Capital budgeting in organizational context

By Hyoung Goo Kang

Abstract: This paper suggests an organizational theory of capital budgeting

that admits the traditional financial capital budgeting model as a special
theory and that understands capital budgeting as an organizational phenomenon.
Therefore, the organizational theory not only broadens the traditional model,
but also explains the behaviors of firms using quasi/non-financial version of
capital budgeting. I demonstrate the validity of the model with ethnographic
data. The ethnography supports and enriches the organizational model. The
model in turn contextualizes the ethnographic data.

Key words: Capital budgeting, controversy, internal capital market,

organization, uncertainty

This paper is short-version of my thesis. Full thesis is available at

http://hyoungkang.googlepages.com/ocbm.html

1
Introduction
Dot-com bubble in late 90s was a fascinating period of valuation and capital
budgeting. Many internet firms had no profit despite their surging stock
prices. Traditional valuation in which analysts assess firms with expected
cash flows or earnings hardly rationalize the stock prices. Initially,
analysts began relying on multiples, revenue-based model and non-financial
metrics slightly extending traditional models. Shortly, innovative valuation
methods flourished as the analysts match their models with the rising stock
prices ex-post. Mary Meeker, a star analyst in Morgan Stanley Dean Witter,
dubbed Queen of the Net, was one of the well-known innovators. She valued
internet firms with eyeball, customer database and pages views. Other
analysts and venture capitalists were as creative as she was. They developed
metrics based on clicks, customer lifetime values, unique visitors, etc. (The
Wall Street Journal Nov 22, 1999; Fortune May 14, 2001). Those innovative
valuations have only loose relation with standard valuation approach. They
are more intuition-based than number driven. With the support of the novel
valuations, large investments in the dot-com sector had occurred. Webvan, a
bankrupt online grocery firm, is a dramatic example of such valuation and
capital budgeting.

We can interpret such innovative valuations in two ways: either manipulative

or sincere valuation. First, stock analysts might have intentionally mis-
assessed internet firms. The analysts had incentives to assist IPO deals with
favorable appraisals because of fee structure and coverage. Such concern
about the conflict of interest leads to The Global Settlement between top
investment banks and U.S. regulatory agencies (Apr 23, 2003). Second, the
innovative valuation might have been the sincere effort of analysts under
uncertainty and controversy. To support, Mary Meeker had not been charged for
her creative valuations. Similarly, economists had the related controversy
about New Economy (Gordon 2000). As New Economy was the frame for certain
economists to understand the miracle of U.S. economy, the innovative
valuation methods could be the frame for stock analysts to understand the
boom in internet sector.

Judging from the legal cases and academic debates, I suggest that the
creative valuations during dot-com bubble reflect both manipulative motives
and sincere efforts of stock analysts. This poses some challenges to
traditional valuation models. If the innovative valuation was the rational
behaviors of analysts, does it mean that the traditional models had broken
down? What valuation models can explain the valuation practices in general?

The Internet boom and coincident productivity growth in the late 1990 were
new and puzzling phenomenon. For instance, Lawrence Summers called them
Paradigm Uncertainty (Business Week April 10, 2000, Gordon 2000). In this
situation, the creative valuation might have been the useful frames to
conduct research about growing sectors and to understand the uncertain
environments. In addition, the valuation frames might have been the useful
communication tools to obtain the legitimacy of research results in face of
controversial problems. It is doubtful whether market would have accepted
negative NPV number for internet stocks as legitimate during the late
nineties. Few people may have thought the values of surging internet
companies are zero.

My research question is to understand valuation and capital budgeting

behaviors. I investigate the situation in which agents rationally choose

2
valuation methods. This setting contrasts the more conventional view that NPV
is the optimal valuation tool and should be used during capital budgeting
process. I argue that valuation methods are the frames through which actors
perceive objects such as investment opportunities. NPV is only one of the
frames, and its optimality is not guaranteed. This separation between frame
and objects require us to consider valuation and capital budgeting behaviors
in substantially novel manner, but helps us to understand them from general
perspective.

I propose that the standard capital-investment and capital budgeting model

(NPV and its variations; SCBM hereafter) are effective to account observed
practices in special cases only. SCBM offers very specific theories about
what is the best practice of investment-decision making (e.g. Brealey and
Myers 2002, Brealey, Myers and Allen 2005). For instance, the titles of
Chapter 5 and 6 of Brealey et al (2005) are: "Chapter 5: Why Net Present
Value Leads to Better Investment Decisions than Other Criteria" and "Chapter
6: Making Investment Decisions with the Net Present Value Rule." Nevertheless,
while SCBM presumes NPV and its variations as normative and common practices,
the real capital allocation is much richer and complicated. Bower (1970,
1986)’s book about resource allocation process is a good example. More
generally, the decision-making processes as important as capital allocation
are quite subtle as abundant literature about organizational decision-making
suggests (e.g. Burton, DeSantics & Obel 2006).

This paper investigates what theory of capital budgeting would explain the
resource allocation patterns in internal capital market. I aim to introduce a
new concept: Organizational Capital Budgeting Model (OCBM). OCBM is a general
theory of capital budgeting that admits the traditional financial capital
budgeting model as a special theory and that understands capital budgeting as
an organizational phenomenon. Therefore, OCBM intends not only to broaden the
traditional model, but also to explain the behaviors of firms using
quasi/non-financial version of capital budgeting. The unit of analysis is
method. I will demonstrate the validity of OCBM with qualitative studies.

Standard capital budgeting model (SCBM)

The attractiveness of a project is Value := E(Cash) - Rd*Inv, using forward
notation. Rd denotes opportunity cost of capital. Inv is the size of
investment. Since risky projects are less desirable, we should reduce E(Cash),
increase Rd, increase Inv or the combination of them as risk increases.

Decrease E(cash): We can decrease the expected future cash term by changing
probability distribution of future cash stream. This approach is similar to
risk-neutral probability measure approach, which is particularly useful in
derivative pricing.

Increase Rd: We can adjust the required rate of return by modeling it as the
increasing function of risk. Weighted average cost of capital (WACC) is the
stylized way to incorporate risks. WACC is the linear combination of the
opportunity costs of equity, debt and other securities. The ratio of the
costs is a function of capital structure and tax rate. The costs of equity
and debts are modeled with single/multi-factor models. CAPM is an example.

Increase Inv: Alternative way is to transform the amount of capital invested.

Inv is modified by adding capital charge to the true amount of investment in
response to risk. The higher risk is, the higher the capital charge is. This

3
method is particularly popular among financial institutes to meet the capital
requirement stipulated in regulations. Basel II is the industry standard.
Basel II agreement requires capital to be over 8% of risk-weighted assets
plus 12.5 x (capital charges) for market and operation risk. Risk-weighted
asset is (exposure) x (risk weight). Thus, as the risk of asset increases,
the capital requirement increases. The capital charge denotes the difference
between capital requirement and current capital.

Above characterization covers standard capital budgeting models (SCBM). SCBM

has important characteristics. First, decision-making is straightforward.
Given investment opportunities, SCBM is applied to compute the value. Second,
dichotomy between investment opportunity and decision-making process is
assumed. The functional form of SCBM remains essentially same irrespective of
investment opportunities or the contexts in which the valuation occurs. The
first and second are related. The decision-making is trivial because
investment opportunities are exogenously given, and SCBM is the formula for
the investment.

Real option technique is also the part of SCBM. While real option technique
involves decision-making over the course of projects, it assumes the
dichotomy also. First, Abel, Dixit, Everly and Pindyck (1996) show that one
can apply NPV instead of real option technique if a decision maker accounts
for future marginal returns to capital at the future optimal levels of the
capital stock. Second, the decision-making process is still trivial and
mechanical in that the value of underlying investment opportunities follow
some exogenous processes, and a decision maker exercises the real option (i.e.
investment) when the underlying value hits some critical levels.

Value space
I modify SCBM by expanding the choice set from investment opportunities to
value space. Value space is two-dimensional space of investment opportunities
and valuation strategies. Thus, I suggest firms make choice not only about
investment opportunities, but also about valuation methods.

The notion of value space touches one of the fundamental issues about
knowledge. Valuation method and investment opportunities are frames and
objects respectively in epistemology. I will use the terms, frame and object,
in this section in order to stress the link between value space and
epistemology. Epistemology is a branch of philosophy and studies the nature
of knowledge, truth and belief. Value space idea implies that actors perceive
objects with frames. The frames can be a deliberate choice of the actors or
be enforced. I take an intermediate view. There is a boundary of frames,
which we can model with the set of valuation strategies. The set is
exogenously given, but actors choose a subset of frames in the given set. If
the set is singleton, there is no freewill in the choice. If the set includes
all the possible frames, there exists complete freedom. The value space is
the space spanned with frames and objects. Value space is the central concept
to build a general model of capital budgeting and to overcome SCBM.

Value space presumes a strong version of constructive epistemology since it

can allow environmental variables to influence the choice of frames in order
to perceive reality. Berger and Luckmann (1966) are well-known proponents of
social constructionist argument. However, my value space idea is subtly
different from theirs discussed in their 1966 book, The Social Construction
of Reality. The book argues that social relations construct knowledge. Social

4
interactions result in institutionalization, the process of which embeds
meaning in society. Since people live in the society covered with the layers
of institutions, they perceive socially constructed reality.

