art2 mckinsey irr
DESCRIPTION
.....TRANSCRIPT
Internal rate of return:
A c autionary tale
Temp ted b y a p ro ject w ith a h ig h in tern al rate
o f return ? B etter ch eck th o se in terim cash
fl o w s ag ain .
M ay b e fi nanc e m anag ers just en jo y liv in g
o n th e ed g e. W h at else w o uld ex p lain th eir
w eak n ess fo r usin g th e in tern al rate o f
return (IR R ) to assess cap ital p ro jects? F o r
d ecad es, fin an ce tex tb o o k s an d acad emics
h av e w arn ed th at ty p ical IR R calculatio n s
b uild in rein v estmen t assump tio n s th at
mak e b ad p ro jects lo o k b etter an d g o o d
o n es lo o k g reat. Y et as recen tly as 19 9 9 ,
acad emic research fo un d th at th ree-q uarters
o f C F O s alw ay s o r almo st alw ay s use IR R
w h en ev aluatin g cap ital p ro jects.1
O ur o w n research un d erlin ed th is p ro cliv ity
to risk y b eh av io r. In an in fo rmal surv ey o f
3 0 ex ecutiv es at co rp o ratio n s, h ed g e fun d s,
an d v en ture cap ital firms, w e fo un d o n ly 6
w h o w ere fully aw are o f IR R ’s mo st critical
d eficien cies. O ur n ex t surp rise came w h en
w e rean aly z ed so me tw o d o z en actual
in v estmen ts th at o n e co mp an y mad e o n th e
b asis o f attractiv e in tern al rates o f return .
If th e IR R calculated to justif y th ese
in v estmen t d ecisio n s h ad b een co rrected fo r
th e measure’s n atural flaw s, man ag emen t’s
p rio ritiz atio n o f its p ro jects, as w ell as its
v iew o f th eir o v erall attractiv en ess, w o uld
h av e ch an g ed co n sid erab ly .
S o w h y d o fin an ce p ro s co n tin ue to d o
w h at th ey k n o w th ey sh o uld n ’t? IR R d o es
h av e its allure, o fferin g w h at seems to b e a
straig h tfo rw ard co mp ariso n o f, say , th e
3 0 p ercen t an n ual return o f a sp ecific
p ro ject w ith th e 8 o r 18 p ercen t rate th at
mo st p eo p le p ay o n th eir car lo an s o r
cred it card s. Th at ease o f co mp ariso n
seems to o utw eig h w h at mo st man ag ers
v iew as larg ely tech n ical d eficien cies th at
create immaterial d isto rtio n s in relativ ely
iso lated circumstan ces.
Ad mitted ly , so me o f th e measure’s
d eficien cies are tech n ical, ev en arcan e,2 b ut
th e mo st d an g ero us p ro b lems w ith IR R are
n eith er iso lated n o r immaterial, an d th ey can
h av e serio us imp licatio n s fo r cap ital b ud g et
man ag ers. W h en man ag ers d ecid e to fin an ce
o n ly th e p ro jects w ith th e h ig h est IR R s, th ey
may b e lo o k in g at th e mo st d isto rted
calculatio n s— an d th ereb y d estro y in g
sh areh o ld er v alue b y selectin g th e w ro n g
p ro jects alto g eth er. C o mp an ies also risk
creatin g un realistic ex p ectatio n s fo r
th emselv es an d fo r sh areh o ld ers, p o ten tially
co n fusin g in v esto r co mmun icatio n s an d
in flatin g man ag erial rew ard s.
W e b eliev e th at man ag ers must eith er av o id
usin g IR R en tirely o r at least mak e
ad justmen ts fo r th e measure’s mo st
d an g ero us assump tio n : th at in terim cash
flo w s w ill b e rein v ested at th e same h ig h
rates o f return .
T h e troub le w ith IR R
P ractitio n ers o f ten in terp ret in tern al rate o f
return as th e an n ual eq uiv alen t return o n a
g iv en in v estmen t; th is easy an alo g y is th e
so urce o f its in tuitiv e ap p eal. B ut in fact,
IR R is a true in d icatio n o f a p ro ject’s
an n ual return o n in v estmen t o n ly w h en th e
p ro ject g en erates n o in terim cash flo w s— o r
1 6 | M c K ins ey on F inanc e | S u m m e r 2 0 0 4
John C. Kelleher
a nd Ju s tin J.
M a c Corm a c k
when those interim cash flows really can be
invested at the actual IRR.
