The Quarterly Review of Economics and Finance 50 (2010) 240–243

Contents lists available at ScienceDirect

The Quarterly Review of Economics and Finance

journal homepage: www.elsevier.com/locate/qref

WACC and free cash flows: A simple adjustment for capitalized interest costs
Axel Pierru a,∗ , Denis Babusiaux b
a
IFP, Economics Department, 228-232 Avenue Napoleon Bonaparte, 92852 Rueil-Malmaison, France
b
IFP, IFP School, France

a r t i c l e i n f o a b s t r a c t

Article history: This paper shows how to value investment projects involving capitalization of interest costs by using the
Received 4 December 2009 standard WACC method. Whenever capitalized interest costs do not immediately generate proportionate
Accepted 11 December 2009 tax shields, one of the assumptions that justify the use of the after-tax weighted average cost-of-capital
Available online 23 December 2009
formula is violated. As an offset to this violation, the project’s free cash flows have to be adjusted. We
here derive and interpret a simple adjustment formula. A numerical illustration is provided.
JEL classification:
© 2009 The Board of Trustees of the University of Illinois. Published by Elsevier B.V. All rights reserved.
G11
G3

Keywords:
Capitalized interest
WACC
APV
Interest tax shield
Capitalization

1. Introduction a consequence, to the best of our knowledge, practitioners3 do not

In this paper, we show how to value investment projects involv-
ing capitalization of interest costs with the standard WACC method.
Discounting free cash flows1 at an after-tax Weighted Average Cost
of Capital (WACC) relies on the assumption that every year the
interest cost immediately generates a proportionate tax shield. This
assumption is in general not valid when some interest costs are
not paid but capitalized. For instance, in many countries capital-
ized interest costs are depreciated according to the same rule as
that applied to the project’s capital expenditures, and they there-
fore generate deferred tax shields. Surprisingly enough, this issue
has been so far ignored2 by the corporate-finance literature, which
results in the absence of a well-founded methodology for the treat-
ment of capitalized interest in the context of project valuation. As
∗ Corresponding author. Tel.: +33 147526408; fax: +33 147527082.
E-mail address: axel.pierru@ifp.fr (A. Pierru).
1
The term “free cash flow” (also called “after-tax operating cash flow”) refers to
the cash flow of the project before any financial claims are paid. For tax purposes,
the taxable income used is defined as the earnings before interest and taxes, which
means that the free cash flow includes no interest tax shields.
2 6
Here we do not deal with accounting issues such as those studied by Bowen,
Noreen, and Lacey (1981) and Peasnell (1993). A possible explanation for this
absence of literature is that adjusting for capitalized interest will in general have a
small impact on a project’s value. However, the adjustment formula we propose is
easy to apply and leads to a rigorous calculation of the project’s value.
adjust free cash flows for capitalized interest.
We here consider a firm that sets4 a target debt-to-value ratio on
the corporate scale, for all projects in the same risk class. The firm’s
WACC is calculated with this target debt ratio. Investment projects
are valued by discounting their free cash flows at this WACC value.
In practice, an “apparent loan5 ” – and its corresponding repayment
schedule – may be attributed to a project involving high capital
expenditures. Although contracted to finance the project consid-
ered, this loan is guaranteed by the firm, consolidated with other
corporate-finance loans and included6 in the calculation of the debt
ratio targeted by the firm. When capital expenditures are spread
3
For example, the Asian Development Bank’s guidelines – titled “Financial man-
agement and analysis of projects” – describe the free cash flows as “excluding any
financing flows such as interest on debt and other financing charges during construc-
tion” (see http://www.adb.org/documents/guidelines/financial/part030402.asp).
4
Graham and Harvey (2001) report that about 80% of firms have some form of
target debt-to-value ratio, and that the range around the target is tighter for larger
firms.
5
This term is used by Pierru (2009) who mentions capital leases, subsidized loans
and loans associated with oil and gas projects for fiscal purposes as other instances
of apparent loans.
The project’s actual contribution to the firm’s debt capacity, equal to the firm’s
target debt ratio times the project’s value, is likely to differ from the amount of this
apparent loan. As emphasized by Pierru (2009), one should consider that the firm
compensates a positive (negative) difference by issuing more (less) corporate bonds
or by increasing (decreasing) the amount of another project’s apparent loan.