Value space idea is more abstract than that of Berger and Luckmann. Since it
allows people to select frames, it includes the neoclassical economist’s view
on choices. It is possible that people choose the frames of perceiving object
in order to optimize their goals such as preference. Thus, value space idea
incorporates the views of both neoclassical economists and sociologists like
Berger and Luckmann. If the institutional restriction on the set of frames is
strong, value space idea converges to the sociologists’ views. If the
restriction is weak enough, it becomes neoclassical views. Of course, it is
not entirely rigorous to distinguish sociologists and economists perspectives
only with the restriction on the set of frames. For instance, standard
financial economics argues that NPV is the optimal frame for valuation. It
regards other methods as suboptimal. In that sense, the standard view imposes
strong restriction on the set of frames. However, since the choice of NPV
does not involve any social influences explicitly, NPV is not sociological.

To summarize, value space is an important concept in valuation. It stresses

that firms make choice in two-dimensional space of valuation strategies and
investment opportunities. The key idea of value space is the separation of
objects and frames, in which actors observe objects through the choice of
frames. Since objects and frames can be correlated with each other, value
space rejects the assumption of the dichotomy between objects and frames. In
this sense, the intuition of value space becomes similar to The Social
Construction of Reality (Berger & Luckmann 1966) as social factors influence
the choice of frames.

Social factors and valuation

Simple model can illustrate how value space idea is operationalized and how
social factors determine the knowledge about investment. Suppose there is a
project VI. If two agents (A1, A2) conduct VI, the project generates positive
value V, V = VH + VL in all cases. Thus, the project is riskless in aggregate.
However,(A1, A2) individually perceive VI risky. Two states {w1, w2} occur with
probability 1/2 each. If w1 occurs, the payoff to (A1, A2) is (VH, VL). If w2
occurs, the payoff to (A1, A2) is (VL, VH).

(state, probability) (w1, 1/2) (w2, 1/2)

Payoff to (A1, A2) (VH, VL) (VL, VH)

VH is positive, but VL is negative as:

VH > (VH + VL)/2 > 0 > VL.

I do not assume that (A1, A2) can contract on the realization of VI. Let us
view this outcome distribution as the typical payoff matrix in game theory.
Thus, {w1, w2} is possibly observable, but not contractible. In this setting,
risk neutral (A1, A2) conduct the project because the expected value is
positive to both of them. Risk neutrality occurs when A1 and A2 have risk
neutral preference or when the realization of {w1, w2} is uncorrelated with
systematic factor. In addition, because the project generates positive value
in aggregate without risk, it is better to execute the project (VI) from the
view of social planner.

5
Valuation method

(A1, A2) radically change their behaviors once we introduce finer valuation
methods. Coarse method cannot distinguish between {w1, w2} and can compute
only the expected values for each player, (VH + VL)/2. I set is as default
method. Finer method can distinguish the states {w1, w2}. Using the finer
method, (A1, A2) know whether they are in w1 or w2.

The finer method makes VI discarded. With finer method, if the state is w1, A2
rejects VI. If w2 occurs, A1 rejects it. Thus, VI is unimplemented in all
states when (A1, A2) uses finer method.

Next, suppose (A1, A2) have the option to choose fine method. Let us do
backward induction. If (A1, A2) choose fine method, their expected value is
zero because one of them rejects project in all states. If they stick to
coarse one, their expected value becomes (VH + VL)/2 > 0. Therefore, (A1, A2)
will choose coarse method and conduct VI.

This behavior illustrates how actors can construct the knowledge about the
investment opportunities under social relation. The actors can construct the
knowledge themselves because they choose valuation method. They choose the
method under the influence of social relation specified with {VH, VL}. (A1, A2)
intentionally stay ignorant of the state although it is costless to learn it.
We can stipulate the knowledge of (A1, A2) using information set. (A1, A2)
construct their knowledge about the project by choosing the set {∅, {w1, w2}}
instead of {∅, {w1}, {w2}, {w1, w2}}.

Method Knowledge Expected payoff Choice

Fine {∅, {w1}, {w2}, {w1, w2}} 0
Coarse {∅, {w1, w2}} + V

For further illustration, let us introduce the incentive to learn. If A1 and

A2 invest in a valuation method X, they can ex-ante distinguish state w1 and w2.
c(X) is the cost of the method to (A1, A2) respectively. If c(X) < 0, we can
interpret it as a subsidy to conduct the method X.

(state, probability) (w1, 1/2) (w2, 1/2)

Payoff to (A1, A2) (VH, VL) - c(X)*(1,1) (VL, VH) - c(X)*(1,1)

Clearly, (A1, A2) use the valuation method X only when they receive the
subsidy large enough to cover their loss in unfavorable states, VL - c(X) > 0.
Since both need to receive c(X), the total subsidy is 2c(X).

Next, consider that only A1 receives the subsidy. If the state is w1, A2 loses
and A1 wins. Then, A2 still rejects the project. If the state is w2, both
accept because A2 wins and A1 receives the subsidy. Interestingly, A2 accept
the method X because A2 has the option to reject the project if w1 occurs and
enjoys the value if w2 occurs. The expected value of A2 is VH/2, which is
greater than (VH + VL)/2 at coarse valuation. Thus, the subsidy to A1 enhances
the payoff of A2.

On the other hand, although A1 receives the subsidy, A1 is vulnerable to the

risk of losing opportunity at w1. Thus, the subsidy c(X) = VL is not enough.
The subsidy should compensate the possibility of losing opportunity at w1.
Thus, the following equation should hold for A1 to accept X.

6
P(w1) 0 + P(w2) (VL - c(X)) > P(w1) VH + P(w2) VL.

Left-hand side is A1's payoff of using X. Right-hand side is A1's payoff using
coarse method. To solve the equation, the subsidy should be at least as great
as P(w1)VH/P(w2).

-c(X) > P(w1)VH/P(w2).

In our example, since P(w1) = P(w2) = 1/2, the subsidy should be greater than
VH. Since VH > -VL, A1 should receive more subsidy if A1 receives alone than if
(A1, A2) receives together. In particular, if VH > -2VL, it is more expensive
to subsidize one than two actors! The better the value in higher state
becomes, the more expensive it becomes to subsidize only one. More generally,
if P(w1) VH/P(w2) > -2VL, it is more costly to subsidize only A1 than (A1, A2).

What would happen if the subsidy to A1 were smaller than VH? Then, A2 rejects
the project in the beginning because A2 knows that it will never receive VH.
Under W2, A1 always rejects the project, so the expected payoff to A2 is just
negative, VL/2. Expecting such response, A1 would reject the subsidy and
valuation if it has the option to do so. In other words, if we offer A1 with
the information about state, A1 avoid the offer.

Let us summarize implications before moving to next step. First, if it is not

possible to distinguish {w1, w2}, (A1, A2) invest in VI ((VH + VL)/2 >
0). Second, suppose that the finer method X to distinguish {w1, w2} is
available. If (A1, A2) choose not to invest in X, they will invest in VI. If
(A1, A2) choose to invest in X, investment occurs only when (A1, A2) receives
the subsidy to remove their negative payoff, VL > c(X). Third, from the second,
(A1, A2) will invest in VI, but not in X, if X is costly. Thus, (A1, A2) will
choose to stay ignorant even if X is free. (A1, A2) deliberately conduct
inferior valuation for VI. Fourth, the subsidy to an actor becomes higher if
the actor is the only one to receive the subsidy. Moreover, it can be more
expensive to subsidize only one than two if the subsidy-receiving actor’s
payoff in favorable state is big enough. Fifth, if we subsidize only one
actor, the receiving actor, as well as the other actor, may reject the
subsidy.

Value of truth/ cost of curiosity

I have showed that the effort to know the truth is neither always desirable
nor wanted. The curiosity toward truth may expropriate the valuable
investment opportunities. Nevertheless, as human being, (A1, A2) are endowed
with curiosity. Thus, they may want to know the truth even if it hurts them.
In experimental setting, it is possible to measure what is the critical value
of c(X) at which people choose to know or not to know the truth. In addition,
it is possible to investigate what other factors determine the critical value
of c(X).

Similarly, in principal-agent setting, (A1, A2) may want to invest in the

valuation method in order to avoid responsibilities or blames if they are
decision makers. In this case, the monitoring to know truth becomes the
consequence of moral hazard.

Construction and valuation

7
Valuation also affects how to construct the architecture of projects. To
illustrate, I extend the previous case slightly. Consider two project P1 and
P2. P1 produces (VH, VL) for (A1, A2) at w1, but produces zero at w2. P2 produces
zero for (A1, A2) at w1, but (VL, VH) at w2. Since VH + VL > 0, both projects
are valuable in aggregate.

(state, probability) (w1, 1/2) (w2, 1/2)

Payoff of P1 to (A1, A2) (VH, VL) (0, 0)
Payoff of P2 to (A1, A2) (0, 0) (VL, VH)

Pooling (P1, P2) into P1+2 converts this case into the previous case (P1+2 = VI).
Pooling can occur due to acquisitions, bundling, securitization, etc. The
pooling changes the behaviors of (A1, A2) much. Before pooling, regardless of
the knowledge about {w1, w2}, both projects are rejected. (A1, A2) rejects (P2,
P1) respectively. On the other hand, if we bundle P1 and P2 into P1+2, (A1, A2)
accept P1+2 on the condition that they are ignorant of {w1, w2}.

