Cost of Capital  405
target capital structure, and estimate its WACC. For the corporate headquar-
ters cash flows, use a weighted average of the business units’ costs of capital. 
Most of ConsumerCo’s businesses have similar betas in a range of 1.1 to 1.2, 
with resulting WACC estimates between 8.6 and 9.1 percent. An exception is 
the devices business, which is more cyclical at a beta of around 1.5 and a cost 
of capital of 10.1 percent. For ConsumerCo’s customer-finance subsidiary, we 
directly estimated the equity beta of its peers in retail banking at 1.2, leading 
to an estimated cost of equity of 10.5 percent.
Finally, using the debt levels based on industry medians, aggregate the 
business unit debt to see how the total compares with the company’s total 
target debt level.7 Set the headquarters target D/E at a weighted average of 
the business units’ D/Es, as its negative cash flow is reducing the company’s 
overall debt capacity. If the sum of business unit target debt differs from 
the consolidated company’s actual debt, we typically record the difference 
as a corporate item, valuing its tax shield separately (or its tax cost when 
the company is more conservatively financed). Remember that the business 
units’ valuations are based on target, not actual, capital structure.
In ConsumerCo’s case, the resulting aggregate target debt level for 
its business units and finance subsidiary is $3,220 million. That amount is 
above its total current net debt of $2,730 million, or $2,980 million debt, net of ­ 
$250 million excess cash (see Exhibit 19.8). If ConsumerCo held on to its cur-
rent leverage, it would realize a loss in value relative to the value of its parts. To 
estimate this loss, project the lost tax shields from the company’s current 
­below-peer-level leverage into perpetuity at the overall revenue growth rate, 
and discount these at the unlevered cost of equity.8
When you value a company by summing the business unit values, there is 
no need to estimate a corporate-wide cost of capital or to reconcile the busi-
ness unit betas with the corporate beta. The individual business unit betas are 
more relevant than the corporate beta, which is subject to significant ­estimation 
7 The allocation of debt among business units for legal or internal corporate purposes is generally ir-
relevant to the economic analysis of the business units. The legal or internal debt is generally driven 
by tax purposes or is an accident of history (cash-consuming units have lots of debt). These allocations 
rarely are economically meaningful and should be ignored.
8 Recall from Chapter 15 that using the cost of debt to discount tax shields significantly overestimates 
their value. In theory, a company’s unlevered cost of equity is a complex average of the unlevered 
cost of equity of its underlying businesses that changes over time. You can use a simple average of the 
unlevered costs of equity of the underlying businesses as an approximation, as any associated error 
has very small impact on value. Assuming a tax rate of 35 percent and an interest rate of 6 percent, 
the tax shields lost from $490 million in debt below target are $10.3 million for 2020. Assuming future 
tax shield losses would roughly grow in line with overall revenues, and discounting at ConsumerCo’s 
unlevered cost of equity of around 9.5 percent, the loss in value would amount to around $190 million.

406  Valuation by Parts
error and is likely to change over time as the weights of the underlying busi-
nesses in the company portfolio change.9
Testing the Value Based on Multiples of Peers
Whenever possible, triangulate the discounted cash flow results with valua-
tion multiples, following the recommendations made in Chapter 18. For each 
of the company’s segments, carefully select a group of companies that are 
comparable not only in terms of sector but also in terms of return on capital 
and growth. Do not simply take an average or median of the peer group mul-
tiples. Instead, always eliminate outliers with multiples that are out of line 
with their underlying economics, and where possible, estimate a median of 
close peers with similar returns and growth. Furthermore, we recommend 
using NOPAT-based instead of EBITA-based multiples, as the latter can be 
distorted by tax differences across companies.
Exhibit 19.9 shows the EV-to-earnings multiples and underlying ROIC 
and growth for the competitors of ConsumerCo’s branded-products business. 
Eliminated from the sample are two outliers with valuation multiples that 
are far above all other peers and not justified by their growth and return on 
capital. Note how the spread in the EBITA multiples is larger than that of the 
NOPAT multiples because of different company tax rates. NOPAT multiples 
are therefore a more reliable basis for the valuation.
9 The implied cost of capital for ConsumerCo as a whole for each future year is around 8.8 percent. It 
can be derived by backing it out from the sum of the underlying business units’ free cash flows and the 
sum of the discounted values of these free cash flows.
EXHIBIT 19.9  ConsumerCo: Multiples for Peer Branded-Product Companies, 
January 2020
Company
ROIC 2020, 
%
Growth, 2020–2025, 
%
EV/EBITA
EV/NOPAT
Peer    1
31.0
4.7
16.5
22.0
Top-peer
average:
21.0
Peer    2
29.6
4.4
15.8
22.6
Peer    3
28.5
4.5
14.3
20.4
Peer    4
27.1
3.9
12.8
19.1
Peer    5
22.0
3.0
12.0
16.0
Close-peer
average:
15.6
Peer    6
21.1
2.8
10.5
15.7
Peer    7
19.7
2.4
11.0
15.7
Peer    8
19.0
2.1
10.0
15.4
Peer    9
18.5
2.2
9.5
15.1
Peer    10
18.3
4.0
20
26.7
Outliers
Peer    11
9.0
3.4
32
45.7
Overall average
18.0
Excluding outliers
21.3
Including outliers