1062-9769/$ – see front matter © 2009 The Board of Trustees of the University of Illinois. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.qref.2009.12.005
 A. Pierru, D. Babusiaux / The Quarterly Review of Economics and Finance 50 (2010) 240–243 241

over several years, as is often the case for big projects, the repay- • a component Bn proportional to Vn , with B0 + C = wV0 and
ment schedule of this apparent loan may involve capitalization of Bn = wVn (n ∈ {1, . . . , T − 1}), whose amount is, to some extent,
interest costs, for instance until the start of production. This is typi- exposed to the firm’s operating risk.
cally the case when debt covenants allow the payment of interest to
begin only when the project starts to produce. Until this date, the Let us first compute V0 which is equal to the present value of free
interest costs produced every year are added to the outstanding cash flows Fn (n ∈ {1, . . . , T }) plus the present value of the interest
loan amount (i.e., they are capitalized) and are, in general, subject tax shields generated by Bn (n ∈ {0, . . . , T − 1}). In a Miles–Ezzell’s
to a special fiscal treatment. world, due to the expectation revision process, the interest tax
As corporate-finance theory recommends, whenever debt shields expected in year k are exposed to the operating risk during
financing is susceptible to have a special impact, we first derive a k − 1 years. According to Miles and Ezzell (1985), in the absence of
valuation formula within a standard Adjusted Present Value (APV) capitalized interest costs, we therefore have:
framework, by considering a Miles–Ezzell’s world (1985) where
interest costs are certain over one period. The resulting adjustment 
T
Fk rBk−1
of free cash flows is then interpreted and discussed. For practical Vn = + , n = 0, . . . , T − 1
(1 + )k−n (1 + r)(1 + )k−n−1
purposes, when the standard WACC method is used, a simple but k=n+1
consistent formula is proposed. A numerical illustration is given in (1)
the last section.
According to (1), Vn and Vn+1 are linked by the following equation
(with VT = 0):
2. Valuation formula with an APV approach Vn+1 + Fn+1 rBn
Vn = + , n = 0, . . . , T − 1 (2)
1+ 1+r
Let us consider a firm expecting the free cash flow Fn in year n
(n ∈ {0, 1, . . . , T }). The firm’s unlevered equity cost, relating to its Since we have Bn = wVn (n ∈ {1, . . . , T − 1}) (2) gives:
operating risk, is . All the debt, assumed free of default risk, is
Vn+1 + Fn+1
contracted at the risk-free interest rate r. According to the APV, the Vn = , n = 1, . . . , T − 1 (3)
1 +  − wr((1 + )/(1 + r))
firm’s value is equal to the present value of its free cash flows plus
that of all tax shields generated by interest costs. By recurrence (3) immediately gives:
There are here two types of tax shield: those generated by inter-
est payments and those generated by the depreciation of unpaid 
T
Fk
(i.e., capitalized) interest costs. We assume that the interest pay- Vn = (4)
k−n
(1 +  − wr((1 + )/(1 + r)))
ments are deductible from the project’s taxable income which is k=n+1
subject to the tax rate . On the other hand, capitalized interest
The denominator in the right-hand side of (4) is the adjusted-
costs are assumed to be depreciated. For ease of notation, we first
discount-rate formula derived by Miles and Ezzell (1985) (see also
assume7 that interest costs are capitalized over 1 year only (in year
Taggart (1991) and Brealey and Myers (2003)) for the firm’s WACC.
1) and paid from year 2 on. Let C be the portion of the firm’s debt
For ease of notation, let us denote this adjusted discount rate as i:
(contracted in year 0) whose interest costs rC in year 1 are cap-
italized. The amount of capitalized interest depreciated in year n
1 + 
i =  − wr (5)
(n ∈ {2, . . . , T }) is denoted as Dn (rC), with a resulting depreciation 1+r
tax shield of Dn (rC).
The firm is assumed to target a debt-to-value ratio w at By combining (4) and (5), we obtain:
the corporate scale every year. We here follow the Miles–Ezzell