Let us consider a related case. Consider two project P1 and P2. P1 produces (VH,
0) for (A1, A2) at w1, but produces (VL, 0) at w2. P2 produces (0, VL) for (A1,
A2) at w1, but (0, VH) at w2. Since VH + VL > 0, both projects are valuable in
aggregate. Both (A1, A2) have the incentive to conduct (P1, P2) respectively.

(state, probability) (w1, 1/2) (w2, 1/2)

Payoff of P1 to (A1, A2) (VH, 0) (VL, 0)
Payoff of P2 to (A1, A2) (0, VL) (0, VH)

If we introduce valuation method X, interesting issue arises. If VL/2 - c(X) >

0, A1 has incentive to conduct valuation on P1. Same argument holds to A2 about
P2. Thus, the subsidy should be greater than -VL/2. However, neither A1 nor A2
has incentive to conduct valuation for the bundled projects P1+2, unless VL -
c(X) > 0 to both of them. The required subsidy is -VL. Since VL is negative,
we have -VL/2 < -VL. Therefore, the bundling reduces the incentive to conduct
valuation by (A1, A2).

To reverse the argument, if c(X) is less than VL/2, it does not matter whether
to bundle or not. If c(X) is between (-VL/2, VL), the projects should be
unbundled. If c(X) is greater than VL, it does not matter again whether to
bundle or not. Thus, the type of valuation method affects the bundling, or,
in more general term, the construction of business architecture.

To summarize implication; first, the bundling can change investment decision.

Second, the bundling can change valuation incentive. Third, the type of
valuation methods changes the bundling decision. All changes occur without
modifying payoff structure at all.

Relatedness and valuation

Relatedness is an important concept in organization researches. We can define

the relatedness in terms of information or product. I will simplify the
concept as the correlation in performances. Correlation can occur in many
situations. For instance, positive correlation can denote related
diversification and synergy. This section shows how the extent of relatedness
influences the choice of valuation methods.

I extend the previous model to incorporate the correlation structure. Two

agents (A1, A2) exist. Four states specify random variable {wLL, wLH, wHL, wHH}.

8
In state wAB, (A1, A2) receives (VA, VB). The probability measures of {wLL, wLH,
wHL, wHH} are {(1+ρ)/4, (1-ρ)/4, (1-ρ)/4, (1+ρ)/4} respectively. ρ exists
between [-1, 1] and denotes correlation.

wH• wHL (1-ρ)/4: (VH, VL) wHH (1+ρ)/4: (VH, VH)

wL• wLL (1+ρ)/4: (VL, VL) wLH (1-ρ)/4: (VL, VH)
Payoff to (A1, A2) w•L w•H

Suppose the knowledge of (A1, A2) is {∅, {wLL, wLH, wHL, wHH}}, same to the case
of coarse valuation method. In this situation, nothing changes from the
perspectives of (A1, A2) individually. They receive VH or VL with 1/2
probability each. In sum, with coarse method, the correlation ρ does not
matter.

However, the correlation allows us to introduce diverse valuation methods. We

can characterize each valuation method with the manner it partitions the
states. Let us analyze all possible methods. For simplicity, valuation cost
is set to zero.

((wLL, wHH)(wHL, wLH)): This valuation method is irrelevant because it does not
change anything. Conditional on each partition, (A1, A2) will earn VH or VL
with the probability 1/2 each.

((wLL)(wHH)(wHL, wLH)): This method is valuable. If the method identifies the

state as {wLL}, both will reject the project. In other state, both accept the
project. The value of the method is -(1+ρ)VL/4 to both (A1, A2) because (A1,
A2) can increase their payoff by that amount with the method. The larger the
correlation, the more valuable is the method.

((wLL)(wHH, wHL, wLH)): This method is valuable because both agree rejecting
the project at {wLL}. In comparison, the expected payoff at {wHH, wHL, wLH} is
positive leading (A1, A2) to accept the project at it. Expected payoff at {wHL,
wLH} is positive by assumption. Adding {wHH} makes the expected value larger.
The value of method is same to the above because partitioning out {wHH} does
not change the behaviors.

((wLL)(wHH)(wHL)(wLH)}: This method is most elaborate because it informs exactly

in which state (A1, A2) are. In all states except wHH, either A1 or A2 reject
the project. The value of the method is -(2VL+(1-ρ)VH)/4 to both. Thus, this
method is useful only when (1-ρ)VH < -2VL. The larger the correlation, the
more likely is the method adopted. In addition, the larger the correlation,
the more valuable is the method. Intuitively, under high correlation,
{wLL}{wHH} occur frequently, which are not controversial states.

((wHL, wHH)(wLL, wLH)): This method is quite interesting. It means the method
teaches A1 the state, but does not teach A2. Under naïve reasoning, we may
think A2 accepts the project, but A1 decides conditional on the state. However,
the type of the valuation method is the common knowledge to (A1, A2). A2 can
observe the behavior of A1 and may infer the state. In this case, the order of
decision-making between (A1, A2) matters. If they move simultaneously or if A2
moves first, the naïve reasoning is correct. If A1 moves first, we need to
find perfect Bayesian equilibrium (PBE). To make argument simple, let us
assume simultaneous move. Ex-ante, the value of the method is -((1- ρ)VH+(1+
ρ)VL)/4, possibly negative to A2. It is increasing function of ρ because of wHH.
To A1, the value is always positive, -VL/2, because it can remove all

9
unfavorable state. In addition, the valuation is never used if the expected
value at {wHL, wHH} to A2 is negative. At {wHL, wHH}, if the expected value is
negative to A2, it will reject the project. At {wLL, wLH}, A1 rejects the
project always. Then, the project is not accepted at any state. By backward
induction, the valuation method is not used in that case.

((wLH, wHH)(wHL, wLL)): Same argument to above.

((wHH)(wLL, wHL, wLH)): At wHH, both accept the project. At the other state, the
decision depends on the expected value. If the expected payoff at {wLL, wHL,
wLH} is positive, the valuation is irrelevant because (A1, A2)’s decision does
not depend on the valuation result. If the expected payoff at it is negative,
both will reject the project at it. Then, this case produces exactly same
result to the case of {wLL}{wHH}{wHL}{wLH}. It does not make any difference to
fully partition the state in {wLL, wHL, wLH}. In other words, it does not matter
to be measurable with respect to {wLL, wHL, wLH} or not.

((wHL)(wLL, wHH, wLH)): We assume simultaneous move for simplicity. If the

expected value of A1 at {wLL, wHH, wLH} is negative, the project is never
implemented. It is because A2 rejects the project at wHL always, and A1 rejects
the project at the other states. If the expected value of A1 at {wLL, wHH, wLH}
is positive, A1 accepts. At wHL, since A2 rejects the project, A2 can insure
itself from loss, but A1 loses value. Thus, the value of the method to (A1, A2)
is ((ρ-1)VH, (ρ-1)VL)/4. The higher the correlation, the more valuable to A1,
but less valuable to A2 is the valuation method. It is intuitive result: wHL is
less likely to occur when ρ is high.

((wLH)(wLL, wHH, wHL)): Inverse to the previous case.

Method Value Δy /Δρ

{wLL, wHH, wHL, wLH} Benchmark 0
{wLL, wHH}{wHL, wLH} 0 0
{wLL}{wHH}{wHL, wLH} Both: -(1+ρ)VL/4 +
{wLL}{wHH, wHL, wLH} Both: -(1+ρ)VL/4 +
{wLL}{wHH}{wHL}{wLH} Both: -(2VL+(1-ρ)VH)/4 +
{wHL, wHH}{wLL, wLH} Rejected or (-VL/2, -((1- ρ)VH+(1+ ρ)VL)/4) NA/(0,+)
{wLH, wHH}{wHL, wLL} Rejected or (-((1- ρ)VH+(1+ ρ)VL)/4, -VL/2) NA/(+,0)
{wHH}{wLL, wHL, wLH} 0 or both: -(2VL+(1-ρ)VH)/4 0/+
{wHL}{wLL, wHH, wLH} Rejected or ((ρ-1)VH, (ρ-1)VL)/4 NA/(+,-)
{wLH}{wLL, wHH, wHL} Rejected or ((ρ-1)VL, (ρ-1)VH)/4 NA/(-,+)

Let us inverse the problem. If (A1, A2) can choose the method, which method
they will choose? And, how does the relatedness affect the value of having
the option to choose method? Broadly speaking, they do not want to know
controversial states such as {wHL, wLH}, but attempt to know uncontroversial
states such as {wLL, wHH}. They are more eager to know the state as the
relatedness increases.

First, (A1, A2) always want to know {wLL} because they have the option to
reject the project in the state. The knowledge of the state increases the
payoff by -(1+ρ)VL/4 to both. It is intuitive because {wLL} occurs with (1+ρ)/4
and with loss VL.

10
Second, the players weakly prefer knowing {wHH}. Knowing {wHH} creates positive
payoff if its complementary information {wLL, wHL, wLH} is relevant. The
complementary information is relevant only if its payoff is negative because
the players can avoid the project in the state. If its payoff is positive,
the valuation is irrelevant because the knowledge does not affect behaviors.

