta (

) times the market risk premium

equals the expected return on that security

:

(11)

Note that the expected return is

replaceable with the cost of equity since a firm’s investors will earn the cost

of equity if the firm meets expectations.

2.1.5.1.1.1

Risk-free Rate

For

an investment to be considered risk-free, two requirements must be fulfilled:

the inexistence of default and reinvestment risk (Damodaran, 1999). Fernández

(2004) denotes the importance of matching the maturity of a risk-free security

to each year’s cash flow in the explicit forecast period. However, since using multiple

rates can be troublesome, Koller et al. (2015) suggest the use of a single

yield to maturity (YTM) that best corresponds to the stream of cash flows being

valued. In practical terms, 10-year government bonds yields for U.S.-based firm

valuations are suggested.

2.1.5.1.1.2

Market Risk Premium

The reasoning behind the market risk premium

(MRP) is simple: investors require a compensation for bearing risk. Since

stocks are riskier than treasury bonds, they should compensate investors with

higher returns (Siegel and Thaler, 1997). Nonetheless, there is no general

agreement within the literature about the estimation of the MRP. Koller et al.

(2015) indicate that most practitioners use the historical MRP or a market

implied cost of equity. The first uses historical excess returns on the market

the company operates. The authors recommend adding today’s long-term government

bond rate to the historical average of past returns, since past returns are

heavily affected by the inflation rate prevalent during that period, to reach a

better estimate of MRP. The second approach computes the implied cost of equity

derived from the relationship between current stock prices and aggregate

fundamental performance of a large sample of companies. The authors advocate

that this method is reliable because it embodies up-to-date market prices. The

fundamental formula is derived from Formula 9, such that:

(12)

where Earnings = equity earnings

g = expected growth in

real gross domestic product (GDP)

ROE = expected return

on equity

= cost of equity

Solving the equation for

, gives

the cost of equity. Subtracting the risk-free rate yields the MRP.

Additionally, Damodaran (2017) suggests that

companies with multinational operations are exposed to country risk, thus

making country risk an essential constituent in their valuation. The MRP is

then calculated as the sum of a mature market equity risk premium and an

additional country risk premium based upon the risk of the country in question.

Concluding, although there is no consensus in

estimating the MRP, the evidence suggests its value ranges around 5% (Koller et

al., 2015).

2.1.5.1.1.3

Beta

Beta

measures the tendency of a security’s returns to swings in the market. In order

to estimate the future beta of a company, Koller et al. (2015) suggest the use

of an industry peer median beta instead of relying on individual company’s

betas, since these can be largely influenced by nonrepeatable events. Firms in

the same industry confront the same business risks, so their betas should be

resembling. As a first step, the authors advise the use of the market model to

estimate the “raw” beta for each company in the peer group:

(13)

In

this model, the stock’s return (

) is regressed against the return on the

market portfolio (

). In practical terms, the market

portfolio should be a “value-weighted, well-diversified market portfolio” such

as the S&P 500 index. The use of at least five years of monthly returns is

also advised to reduce systemic biases. After gathering the “raw” betas for the

peer group, each company’s beta should be unlevered because beta is a function

of financial and operational risk. Therefore, to better compare the operational

risk across companies, the effect of leverage must be removed from their betas.

The following formula is used:

(14)

where

=

levered beta

=

unlevered beta

D = market value of debt

E = market value of equity

Afterwards,

a median of the unlevered betas should be computed to ultimately relever the

industry beta to the company’s target capital structure using Formula 14.

As

a last note, some practitioners “smooth” the calculated betas towards the

overall average of all companies. Blume (1975) showed evidence of the fact that

betas are mean-reverting. Therefore, the following adjustment is proposed:

(15)

2.1.5.1.2

Fama-French three-factor Model

Fama and French (1992)

assert that stock’s returns are negatively correlated with the size of a

company (measured by market capitalization) and positively correlated with the

ratio of a firm’s book value to its market value of equity. The authors suggest

that stock’s excess returns

can

be explained by the excess market returns

, the difference between the returns of a

portfolio containing small stocks and another containing large stocks (SMB) and

the difference between the returns of a portfolio composed by high

book-to-market stocks and another by low book-to-market stocks (HML). The

following equation describes the model:

(16)

where

=

coefficient related to market excess return

=

coefficient related to size

=

coefficient related to value

Lastly,

Koller et al. (2015) advert that to best implement the CAPM, an industry beta

is used since “regression results are imprecise”. Since the Fama-French model

has three coefficients and that they are dependent on each other, the industry

beta cannot be generated in a clear way.

2.1.5.2 Cost of Debt

The cost of debt

represents the marginal cost to the firm of issuing new debt. Koller et al.

(2015) present two approaches to estimate this value. The first states that if

the company has investment-grade debt, the YTM on “liquid, option-free,

long-term debt” should serve as a reasonable proxy for the cost of debt. The

second approach relates to credit ratings. For companies whose debt rarely

trades, or for nontraded debt, the YTM on a portfolio of long-term bonds with

the same credit rating as the company’s unsecured long-term debt should be used

as an approximation for the implied YTM on long-term debt.