T
Fk
analysis (1980, 1985): the current debt level, which is based V1 = (6)
k−1
on firm’s current value, is known, so in the absence of default (1 + i)
k=2
risk all the interest costs – including rC – at the end of year 1
are certain. We consequently consider that the depreciation tax Since we have: B0 + C = wV0 (2) gives in year 0:
shields Dn (rC)(n ∈ {2, . . . , T }) are also certain and should there-
V1 + F1 r(wV0 − C)
fore be discounted at the rate r. This assumption is discussed V0 = + (7)
later. 1+ 1+r
Every year n, the firm’s value can therefore be broken down into By combining (7), (5) and (6), we finally have:
a part Vn that is subject to the (operating) free cash flow risk and a
part V̄n relating to the risk-less capitalized interest depreciation tax rC  Fn T

shields. Targeting the debt ratio w therefore results in a linear debt V0 = − + n (8)
1 + (1 − w)r (1 + i)
policy8 where in every year n (n ∈ {0, . . . , T − 1}) the firm’s debt is n=1
the sum of:
Let us now determine V̄0 which is equal to the present value of the
capitalized interest depreciation tax shields plus the present value
• a component B̄n , equal to wV̄n , whose amount (and corresponding of the interest tax shields generated by B̄n . As all these tax shields
interest payment) is certain, are risk-less, we have:

r B̄0  Dn (rC) + r B̄n−1

T
V̄0 = + (9)
7
This assumption is relaxed later in the paper.
1+r (1 + r)n
n=2
8
Ruback (2002) describes a linear debt policy as including a fixed component
(whose outstanding amount is directly targeted and interest tax shield is therefore 
T
Dk (rC) + r B̄k−1
certain) and a proportional-to-value component (whose interest tax shield is subject
V̄n = , n = 1, . . . , T − 1 (10)
to the firm’s operating risk). Here B̄n is equivalent to a fixed debt component since (1 + r)k−n
its amount is defined as proportional to the value of riskless cash flows. k=n+1
242 A. Pierru, D. Babusiaux / The Quarterly Review of Economics and Finance 50 (2010) 240–243

(9) and (10) are equivalent to the following recurrence (with V̄T = taxable income is positive. This shows that these tax shields are,
0): to some extent, subject to the operating risk. Moreover, we have
derived the valuation formula (17) in a Miles–Ezzell’s world where
V̄n+1 + Dn+1 (rC) + r B̄n
V̄n = , n = 1, . . . , T − 1 (11) interest tax shields are certain over one period. Alternatively, we
1+r could have considered a Harris–Pringle’s world10 (1985) where the
V̄1 + r B̄0 target debt ratio is instantaneously satisfied and interest tax shields
V̄0 = (12)
1+r always uncertain. In such a world, there would be no adjusting
factor in the first year.
Since B̄n = wV̄n in every year n (11) and (12) give respectively:
All these considerations imply that in every year n the relevant
V̄n+1 + Dn+1 (rC) adjusting factor lies certainly somewhere between 1 (i.e., no adjust-
V̄n = , (n = 1, ...., T − 1) (13) n
1 + (1 − w)r ing factor) and ((1 + )/(1 + r)) . To determine its true value seems
a very difficult – if not impossible – task. For this reason, we suggest
V̄1
V̄0 = (14) that this adjusting factor should simply be set equal to one, since
1 + (1 − w)r the resulting valuation formula (18) has then a clear interpretation
The following formula is immediately deduced from (13) and (14): and can be very easily implemented:


T
Dn (rC) F1 − rC  Fn + Dn (rC)
T
V̄0 = n (15) F0 + + n (18)
(1 + (1 − w)r) 1+i (1 + i)
n=2 n=2

According to (8) and (15), the net present value generated by the (18) has a straightforward interpretation: discounting the firm’s
firm, equal to F0 + V0 + V̄0 , is therefore: free cash flows at the rate i relies on the assumption that every year
 T  the firm’s total interest cost11 generates an interest tax shield at