Third, the players do not want to know {wHL} or {wLH} if the knowledge
generates the controversy in its complementary state too. In {wLH}, one likes,
but the other dislikes the project. If the complementary state of {wLH} also
makes the players disagree, the project is rejected. Since the project is
valuable unconditionally, the rejection hurts both players. Same argument
holds for {wHL}.

Fourth, knowing only {wLL}{wHH} is valuable. The payoff to (A1, A2) increases if
the relatedness is high and if valuation separates only {wLL} or {wHH}.
Intuitively, if {wLL} and {wHH} do not occur often, the ability to knowing the
states is not so valuable ex-ante. {wLL}{wHH} do not occur often if relatedness
is low.

This result implies the possibility that under related diversification, firms
have more incentive to conduct sophisticated valuation. In addition, since
this option to conduct valuation becomes less valuation with un-relatedness,
this model suggests a source of diversification discount.

To summarize intuitions, first, valuation methods can be very controversial

to the players. One player may want to implement a method, but the other may
reject it. Second, more information is not always good. The players can be
indifferent between fine and coarse methods. Often, they even prefer coarse
methods. Third, the relatedness between players influences the choice of
methods. Depending on the correlation, some methods are more preferable than
others are. Fourth, players tend to choose the method that reveals less
controversial state only. Players tend to dislike knowing controversial state.
Thus, related diversification enhances the incentive to conduct elaborated
valuation than unrelated diversification does.

We can generalize the intuition into other distribution. First, the valuation
method becomes useful to players if it partition out the sample space that
produces negative payoff to all. Second, as far as there is a partition in
which the expected payoff to all players is positive, the project is not
rejected until the partition occurs through the valuation. Third, the higher
is the relatedness, the more valuable becomes the ability to partition to
uncontroversial sample space.

Game and valuation

Let us consider the two players are in game situation, in which valuation is
strategic behaviors of the players. For simplicity, let us assume
simultaneous game. Same to the first case, if w1 occurs with probability 1/2,
the payoff to (A1, A2) is (VH, VL). If w2 occurs with probability 1/2, the
payoff to (A1, A2) is (VL, VH). If valuation is costless, Nash equilibrium is
(1) both players conduct valuation, and (2) project is never implemented. If
A2 does not conduct valuation, A1 can increase its payoff by rejecting the
project under w2 and by accepting it under w1. If A2 conducts valuation, A1
never receives VH because A2 rejects the project under w1. While A2 does not
reject under w2, A1 receives negative payoff VL in this state. Thus, the
dominant strategy of A1 is to conduct valuation. Same logic applies to A2.

11
Since valuation is dominant strategies, the project is not implemented. This
game is Prisoner’s Dillema.

To generalize the problem, let us denote the types of (A1, A2) as (θ1, θ2) such
that the payoffs of the players change by θi as they conduct valuation. Thus,
if both conduct valuation, the payoff becomes (θ1, θ2) while the project is
rejected. Similarly, if A1 conducts valuation, but A2 does not, their expected
payoffs become (VH/2 + θ1, VL/2).

(A1, A2) Valuation No valuation

Valuation (θ1, θ2) (VH/2 + θ1, VL/2)
No valuation (VL/2, VH/2 + θ2) ((VH + VL)/2, (VH + VL)/2)

It does not change the situation much even if we introduce the types. If θi >
VL/2, Ai always conducts valuation. Otherwise, Ai rejects the valuation. The
players conduct the strategy whether they know the other’s type or not. This
analysis shows that the players need to cooperate in order to hide the truth
and to implement profitable projects. Such cooperation can occur in many
situations such as iterated games and cooperative games.

Organizational capital budgeting model (OCBM)

This section describes Organizational Capital Budgeting Model (OCBM)
extending value space idea. I hypothesize that a firm performs following
OCBM-optimization when the set of valuation methods (S) and raw investment
opportunities (K) are given. OCBM is defined in association.

Definition: OCBM-optimization means,

• Maximize: G(V, B, C, U, X), in which B ≡ A - E
• With respect to s Є S and k Є K

Definition: OCBM is the set of s that solves OCBM-optimization given k.

OCBM(k) ≡ arg.max{s} OCBM-optimization for a given k and s Є S.

Notations: {V, B, A, E, C, U} are the functions of s and k. They are the

perceived value (V), net accuracy (B), accuracy (A), effort (E), controversy
(C) and uncertainty (U). Net accuracy is gross accuracy minus effort to
obtain the gross accuracy. X denotes situation factors and may include S and
K. Let us call {C, U} as social factors. I assume well-defined second order
conditions. C and U capture internal and external influences in a broader
view. V = V(s, k) is the derived value of a project k using a valuation
method s. For instance, V can increase with mean, but decrease with its
variance if s is mean-variance optimization. If s is NPV method, V increases
with expected cash flow, but decreases with the beta of the cash flow.
Controversy has been important concepts in Hobbes (1881), Cyert and March
(1963), Bower (1970), Rawls (1971), Nelson and Winter (1982), Weick (2001),
etc. Knightian uncertainty has been in Knight (1921), Keynes (1921), Simon
(1945, etc.)

Once we presume the choice set as S x K (value space) instead of just K, it

is clear that simply maximizing perceived firm value (V) is unrealistic.
Otherwise, firm would choose the most optimistic valuation method for a given
k. This situation may hold only in particular contexts such as exaggeration,
strategic disclosure or moral hazard. Instead, the OCBM-optimization
hypothesizes that organizations conduct capital budgeting in order to

12
accomplish goals (G). Cyert and March (1963) argues that firms pursue goals
or aspiration levels. Four of the most important goals are (a) choosing high
value project, (b) increasing the accuracy of the investment decision, while
reducing the organizational resources required for the decision, (c) reducing
the controversy in investment decision, and (d) decreasing uncertainty in
investment. Next assumption expresses them.

Assumption: GV ≥ 0, GB ≥ 0, GC ≤ 0 and GU ≤ 0.

Organizations make decision on the value space S x K which is the Cartesian

product of valuation strategies and investment opportunities. A raw
investment opportunity becomes a perceived opportunity after it meets a
valuation strategy. In other words, a valuation strategy (s Є S) determines
the perception of an investment opportunity (k Є K) in an organization. S
includes intuition, heuristics exploration, analysis and communication.
Exploration includes fieldwork and the aggregation of information from both
in and outside of an organization. Analysis includes qualitative studies as
well as quantitative tools of SCBM. Communication matters because it affects
the perception of subgroups in an organization.

SCBM is a special case of OCBM

Let us start from the formal definition of standard capital budgeting model
(SCBM).

Definition: SCBM is the solution of s to maximize B given k.

SCBM(k) = arg.max {s} B for a given k and s Є S.

Under well defined first- and second-order conditions, SCBM(k) solves Bs(s, k)
= 0. I define SCBM as the most accurate method of valuating a project in
consideration of effort. Numerous textbooks and articles discuss NPV and its
variations ending up rationalizing them as the accurate tool for investment
decision. In addition, NPV is very flexible to account for the cognitive or
organizational costs. An analyst can change the assumption or a specification
of an NPV model, so that she can make it simple or sophisticated depending on
the preference and the opportunity cost of doing the analysis. For instance,
it is possible that she performs valuation only up to satisficing level to
reduce the cognitive resource. Thus, SCBM covers the case of bounded
rationality and heuristics.

It is clear that SCBM is not always the best valuation method to maximize the
organizational goal. To the contrary, SCBM is optimal only in special
situation.

Lemma: SCBM (s*) is the optimal valuation strategy if and only if

G (V(s*, k), B(s*, k), C(s*, k), U(s*, k), X)
≥ G (V(s, k), B(s, k), C(s, k), U(s, k), X) for all s Є S and k Є K.

Notice 'if and only if' in the proposition. SCBM is in general suboptimal
valuation strategy (even without social factors). Let me illustrate why we
need 'if and only if'.

Suppose that a firm has two projects and two valuation methods: {alliance,
integration} x {NPV, interview}. NPV is a representative SCBM. Expert
interview is a popular qualitative method that firms use before launching new
projects. The sets of projects and valuation methods generate the following

13
hypothetical table as a value space. The first and second numbers in
parenthesis are value and net accuracy respectively.

(V, B) Alliance Integration

NPV (-1, 1) (1, 1.2)
interview (10, 1/2) (-10, 1.2/2)

Clearly, the firm is in dilemma. NPV is always more accurate and supports
integration. On the other hand, while interview is less accurate, it is
overwhelmingly for alliance and against integration. Thus, even if the firm
perceives NPV as the most accurate valuation strategy, the firm may not
always choose (integration, NPV) pair instead of (alliance, interview). One
solution is Bayesian encompass: taking weighted average of the qualitative
and quantitative information in the example. However, it is a semantic issue
whether we can call it as SCBM. We may not want to call the linear
combination of NPV and a qualitative index as SCBM. Rather than arguing
everything is the variation of NPV, I regard SCBM specific in order to find
richer implications. The problem in the example is that the ordering of
perceived value (V) changes significantly as the valuation method (s) changes.
If all s Є S generate similar ordering, such problem would not exist. In this
case, however, there is no reason to stick to NPV over other methods because
all methods produce similar results anyway.

Lemma: Suppose social factors do not exist. If Vs = 0 for all s and k, SCBM is
always the optimal valuation strategy.

