CP

(National Geographic (Little) Kids) #1

234 CHAPTER 6 The Cost of Capital


First, there is no theoretical guidance as to the correct holding period over which
to measure returns. The returns for a company can be calculated using daily, weekly,
or monthly time periods, and the resulting estimates of beta will differ. Beta is also
sensitive to the number of observations used in the regression. With too few observa-
tions, the regression loses statistical power, but with too many, the “true” beta may
have changed during the sample period. In practice, it is common to use either four to
five years of monthly returns or one to two years of weekly returns.
Second, the market return should, theoretically, reflect every asset, even the hu-
man capital being built by students. In practice, however, it is common to use only an
index of common stocks such as the S&P 500, the NYSE Composite, or the Wilshire


  1. Even though these indexes are highly correlated with one another, using differ-
    ent indexes in the regression will often result in different estimates of beta.
    Third, some organizations modify the calculated historical beta in order to pro-
    duce what they deem to be a more accurate estimate of the “true” beta, where the true
    beta is the one that reflects the risk perceptions of the marginal investor. One modifi-
    cation, called an adjusted beta, attempts to correct a possible statistical bias by adjusting
    the historical beta to make it closer to the average beta of 1.0. Another modification,
    called a fundamental beta, incorporates information about the company, such as
    changes in its product lines and capital structure.
    Fourth, even the best estimates of beta for an individual company are statistically
    imprecise. The average company has an estimated beta of 1.0, but the 95 percent con-
    fidence interval ranges from about 0.6 to 1.4. For example, if your regression produces
    an estimated beta of 1.0, then you can be 95 percent sure that the true beta is in the
    range of 0.6 to 1.4.
    So, you should always bear in mind that while the estimated beta is useful when
    calculating the required return on stock, it is not absolutely correct. Therefore, man-
    agers and financial analysts must learn to live with some uncertainty when estimating
    the cost of capital.


An Illustration of the CAPM Approach

To illustrate the CAPM approach for NCC, assume that rRF8%, RPM6%, and
bi1.1, indicating that NCC is somewhat riskier than average. Therefore, NCC’s
cost of equity is 14.6 percent:

(6-3a)

It should be noted that although the CAPM approach appears to yield an ac-
curate, precise estimate of rs, it is hard to know the correct estimates of the in-
puts required to make it operational because (1) it is hard to estimate the beta
that investors expect the company to have in the future, and (2) it is difficult
to estimate the market risk premium. Despite these difficulties, surveys indicate
that CAPM is the preferred choice for the vast majority of companies.

What is generally considered to be the most appropriate estimate of the risk-
free rate, the yield on a short-term T-bill or the yield on a long-term
T-bond?
Explain the two methods for estimating the market risk premium, that is, the his-
torical data approach and the forward-looking approach.
What are some of the problems encountered when estimating beta?

14.6%.

8%6.6%

rs8%(6%)(1.1)

The Cost of Capital 231
Free download pdf