T
Fn  Dn (rC) rC
+ rate , whereas the interest cost rC does not produce any tax shield
n n − (16)
(1 + i) (1 + (1 − w)r) 1 + (1 − w)r in year 1. Consequently, to remain consistent with the WACC, the
n=0 n=2
tax shield lost rC is subtracted from the free cash flow of year 1.
Since according to (5) we have 1 + (1 − w)r = ((1 + r)/(1 + ))(1 + On the other hand, the tax shields generated by the depreciation of
i), formula (16) is equal to: the capitalized interest costs are added to the firm’s free cash flow.
F1 − ((1 + )/(1 + r))rC For projects where capitalized interest costs produce no tax
F0 + shields (i.e., Dn (rC) = 0 in every year n), the adjustment simply con-
1+i
sists in subtracting the lost interest tax shield rC from the free cash

T n
Fn + ((1 + )/(1 + r)) Dn (rC) flow of year 1.
+ n , (17)
(1 + i) Note that the free cash flow adjustment only accounts for the
n=2
fiscal treatment of capitalized interest. In countries where, for fis-
where i is the firm’s WACC. cal purposes, capitalized interest costs are treated like interest
payments (i.e., they are directly deducted from the firm’s taxable
3. Interpretation, discussion and proposal of a simple free income), no adjustment is required. Since the target debt ratio has
cash flow adjustment to be satisfied, every dollar of interest capitalized simply replaces a
dollar of debt that otherwise would have been contracted. Paying
The valuation formula (17) has been derived within a standard $1 of interest and borrowing $1 of debt simultaneously is equiva-
APV framework. A crucial assumption we have made is that the lent to capitalizing this dollar. The capitalization of interest costs
tax shields generated by the depreciation of capitalized financial has then no impact on the firm’s value.
charges are riskless and have therefore to be discounted at the By following this simple interpretation (18) can be easily
interest rate r. As a result, in the net present value formula (16), extended to the usual real-life case where no interest payment
all cash flows relating to capitalized interest costs are discounted occurs before the start of production. Let us consider that, for the
at a rate that is smaller than the firm’s WACC. Equivalently, in for- project under consideration, construction requires t years and pro-
mula (17) all cash flows are discounted at the firm’s WACC, but, duction starts in year t + 1. All interest costs generated until year t
in every year n, the cash flow adjustment relating to capitalized are therefore capitalized. Let Cn (n = 0, . . . , t − 1) be the outstand-
n
interest costs includes the adjusting factor ((1 + )/(1 + r)) . ing debt in year n whose interest costs in year n + 1 are capitalized
This assumption, although consistent with the literature on APV (for ease of notation we use the notation C−1 = 0). The project’s net
(e.g., Myers (1974), Taggart (1991) and Ruback (2002)), can be dis- present value is then as follows:
cussed. For instance, one could argue here that all depreciation

t

T t−1
tax shields have the same risk, whatever their origin9 (investment Fn − rCn−1 Fn + k=0
Dn,k (rCk )
n + n , (19)
outlay or capitalized financial charges). If we follow this line of rea- (1 + i) (1 + i)
n=0 n=t+1
soning, formula (17) then operates an artificial distinction between
the tax shields generated by the depreciation of the investment where Dn,k (rCk ) is the depreciation amount in year n resulting from
(included in the free cash flow and therefore discounted at the the capitalization of interest costs rCk in year k + 1.
unlevered cost of equity) and those generated by the depreciation The project’s free cash flows are here adjusted to capitalized
of the capitalized interest costs (discounted at a smaller rate). In interest in a very simple way: every year during construction, lost
addition, from a marginal perspective, the depreciation of capital-
ized interest costs will result in actual tax shields only if the firm’s
10
As shown by Taggart (1991), the adjusted-discount-rate formula derived by
Harris and Pringle (1985) for the firm’s WACC is a continuous-time version of that
9
This is also consistent with the observation that capitalized financial charges derived by Miles and Ezzell (1985).
11
are often depreciated according to the same rule as the corresponding investment i.e., the interest produced by the loan amount that satisfies the debt ratio tar-
outlays. geted by the firm (with respect to the firm’s value).
 A. Pierru, D. Babusiaux / The Quarterly Review of Economics and Finance 50 (2010) 240–243 243

Table 1
Lost interest tax shields, depreciation tax shields and adjusted free cash flows ($ millions).