I allow more general class SCBM than commonly presumed ones. For instance, I
allow the possibility that a firm may not always choose the highest NPV
project. Suppose NPV is the only valuation method a firm has. (V, B) pair for
project can be,

(V, B) Alliance Integration

NPV (1, 2) (2, 1)

In this situation, the firm values integration the twice of alliance, but
only with the half of confidence/ accuracy. It might be hard to model the
post-integration situation or to consider all complexities involved in
integration contracts. It is uncertain which strategy the firm selects. Then,
when does a firm select the highest NPV project? Next lemma answers the
question.

Lemma: Suppose social factors do not exist. If Vs = 0 and ∆Vk/∆Bk ≥ 0 for all s
and k, a firm selects the project with the highest value measured with SCBM.

If the perceived value and accuracy tend to move with the types of projects,
we have the well-known situation of choosing highest NPV projects.

In sum, SCBM is in general suboptimal method of valuation. In addition, even

if SCBM is the only valuation tool a firm has, it is generally suboptimal to
select a project with highest NPV.

Social factors and the deviation from SCBM

Next, let us discuss the importance of organizational/ social factors.

Without loss of generality, let us assume Vs = 0 and introduce only the
controversy with a weighting parameter δ. Then the goal becomes G(V, B, δC).

14
The formula reverts to the previous with δ = 0. I will show that δ increases
the deviation from SCBM. The first order condition with respect to s and k
becomes:

GBBs + δGCCs = 0.
GVVk + GBBk + δGCCk = 0.

The first equation in the first-order conditions is relevant because it

expresses s as the implicit function k in consideration of B and C. Let us
rewrite it as: Bs = -δGCCs/GB. In contrast to SCBM, we do not have Bs = 0
unless δ, Gc or Cs is zero. δ and Gc have the same intuition to determine the
relative importance of controversy. Cs specifies how sensitively subgroups
respond to the change of valuation method. Same intuition holds for
uncertainty (U). Thus, we have following proposition.

Proposition: Suppose Vs = 0 for all s and k. SCBM is always optimal valuation

strategy if
1. The relative importance of social factors, controversy and uncertainty,
are zero. Or,
2. The sensitivities of social factors with respect to valuation methods
are zero. Or,
3. The social factors cancel their effects each other.

As the social factors become important, an organization needs to balance the

impact of valuation strategies on accuracy and on social factors. GBBs term
measures how the change of valuation strategy affects the level of accuracy
and consequently the organizational goal. In case accuracy is the only
concern of valuation, an organization would change valuation method until the
marginal value of further change becomes zero. By definition, such valuation
method is SCBM. Notice GB is always positive.

On the other hand, -δGCCs is the cost of changing valuation strategy. The
change in valuation method generates social impact by δCs, which in turn
affects the organizational goal by GC. Given this cost, an organization cannot
mindlessly pursue accuracy in valuation, but allows valuation method to
deviate from SCBM in order to balance the social factor and accuracy as: GBBs
= -δGCCs. Next empirical predictions summarize such intuition.

Predictions: To denote NSCBM ≡ OCBM\SCBM,

1. The more important the social factors become; the more NSCBM is used.
2. The more sensitive the social factors become with respect to valuation
methods; the more NSCBM is used.
3. The importance of controversy and uncertainty increases the use of
NSCBM.

Illustration about social factors

Let us illustrate the role of social factors with simple examples. Next table
illustrates two distinct valuation strategies in which the perception on the
risk differs. For instance, both methods can be NPV, but with different
intuitions. Or, they can be quantitative and qualitative methods respectively.

15
 Perceived State 1 State 2
(team1, team2)
accuracy (Prob. = 50%) (Prob. = 50%)
Method 1: P1:(100, -60) P1:(-60, 100)
50%
High diffusion P2:(15, 15) P2:(15, 15)
Method 2: P1:(50, -30) P1:(-30, 50)
50%
Low diffusion P2:(15, 15) P2:(15, 15)

Shaded area denotes the selected combination of valuation strategy and

investment opportunity, (Method2, P2). To explain the situation,

• The realization of either state 1 or 2 is independent of systematic

factors, i.e. idiosyncratic. This simplifying assumption prevents cash-
flow fluctuations from influencing the discount rate of the project.
• The first column denotes two valuation strategies. Method 1 argues the
diffuser distribution for P1 than Method 2 does.
• The second column denotes how much confidence the firm has about the
accuracies of each valuation strategy. I assume they are same for
simplicity.
• The third column denotes what will be the returns to (team1, team2)
from the investments P1 and P2 respectively if state 1 occurs. The
fourth column shows the case of state 2.
• The first row shows that state 1 and state 2 are equally likely (50%
each).

P1 creates controversies and two undesirable consequences. Most of

organization can choose P2 over P1. First, the decision-making would be
costly in terms of time and resources. Some subgroups in a firm may reject
the project. The firm may have to design additional contracts to redistribute
payoff after the project is over. Second, such controversies will make
coordination and communication difficult in implementation stages. If
implementation requires the effort of both subgroups, the losing group may
sabotage the project. I exclude the impact of controversy from the payoff
table in order to demonstrate the presence of controversy clearly.

Managers can communicate the choice in two ways. First, they assert that they
used 'Method 1' and made suboptimal choice. Second, they declare that they
used 'Method 2' and made optimal choice.

I argue firms will choose the second way of communication. The second way is
less likely to create cognitive dissonance, easier to communicate and helps
building consensus. For instance, firms can argue P1 does not fit to their
culture, goal or vision, which method 2 may take into account. Thus, a firm
considers social factors and the easiness of communication choosing method 2.

Corporate Social Responsibility (CSR) may imply similar situations. Let us

consider (CSR-Low, CSR-High) projects, and (internal, external) stakeholders.
Internal stakeholder is simply a firm.

(Internal, Perceived State 1 State 2

External) accuracy (Prob. = 50%) Prob. = 50%)
Method 1: CSR-Low :(100, -60) CSR-Low :(50, -30)
60%
High-diffusion CSR-High:(24, -2) CSR-High:(12, -1)
Method 2: Low- CSR-Low :(50, -30) CSR-Low :(25, -15)
50%
diffusion CSR-High:(24, -2) CSR-High:(12, -1)

16
Similar to before, Method 2 evaluates CSR-Low investment at half of benefit/
costs in each state compared with Method 1. Method 1 proposes that CSR-Low is
better than CSR-High from the perspectives of both the firm and social
planner. If the firm uses Method 1 to evaluate CSR-Low, the investment at
state 1 will benefit the firm by 100, but hurts environment by -60. At state
2, the payoff to the firm and environment is (50, -30). Thus, the expected
payoff becomes (75, -45). Similarly, the pair (Method 1, CSR-High) produces
expected payoff (18, -1.5). Thus, CSR-Low is better for the firm (75 > 18)
and for the society (75-45=30 > 18-1.5=16.5). In comparison, when the firm
uses Method 2, CSR-Low is better for the firm (37.5 > 18), but CSR-High is
better for society (37.5-22.5=15 < 18-1.5=16.5). Consequently, without moral
hazard, the firm should choose either (Method 1, CSR-Low) or (Method 2, CSR-
High).

Given the highly negative consequence in state 1, CSR-Low project may

generate controversies. Thus, the firm prefers CSR-High projects. CSR-Low
project may be even infeasible due to the relationship with external
stakeholders. Given the similar accuracy of Method 1 and Method 2, I expect
the firm chooses (Method 2, CSR-High) pair over (Method 1, CSR-Low) or
(Method 1, CSR-High) in consideration of communication.

Next is another example how communication consideration can change valuation

method.

(team1, Perceived State 1 (Prob. = State 2 (Prob. =

team2) accuracy 50%) 50%)
P1:(100, -60) P1:(-60, 100)
FINE 60%
P2:(18, 12) P2:(12, 18)
P1:(20, 20)
COARSE 40%
P2:(15, 15)

The valuation strategy FINE can distinguish state 1 and state 2, but COARSE
cannot. However, P1 always generates higher expected return regardless of the
valuation strategies. I propose that firms communicate the pair (COARSE, P1)
in case (1) managers stop the valuation at COARSE, not proceeding to FINE, or
(2) managers proceed to FINE, but only communicate COARSE as official. Case
(1) arises because COARSE is good enough (satisficing) to make decision and
communicate. Case (2) occurs because managers intend to make P1 legitimate
without generating conflicts between team 1 and team 2, while maximizing firm
value. Simply speaking, it is easier to justify (COARSE, P1) than (FINE,
P1).

Procedural justice generates more parsimonious example as follows.

Perceived
(team1, team2) P1 P2
accuracy
Domination 60% (100, -60) (15, 15)
Participation 50% (50, -30) (15, 15)

In the example, the pair (Participation, P2) is selected over (Domination,

P1). The participation of team 2 in decision-making process may decrease the
accuracy of valuation if the information of team 2 overlaps that of team 1
except pure noise. It is also possible that the participation involves very
different valuation methods from the domination. For instance, the definition
of costs can be politically determined in transfer-pricing conflicts
(Granovetter, 1985). Nevertheless, participation in decision-making raises

17
perceived procedural fairness. Procedural justice increases the satisfaction
of teams and performance as a result (Lind & Tyler, 1988). Communication and
participation also increase group longevity (Katz, 1982). Such social
consideration leads a firm to choose (Participation, P2).