When the calculated IRR is higher than the
true reinvestment rate for interim cash flows,
the measure will overestimate—sometimes
very significantly—the annual equivalent
return from the project. The formula
assumes that the company has additional
projects, with equally attractive prospects, in
which to invest the interim cash flows. In this
case, the calculation implicitly takes credit
for these additional projects. Calculations of
net present value (N PV ), by contrast,
generally assume only that a company can
earn its cost of capital on interim cash flows,
leaving any future incremental project value
with those future projects.
IRR’s assumptions about reinvestment can
lead to major capital budget distortions.
Consider a hypothetical assessment of two
different, mutually exclusive projects, A and
B, with identical cash flows, risk levels, and
durations—as well as identical IRR values
of 4 1 percent. U sing IRR as the decision
yardstick, an executive would feel
confidence in being indifferent toward
choosing between the two projects.
H owever, it would be a mistake to select
either project without examining the
relevant reinvestment rate for interim cash
flows. Suppose that Project B’s interim cash
flows could be redeployed only at a typical
8 percent cost of capital, while Project A’s
cash flows could be invested in an attractive
follow-on project expected to generate a
4 1 percent annual return. In that case,
Project A is unambiguously preferable.
E ven if the interim cash flows really could
be reinvested at the IRR, very few
e x h i b i t 1
Identical IRRs, but very different annual returns
1Compound annual growth rate.
Key assumption: reinvestment rate = IRR Key assumption: reinvestment rate = cost of capital
True return is nearly 5 0 % less b ecause of low er reinvestment rate
Year
C as h flo w s , $ m illio n
IRR
Year
C as h flo w s , $ m illio n
2
5
3
5
4
5
0
– 10
1
5
5
5
$ 5 6million
$ 2 9million
4 1 %
. . . h ow ever, interim cash flow s are reinvested at d ifferent rates
Internal- rate-of- return (IRR) values are id entical for 2 projects . . .
V alu e o f c as h flo w s at y ear 5 if rein v es ted at 41%
Year
C A G R1
C A G R1
2 3 40 1
5
5
41%
41%
41%
41%
4 1 %
5
7
10
14
20
5
5
5
P roject A IRR
2
5
3
5
4
5
0
– 10
1
5
5
5 4 1 %
P roject B
P roject A
V alu e o f c as h flo w s at y ear 5 if rein v es ted at 8 %
Year
C A G R1
C A G R1
2 3 40 1
5
5
8 %
8 %
8 %
8 %
2 4 %
5
5
6
6
7
5
5
5
P roject B
Y ear 5 value of $ 1 0 million investment = Y ear 5 value of $ 1 0 million investment =
Internal rate of return: A cautionary tale | 17
18 | McKinsey on Finance | Summer 2004
practitioners would argue that the value of
future investments should be commingled
with the value of the project being
evaluated. M ost practitioners would agree
that a company’s cost of capital—by
definition, the return available elsewhere to
its shareholders on a similarly risky
investment—is a clearer and more logical
rate to assume for reinvestments of interim
project cash flows (Exhibit 1).
When the cost of capital is used, a project’s
true annual equivalent yield can fall
significantly—again, especially so with
projects that posted high initial IRRs. Of
course, when executives review projects
with IRRs that are close to a company’s
cost of capital, the IRR is less distorted by
the reinvestment-rate assumption. But when
they evaluate projects that claim IRRs of
10 percent or more above their company’s
cost of capital, these may well be
significantly distorted. Ironically, unadjusted
IRRs are particularly treacherous because
the reinvestment-rate distortion is most
egregious precisely when managers tend to
think their projects are most attractive. And
since this amplification is not felt evenly
across all projects,3 managers can’t simply
correct for it by adjusting every IRR by a
standard amount.
How large is the potential impact of a
flawed reinvestment-rate assumption?
M anagers at one large industrial company
approved 23 major capital projects over five
years on the basis of IRRs that averaged
7 7 percent. Recently, however, when we
conducted an analysis with the reinvestment
rate adjusted to the company’s cost of
capital, the true average return fell to just
16 percent. The order of the most attractive
projects also changed considerably. The top-
ranked project based on IRR dropped to
the tenth-most-attractive project. M ost
striking, the company’s highest-rated
projects—showing IRRs of 800, 15 0, and
130 percent—dropped to just 15 , 23, and
22 percent, respectively, once a realistic
reinvestment rate was considered (Exhibit 2).
Unfortunately, these investment decisions
had already been made. Of course, IRRs this
extreme are somewhat unusual. Yet even if a
project’s IRR drops from 25 percent to
15 percent, the impact is considerable.
W hat to d o?