Year, n 0 1 2 3–7
Free cash flow, Fn −100 −150 −150 80
Loan contracted in year n −50 −75 −75
Outstanding debt whose interest costs are capitalized, Cn 50 128
Capitalized interest costs in year n, rCn−1 3 7.68
Lost interest tax shields in year n, −rCn−1 (n = 1, 2) −1.2 −3.07

1

Tax shield from depreciation of capitalized interest costs in year n, Dn,k (rCk ) = ((rC0 + rC1 )/5) (n = 3, . . . , 7) 0.85
k=0
Adjusted free cash flow −100 −151.2 −153.07 80.85

interest tax shields are deducted from the free cash flow, and, every According to formula (19), the present value of the adjustments for
year during production, tax shields generated by the depreciation capitalized interest costs is:
of capitalized interest are added to the free cash flow. In both
0.4 × 0.06 × 128  0.4 × 2.14
7
cases, the adjustment made compensates for a departure from the 0.4 × 0.06 × 50
− − +
assumption that every year the firm’s total interest cost immedi- 1.12 1.122 1.12n
n=3
ately generates a proportionate tax shield, since this assumption
is implicitly made when the WACC is used. The valuation formula = −$1.06 million
(19) therefore extends the field of application of the standard WACC
The project’s net present value is therefore:
method.
153.07  80.85  80
7 12
151.2
4. Numerical illustration −100 − − + + = $5.77 million
1.12 1.122 1.12n 1.12n
n=3 n=8
A firm is studying a project of which capital investments are
spent over 3 years: $100 million in year 0, $150 million in years 1
References
and 2. The project’s free cash flow is assumed to remain constant
and equal to $80 million every year from year 3 (start of produc- Bowen, R. M., Noreen, E. W., & Lacey, J. M. (1981). Determinants of the corpo-
tion) to year 12 (end of the project’s life). Every year (from year rate decision to capitalize interest. Journal of Accounting and Economics, 3,
0 to year 2), the project’s investments are financed up to 50% by 151–179.
Brealey, R. A., & Myers, S. C. (2003). Principles of corporate finance (7th ed.). New
loans contracted at the annual interest rate r = 6%. Until the end of York: McGraw-Hill.
year 2 (here t = 2), interest costs are capitalized and, for fiscal pur- Graham, J. R., & Harvey, C. R. (2001). The theory and practice of corporate finance:
poses, depreciated in a straight-line way from year 3 to year 7. The evidence from the field. Journal of Financial Economics, 60, 187–243.
Harris, R. S., & Pringle, J. J. (1985). Risk-adjusted discount rates—Extensions from the
interest costs produced from year 3 on are paid and (immediately)
average-risk case. Journal of Financial Research, 8, 237–244.
deducted from the project’s taxable income (subject to a tax rate Miles, J., & Ezzell, J. R. (1980). The weighted average cost of capital, perfect capital
 = 40%). All these loans are consolidated with the firm’s debt and markets and project life: A clarification. Journal of Financial and Quantitative
Analysis, 15, 719–730.
are hence included in its target debt ratio calculation. The WACC
Miles, J., & Ezzell, J. R. (1985). Reformulating tax shield valuation: a note. Journal of
used by the firm to discount its free cash flows is i = 12%. All these Finance, 40, 1485–1492.
data are expressed in nominal dollars. Myers, S. C. (1974). Interactions of corporate financing and investment
Table 1 gives the project’s free cash flows adjusted for cap- decisions—Implications for capital budgeting. Journal of Finance, 29, 1–25.
Peasnell, K. V. (1993). Capitalisation of interest. British Accounting Review, 25,
italized interest costs, as well as all intermediate calculations. 17–42.
The capitalized interest costs amount to $3 million in year Pierru, A. (2009). The weighted average cost of capital is not quite right: A rejoinder.
1 (with C0 = $50 million) and $7.68 million in year 2 (with Quarterly Review of Economics and Finance, 49, 1481–1484.
Ruback, R. S. (2002). Capital cash flows: A simple approach to valuing risky cash
C1 = 50 + 75 + 3 = $128 million). The resulting annual depreciation flows. Financial Management, 31, 85–103.
amount (from year 3 to year 7) is therefore: Taggart, R. A. (1991). Consistent valuation and cost of capital expressions with cor-
porate and personal taxes. Financial Management, 20, 8–20.
3 + 7.68
= $2.14 million
5