I have skipped the toy example for uncertainty case and focus on controversy.
Instead, I entirely devote one case study to discuss uncertainty in later
section.

Cases
I present two qualitative studies to illustrate how OCBM can explain capital-
budgeting practices. The first case investigates the practice when Knightian
uncertainty is high. The second case analyzes how a firm conduct capital
budgeting when both uncertainty and controversy are high.

Case #1: Knightian uncertainty and capital allocation

This case shows that the conventional view, presuming the independence of
valuation from organizational context, is neither normatively correct nor
close to reality. Under high Knightian uncertainty, simple application of NPV
for valuing a project can distort organization structure, coordination,
culture, incentives of employees and cognitive variables in organizations and
can prevent firms from identifying valuable investment opportunities. This
case study also shows that the view of the subject firm on valuation process
is different from that of textbook.

To summarize the intuitions; first, valuation is a process to articulate

strategies under uncertainty. Valuation is not just to evaluate the value of
a project that already has concrete strategies, action plans, revenue, costs
and discount rates. Instead, during valuation process, necessary data are
collected, information is exchanged, and business plans are formulated. Such
articulation of strategies is sensitive to what valuation process the firm
uses. In addition, business plan include many qualitative indicators about
strategies, organization and implementation. In this sense, I find that the
firm identifies and constructs value during valuation with valuation methods
and processes. This finding corresponds to a proposition of OCBM: The more
sensitive is uncertainty to valuation strategies, the more non-standard
capital budgeting models are used.

Second, the concept of risk that financial theories stress for valuation does
not capture important practices in reality. The binary distinction between
systematic and idiosyncratic risks is too restrictive. Instead, firms employ
more convoluted and strategic concept of risk tailored to their particular
situations. Such practice is in line with the findings that valuation is a
process to find values rather than to calculate values. The usual valuation
techniques based on the concept of systematic risk are therefore neither
right nor realistic. This finding also confirms the proposition of OCBM: The
more important Knightian uncertainty becomes, the more non-standard capital
budgeting models are used.

In conclusion, this case proposes that we should understand and design

valuation in a broader context in order to incorporate uncertainty-resolution
and strategy-formulation practices. Implementation environment,
organizational situation and value creation also matter. This case study also
describes what methods the firm considers as the alternative to the standard

18
capital budgeting models. In addition, since value is a decision-making
process and means to construct values, sticking to single rule such as NPV
and to the narrow concept of risks such as systematic risk is neither
desirable nor realistic.

Case introduction and key issues of ICM

The firm is one of the major mobile operators with 20 million subscribers and
over 50% market share. In addition, the Firm is one of the leaders in global
wireless industry in terms of users, market capitalization, technology and
overseas operations. For example, it is the first operator in the world to
commercially launch CDMA IS95 1x services.

The firm belongs to a very successful Asian conglomerate. The conglomerate

starts as a tiny textile manufacturer more than 50 years ago. Via vertical
and horizontal diversification, it now own more than 40 affiliates covering
energy, chemicals, telecommunications and services in many countries. In
particular, the energy and the telecommunications subsidiaries are large and
leading their respective markets.

Importance of strategic/ qualitative consideration: ICM and new businesses

The focus of the internal capital allocation of the firm is to develop new
businesses. The firm’s interest in new businesses originates from (1) the
matured wireless market, (2) the technological trend of digital convergence,
(3) large cash from the profitable wireless business, and (4) sizable
customer intelligence about wireless service and energy.

First, the wireless business is now mature. The domestic market cannot grow
much because already 80% of population uses mobile phone as of 2007. The firm
cannot take more market share either because doing so would trigger
government intervention given its large market share. The only way to
increase revenue is to pull out more cash from existing customers and to
diversify revenue sources with global expansion and new business. Yet,
extracting more cash from existing customer using existing business is
difficult because of heavy regulation on prices, competition and the
decreasing trend of the average revenue per user (ARPU) from voices

On the other hand, the revenue from other applications such as data and
value-added services tend to increases, which drive total ARPU. In this
situation, the firm and its competitors have tried to develop broad portfolio
of services and to extract more from existing customers with new applications.
The firm wants to develop and manage large part of such new services and
contents themselves via its own units rather than outsourcing them. The firm
argues that

"We need to develop many contents and services ourselves. Killer

applications change the fate of hi-tech organizations. They become core
competence. Full control of those products will give us strategic
flexibility. I believe we can think the projects as vertical
integration. Vertical integration is efficient because of the close
interaction and the concern about knowledge leakage to competitors."

In sum, given the maturity of market, new businesses such as diversification

and global expansion seem only options to spur revenue growths. The firm’s
insight resembles transaction cost economics (Williamson 1975) and knowledge-
based view argument (Kogut & Zander 1992).

19
Second, new businesses become more significant with the trend of digital
convergence, rendering the wall between technologies or between products/
services less relevant. See the next quote.

"In the middle of convergence among contents, computer, software,

consumer-electronics and communication industry, how can you define
related and unrelated businesses? How can you limit yourself to
particular sector? We need to develop new businesses in order to follow
the trend of digital convergence. Firms in other sectors can become
formidable competitor suddenly. Diversification prepares us for
unexpected competitions."

Indeed, once we take into account digital convergence, it would be hard to

define related/ unrelated diversification/ core businesses. This can be both
concern and opportunity of the firm. The only relevant factor may be the
capability to conduct diversification and to cope with dynamic environments.
Such argument for diversification resembles the view of resource-based view
of the firm (Wernerfelt 1995 for the history of RBV).

Third, the large and stable cash flows from wireless businesses make it
tempting and possible to launch new businesses. Large cash holdings tend to
draw the pressure from investors to use cash in growth investments
opportunities; otherwise, the shareholders will compel the firm to return the
cash via dividends or buy-backs. Similar finding is Lamont (1997) which finds
the investment in non-oil subsidiaries increases as the oil subsidiaries have
more cashes. Lamont presents the case as the proof for ICM existence. To
compare, the interviewees seem to regard hoards of cash as the strategic
resource.

Fourth, the firm recognizes its wireless users as the most valuable resource
to start new business. Its wireless customers are more that 50% of domestic
population with purchasing power. More importantly, the firm has accumulated
the customer database with impressive quality. The firm offers several
membership options among which its users can choose. When a customer chooses
membership option, it reveals its type and consumption penchant. In addition,
the firm can obtain more refined consumption and life-style information as
the customers use the membership card in restaurants, theaters, shopping, etc.
Additively, one of its sister company owns the largest gas station network in
domestic market and has its own membership card system. The combined customer
intelligence will certainly become effective marketing and planning tool for
new businesses.

Fifth, the growth through its new business may hedge regulation risks. As a
market leader, the firm has faced asymmetric disadvantages from regulators.

Characteristics of Capital Budgeting and Issues

This section describes what financial techniques and processes the firm uses
to select projects to invest. It demonstrates how the firm’s goals,
diversification strategies and corporate governance conditions affect
valuation processes. While the firm argues that NPV is its main valuation
tool, I argue that what it uses is not NPV.

The firm regards valuation having multiple purposes. First, valuation should
calculate efficient asset allocation by minimizing risks and maximizing
profits. To define risks, the firm considers not only systematic but also

20
idiosyncratic risks and qualitative uncertainties. In this sense, even if the
firm’s valuation framework receives correct and exhaustive inputs, the result
can be different from what NPV recommends as a firm-value maximizing solution.

There can be several competing explanations. (1) The firm’s valuation

practice may suggest that the firm has different agenda from firm-value
maximization. (2) The valuation may serve other purposes in addition to
computation. (3) The valuation process may create extra values by finding
extra opportunities beyond the inputs it receives.

The key finding is that under Knightian uncertainty, valuation process

becomes more than applying NPV formula to compute value. It is an integrative
framework to generate insights about uncertain environments and to construct
strategies to deal with the uncertainties. I will explain firm’s risk
analysis procedures in later sections in detail.

Second, the firm regards valuation as a framework to make decisions more

efficiently. Valuation process requires gathering and assessing different
sources of information. It also helps breaking down various issues to
tractable problems starting from revenues and cost estimation. This process
enhances the capability of the firm to learn and to cope with uncertainties.

Third, the firm uses valuation techniques not only for making decisions, but
also for enhancing internal communication. Thus, managers can conduct better
joint problem solving and co-work by using valuation tools to share
information. In addition, managers admit that the usage of valuation grants
the perception of authority, rigorousness, justice and fairness. In all,
valuation is a management tool with which managers can govern organizations
better.

Fourth, the firm uses valuation for external communication. It gives the
sense of transparent and objective disclosure about new businesses and their
potential for growth. During the communication with external stakeholders,
valuation builds trusts with them such as shareholders/ board of directors,
creditors, suppliers, customers, regulators and external auditors.

Fifth, by imposing constraints and providing guidelines, the firm can prime
managers toward the corporate objectives and align the different incentives
of them. The review and discussion during valuation processes conduct such
roles. This cognitive value of valuation becomes salient during internal
interactions, which I will describe shortly.

In conclusion, valuation is more than a valuation tool. It is also a

communication, education and decision support tool. Through value, the firm
can construct values as well as calculate values under non-quantifiable risks.
In this sense, since valuation tool incorporate useful concepts to design
decision-making process and to create values, it may not maximize firm value
to apply NPV in parochial or textbook ways in order to assess businesses
opportunities. The same argument can become even more prominent in risk
analysis.