The most straightforward way to avoid
problems with IRR is to avoid it altogether.
Yet given its widespread use, it is unlikely to
be replaced easily. Executives should at the
very least use a modified internal rate of
return. While not perfect, M IRR at least
allows users to set more realistic interim
reinvestment rates and therefore to calculate
a true annual equivalent yield. Even then,
we recommend that all executives who
review projects claiming an attractive IRR
should ask the following two questions.
1. What are the assumed interim-
reinvestment rates? In the vast majority of
cases, an assumption that interim flows can
be reinvested at high rates is at best
overoptimistic and at worst flat wrong.
Particularly when sponsors sell their projects
as “ unique” or “ the opportunity of a
lifetime,” another opportunity of similar
attractiveness probably does not exist; thus
interim flows won’t be reinvested at
sufficiently high rates. For this reason, the
best assumption—and one used by a proper
discounted cash-flow analysis—is that
interim flows can be reinvested at the
company’s cost of capital.
2 . A re interim c ash f lo w s b iased to w ard
the start o r the end o f the p ro jec t? Unless
e x h i b i t 2
A rude surprise
1 Disguised example of large industrial company.
%
200
18 19 20 21 22 23
0
50
100
150
800
850
1Project1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
S tandard IR R calculation M odified IR R with reinvestment at cost of cap ital
Internal rate of return: A cautionary tale | 19
the interim reinvestment rate is correct (in
other words, a true reinvestment rate rather
than the calculated IRR), the IRR distortion
will be greater when interim cash flows
occur sooner. This concept may seem
counterintuitive, since typically we would
prefer to have cash sooner rather than later.
The simple reason for the problem is that
the gap between the actual reinvestment
rate and the assumed IRR exists for a longer
period of time, so the impact of the
distortion accumulates.4
Despite flaws that can lead to poor
investment decisions, IRR will likely
continue to be used widely during capital-
budgeting discussions because of its strong
intuitive appeal. Executives should at least
cast a skeptical eye at IRR measures before
making investment decisions.
T h e a uth o rs w is h to th a n k R o b M c N is h fo r h is
a s s is ta n c e in d ev elo p in g th is a rtic le.
MoF
John Kelleher (J o h n _ K elleh er@ M c K in s ey .c o m)
and Ju s tin M a c C orm a c k (J us tin _ M a c C o rma c k @
M c K in s ey .c o m) are consultants in McKinsey’s
Toronto offi ce.
1 J o h n R o b ert G ra h a m a n d C a mp b ell R . H a rv ey , “ T h e th eo ry
a n d p ra c tic e o f c o rp o ra te fi n a n c e: E v id en c e fro m th e fi eld ,”
D uk e U n iv ers ity w o rk in g p a p er p res en ted a t th e 2001
a n n ua l meetin g o f th e A meric a n F in a n c e A s s o c ia tio n , N ew
O rlea n s (a v a ila b le a t h ttp ://s s rn .c o m/a b s tra c t= 22025 1).
2 A s a res ult o f a n a rc a n e ma th ema tic a l p ro b lem, IR R c a n
g en era te tw o v ery d ifferen t v a lues fo r th e s a me p ro jec t
w h en future c a s h fl o w s s w itc h fro m n eg a tiv e to p o s itiv e (o r
p o s itiv e to n eg a tiv e). A ls o , s in c e IR R is ex p res s ed a s a p er-
c en ta g e, it c a n ma k e s ma ll p ro jec ts a p p ea r mo re a ttra c tiv e
th a n la rg e o n es , ev en th o ug h la rg e p ro jec ts w ith lo w er IR R s
c a n b e mo re a ttra c tiv e o n a n N P V b a s is th a n s ma ller p ro j-
ec ts w ith h ig h er IR R s .
3 T h e a mp lifi c a tio n effec t g ro w s a s a p ro jec t’s fun d a men ta l
h ea lth imp ro v es , a s mea s ured b y N P V , a n d it v a ries
d ep en d in g o n th e un iq ue timin g o f a p ro jec t’s c a s h fl o w s .
4 In teres tin g ly , g iv en tw o p ro jec ts w ith id en tic a l IR R s , a
p ro jec t w ith a s in g le “ b ullet” c a s h fl o w a t th e en d o f th e
in v es tmen t p erio d w o uld b e p refera b le to a p ro jec t w ith
in terim c a s h fl o w s . T h e rea s o n : a la c k o f in terim c a s h
fl o w s c o mp letely immun iz es a p ro jec t fro m th e
rein v es tmen t- ra te ris k .