Next sub-section better displays valuation process in detail. I decompose

valuation process into two parts: pricing and risk analysis. I explain the
parts sequentially.

Pricing Process

21
Overall Evaluation Process

Previous sections say that the key issue of the firm in internal capital
allocation is the investment in new businesses. Next figure shows the overall
process of evaluating new business proposals.

Figure: Overall Evaluation Process. Source: Interview and reconstructed by

the author

The HQ of the conglomerate, which is officially dismantled only and replaced

with similar organization, designed the process. The evaluation process has
several important ingredients. Let us investigate them at the each step of
the process.

At the first step, the firm assesses whether a project corresponds to the
mid/ long term strategies. Short-term strategy and profit are less important
at this stage. The firm does not continue valuation unless a project passes
this step. It shows the importance of long-term strategies such as corporate
visions or objectives. Quantitative projections and short-term financial
performances are insignificant indices in this step. The firm can reject
positive NPV projects as a result.

The second step is economic analysis, the narrow and conventional definition
of valuation. Net Present Value (NPV) and Internal Rate of Return (IRR) are
important metrics. If a project proves profitable, it goes to the next step.
Else, synergy assessment follows. The firm does not discard a project right
away just because it is unprofitable in narrow sense. Instead, it places the
project in the bigger picture to evaluate whether the project creates
positive externalities to other new and existing businesses. Such synergy
analysis can justify many diversifications. In addition, since synergy
analysis follows only after economics analysis rejects a project, the firm
might pass to next steps a project with negative synergies, but positive cash
flow. The firm does not test synergies once a proposal passes economic
analysis.

"One possibility is that a project has large cash-flow potential and

large negative synergies, and passes this step. However, it is easier
to detect negative synergies because other divisions will express their

22
 concern during the valuation process. In addition, analysts are
experienced enough to consider this possibility even in the first
step..."

Once the firm concludes that a project is profitable or creates enough

synergies, the project goes to the third step for risk assessment. Standard
finance theories argue that only systematic risks matters in valuation. The
intuition of why companies do not have to account for idiosyncratic risks is
that investors can simply diversify away or hedge such risks. This case study
reveals that the risk analyses can be far more complicated than the standard
financial theories suggest. Given the importance, I will explain the risk-
assessment practice of the firm in separate section.

Step 4 & 5 are review processes. Related divisions review the new businesses,
and then strategy committee evaluates the projects from multiple perspectives.
The valuation models become communication tool. The bargaining, information
sharing, interaction and joint efforts to solve problems occur intensely in
this stage.

How valuation strategies are applied?

The firm applies financial analysis techniques through four steps: Business
Logic, Data Gathering, Modeling, and Sensitivity Analysis.

Figure: Valuation Technique Application Process. Source: Interview, archival

analysis and reconstructed by the author

In what follows, I explain the process in each step.

Business Logic: Managers develop logics about how sales are generated, how
costs are related with sales, and what are appropriate investment and
financing schedules. Corporate visions and long-term strategies are important
principles to assess such logics. Once the business logics are properly
defined, managers construct variables and quantify them to be incorporated in
valuation framework.

Data Gathering: After defining variables to the reasonable details, managers

gather data. They search information, process the collected data, and
determine the assumptions of the model when proper data are unavailable.

Modeling: Model parameters are estimated or conjectured. In tandem, managers

prepare supporting documents about the estimations. They also review data and
methodologies. Then, managers test the model and improve its performance.

Sensitivity Analysis: Input variables are changed within fair ranges in order
to identify the main movers of financial performance. Once the dominant
variables are identified, scenarios are developed with them. Optimistic,
neutral and pessimistic scenarios are constructed and analyzed.

23
Iteration: Sensitivity analysis often reveals important issues. Managers
investigate and gather data about them. Then, they iterate modeling the
situation and analyzing sensitivities while it is necessary.

Next figure and interviewee comments outline the valuation framework in

relation to the Business Logic Development phase. They show that many input
values in valuation model are targets instead of forecasts.

"When uncertainty is high, it is often infeasible to estimate market

size, costs and discount rates. Instead of trying to estimate the input
values, we regard them as target to achieve or as strategies. Relevant
teams should fill the blank of model… I mean how they will achieve the
targets and what strategies they suggest… NPV may not be an appropriate
framework to design targets…"

Figure: Elements of Business Logic. Source: Interview, archival analysis and

reconstructed by the author

Interviewees particularly emphasize market size forecast/ targeting.

"Market size forecast is most important. If forecast is not possible,

appropriate goals for market share and pricing matter. The market
forecasts set the target for revenues and expenses. The revenue and
expenses plans generate prospect income statement. Investment plan is
subject to market size forecast and expense plans. Investment plans and
prospect income statement determine funding strategies and cash-flow
statements, the basis of quantitative analysis using NPV, IRR and
Payback method."

Risk-Analysis Process under Knightian uncertainty

The third step of the overall evaluation process is risk analysis. The firm
uses risk analysis not only for valuation, but also for research, learning,
planning and strategy formulation. Such practices become salient when the
firm analyzes uncertainties and idiosyncratic risks, which are irrelevant in
standard capital budgeting models.

Valuation requires risk analysis. In detail, valuation means computing the

price of new projects, which is the weighted sum of future cash flows of
projects. The weights of the future cash flows are determined by the discount
rates, which is the function of the correlations (risks) between the future
cash flows and systematic factors. Thus, the risk analysis in valuation

24
process means identifying relevant systematic factors and computing their
correlation with cash flows. Irrelevant in valuation is idiosyncratic risk,
which determines the remaining fluctuations of the cash flows, but not
related with economy-wide risks. Also, qualitative uncertainty is not
explicitly considered in the standard model.

Our case reveals that the simple binary concepts of systematic and
idiosyncratic risks do not capture the reality. The reason is that the risk
analysis is not just the process for the pricing of projects. Instead, the
risk evaluation separately affects capital budgeting as much as NPV does,
which the overall evaluation process chart shows. It means that the financial
techniques are only a partial mediator to relate risk with investment. To
rephrase, the attractiveness of project is the weighted sum of risk analysis
and standard valuation, which includes the analysis of systematic risks.
Therefore, the more risk analysis becomes important, the more valuation
deviates from standard valuation. Risk analysis becomes important when
Knightian uncertainty and idiosyncratic risk becomes important. Next figure
summarizes the argument.

The risk analysis process starts from defining and assessing potential
businesses risks ex-ante. Next, tracking and developing contingency plans
follow. The firm categorizes risks depending on their sources such as macro,
industry, customer and internal sources.

Macro Risk: Macro risks cover macroeconomic risks, political and regulatory
risks, technological risks and socio-cultural risks. Macroeconomic risks
include GDP growth, inflation, interest, exchange, energy, materials, etc.
Political and regulatory risks are related with the change of political
system, regulations, privatization, antitrust law, intellectual property
rights, etc. Technological risks are from the change of IT trends and shocks
to telecom technologies and new web technologies. Socio-cultural risks
concern population growth, mobile subscribers, wireless internet penetration,
mobile commerce trends, contents trends, etc.

Industry Risk: Industry risks are closely related with Porter’s five forces
model (Porter 1985). Customers’ switching costs to substitute services are
particularly important. The shocks from entrants, suppliers, competitors,
demand, substitutes define industry risks. Entrants’ risks comprise the
degree of economy of scale, required investment, product differentiation,

25
cost advantage, competitors’ retaliatory entry, and regulatory barriers.
Suppliers risk is related with supply monopoly, input differentiation, input
substitutes, etc. Competitors pose the risk of consolidation, collusion,
price/ non-price competitions, switching costs, etc. Demand risks include
price sensitivity, income, switching costs, etc. Substitute-risks consist of
the sensitivity to substitutes, relative prices and switching costs.

Customer Risk: Both characteristics and needs of customers are the sources of
risks. Important customer characteristics are price sensitivity, budget,
quality of current handset, service awareness, mobile usage patterns,
internet commerce size, churning cost, etc. Customer needs influence the
demand and migration rate to new services, the reliance on internet
transaction (both wired and wireless), business-to-business/ business-to-
customer transaction patterns, etc.

Internal Risk: The firm’s specific situation can pose risks to projects.
Uncertain financing and investing schedule can increase opportunity cost
suddenly. Staffing and some HR issues may delay new businesses. Sourcing IT
engineers could be difficult. The organization may not be designed as planned.

The firm assesses the risk with the probability to occur, the damage given
the events and controllability.

Issues in Risk Analysis Process

When I asked what valuation methods the firm use, the interviewers answered
they use NPV. Broadly, the aforementioned process may be framed as NPV.
However, the NPV in use does not correspond to the standard concepts in
textbooks. The firm uses NPV in creative and abstract manner to the point
that the method is not NPV any more. This becomes particularly apparent in
defining risks. This experience reveals the problem of using survey to know
what methods firms actually use. It is necessary to investigate in which
context and for what purpose firms use various methods.

It is certain that the risk analysis process aims at not just valuating new
businesses, but also addressing broader issues such as developing better
strategies and organization. The risk analysis becomes broad and creative
especially when Knightian uncertainty is high. Such rationale of risk-
analysis process is in line with our findings that valuation is a process to
create values via better procedures and organizations. Financial theories
argue that new businesses should be evaluated upon the binary notion of risks.
This case argues such view is too narrow.

In relation to usual valuation theories, we can particularly notice several

findings. First, the firm clearly considers the idiosyncratic risks such as
internal risks, customer risks and some of industry risks. Second, many of
the uncertainties are qualitative. The macro risks are a good example.
Although we can regard them systematic, it would be very hard to reflect some
of macro risks in valuation models quantitatively. Third, the risk analysis
process purports both valuation and instruction for new businesses. One
challenge to such practice is that, while instructing employees may help
instilling corporate objectives to them, it may obstruct objective
assessments or the inclusion of diverse knowledge. Fourth, the concept of
risk is very extensive. Such broad notion of risk may promote the firm to try
unrelated diversification in order to reduce total risks. This may sacrifice
the interest of shareholders. On the other hand, the illustrated risk
management process may stress related businesses with the concept of internal

26
risk, industry risk and macro risk. Such trade-off originates from the dual
roles of risk analysis, valuation and strategy formulation. In sum, risk
analysis facilitates strategy formulation by providing research framework,
communication tool, organizational learning through routinization, and
organizational justice. Little researches have analyzed the trade-off
inherent in the dual role of risk analysis.

Many financial institutions have the risk analysis and management processes
similar to the firm’s. Regulators urge financial institutions to adopt
prudent risk management system and process (Basel II). It is important to
understand why regulators enforce such supervision in financial sector.
Financial institutions are special in that a shock to a bank can spread to
other financial institutes and threat the stability of entire financial
system, which in turn affects whole economy. In this sense, in financial
sector, many financial risks are systematic. Despite these important
externalities, because of moral hazard and the properties of financial
stabilities as a public good, individual financial institutes do not have
sufficient incentive to monitor and to appropriately manage their risks from
the perspective of entire economy. Thus, the risk management system should be
stressed and monitored by regulators. Basel II is global standard of such
risk management and is similar to the firm’s risk analysis process in several
respects.

Comparison with Financial Institutes

The risk analysis process resembles that for financial institutes suggested
in the second Consultative Paper published by the Basel Committee on Banking
Supervision in 2001 (Basel II). Basel II proposes three pillars. Pillar I
stipulates minimum capital requirement, Pillar II guides supervisors and
banks to process risks, and Pillar III depicts disclosure requirements about
risks.

Pillar I is particularly relevant to this paper. In Pillar I, the minimum

capital requirements is 8% of risk-weighted capital to buffer shocks to banks.
The appropriate amount of risk-weight capital considers credit, market and
operation risks. The credit risks are computed upon the probability of
default, loss given default and exposure at default. Market risks are
computed with VaR. Operational risk is “the risk of loss resulting from
inadequate or failed internal processes, people and systems or from external
events. This definition includes legal risk, but excludes strategic and
reputational risk,” (Paragraph 644, Consultative documents, 2006). The
operational risk resembles the firm’s internal risk.

The credit-risk analysis of Basel II seems to have connection with the firm’s
risk analysis framework although interviewees deny. Both stress probability
of events and expected damage. The firm additionally includes the extent of
controllability on events, which it is similar to the concept of exposure at
default in Basel II.

Despite the similarities, the firm considers far wider range of risks. Basel
II formally excludes strategic and reputation risk, but the firm regards them
more important. For example, the industry risk of the firm is based on the
strategic concept of Porter’s five forces, which Basel II ignores. The
difference between the two suggests that the purpose of risk management of
the firm is different from that of Basel II. While the latter stresses
measuring and controlling risks, the former focuses on strategy development

27
and organizational implications of new businesses under high Knightian
uncertainty.

Cognitive/ behavioral effect of valuation process

When a firm conducts valuation under uncertainty, it defines and constructs

values by influencing managers’ behavior and by determining rules and
organization structure. The naturally following questions are why it helps
creating values to affect managers’ minds and what the rationales of defining
the rule and organization structure are. For instance, it may not be
necessary to design the valuation process at all, if managers are rational,
committed to the goals of organizations and fully informed. In what follows,
I clarify the causality between valuation process and value creation at
cognitive/ behavioral level.

Schema Building

The valuation process affects the cognitive structure, i.e. schema, of

managers. Let us investigate the issue in intra-managers and inter-managers
perspectives.

Inter-managers perspective: The valuation process assists managers to form

better schema about the social relationship in a firm. Managers need to
comprehend the transactive memory system (Wegner 1987) of a firm to collect
information better and to coordinate among them, which in turn enhances the
transactive memory system of a firm. Transactive memory system comprises the
knowledge of employees and the relationship among the knowledge/ employees.
The valuation process, which guides and presents the relationship between
various parts in decision making, facilitates managers to understand the
networks of knowledge better and, as a result, to form better network schema
and transactive memory system. Better transactive memory system becomes the
core competency of a firm to create further values. Janick and Larrick
(2005)’s research is particularly relevant. First, Janick and Larrick show
that the differences in schematic knowledge of missing relations in social
network affect the accuracy in the encoding and recall of the incompleteness
in the network. Their result thus shows the value of correct network schema
and explains why matters the role of valuation for communication, information
sharing and relationship recognition. Second, they demonstrate that learning
can enhance the schema to recognize missing social relationship. This proves
the educational value of valuation process to guide the managers toward
desirable interaction among them. In addition, valuation process allows
managers at the juncture of incomplete relationship to recognize their status
in the network in order to create more values by bridging the information and
people.

Intra-manager channel: Managers have their own mental models to interpret

information about investment opportunities and to predict situations (Norman
1983). Offering formal frameworks to process information about projects and
to evaluate them financially, a firm implicitly teaches managers
metacognitive strategies in order to accomplish the goal of the organization.

Formalism

Formalism is another mediating channel from ‘using valuation as a broader

concept of a process’ to ‘creating values’. The reasoning is as follows.
First, valuation as a process produces formalism in organizations under
uncertainty. Valuation process appears as guidelines and consultation to

28
employees. Second, the formalism activates procedural justices, but
diminishes organization politics (Ayree et al., 2004). Third, procedural
justice enhances not only task performance, but also context performances
such as interpersonal facilitation and job dedication (Lind & Earley 1992,
Cohen-Charash & Spector 2001, Ayree et al., 2004). Fourth, diminished
organization politics may lessen its destructive effects on performances.

Organization politics can induce employee to engage in unproductive

politicking, and can cause inefficiencies/ resource misallocation.
Organization politics changes the incentives of employees and affects task
performances (Cropanzano et al., 1997; Vigoda, 2001). It also negatively
affects the relationship among co-workers (Witt et al., 2002).

Then, why valuation process creates formalism in a firm? Valuation process

achieves formalism through imposing constraints on managers’ behaviors and
cognition and through standardizing valuation practices.

Constraints on behaviors: Valuation process forces managers to adopt

particular process from identifying investment opportunities to assessing the
attractiveness of projects. For instance, managers should evaluate and
reports risks in very specific ways.

Constraints on cognition: The process influences managers’ recognition of

investments opportunities, attractiveness and risks. First, during the
overall evaluation process, managers should take into account a firm’s mid-
long term strategies. This makes particular investment opportunities more
salient in the mind of managers. Second, a firm recommends managers to apply
valuation-techniques in a specific way. This affects managers’ schema to
judge the attractiveness of projects. Third, managers should categorize risks
considering macro, industry, customer and internal risks, and assess the
risks in terms of expected damage and the likelihood to occur, which in turn
outline the attitude of managers about risks.

Universality: Valuation process generates rules, regulations and guidelines

by stressing shared values and by standardizing evaluation, valuation
technique application, business logic development and risk analysis. This
engenders universality in the organization. In addition, since valuation is
also a communication tool to external and internal audiences, valuation helps
employee sharing values, ideas, information and cooperation. Such interaction
promotes universality.

Organizational learning

Formalization is also related with organizational learning (Levitt & March

1988). Valuation process comprises routines to incorporate various factors
about strategy, organization, implementation, qualitative risks,
idiosyncratic risks, etc. Such routines bear knowledge generated during
capital budgeting practices. The knowledge will increase the performance of
later investments especially when Knightian uncertainty is high. Therefore,
the higher Knightian uncertainty becomes, the more important the benefit of
organizational learning as formalism increases.

Organizational Justice

In many economics and finance literature, internal capital market is

characterized with the competition for the limited supply of internal
resources (e.g. Stein, 1997). Organizational justice literature deals with

29
related issues, but with richer consideration about the relationship during
resource allocation process.

This section examines a firm’s behavior with four dimensions of

organizational justice suggested by Colquitt (2001). The dimensions are
procedural, distributive, interpersonal and informational justices.

Procedural justice

I investigate the relation between valuation process and procedural justice,

using the concepts provided in Leventhal (1976) and Leventhal et al. (1980).

Consistency: Financial analysis determines capital allocation. By making

valuation into routinized process, a firm can ensure capital budgeting occurs
consistently across person and time. This results in consistent functioning
of internal capital market under uncertainty.

Bias suppression: Valuation process suppresses potential biases arising in

the mind of managers during the analysis about uncertainty. The overall
evaluation process starts from checking the correspondence of proposed
projects with the mid/long-term corporate strategies. The emphasis on
corporate strategies reduces arbitrary and biased decisions by individual
managers. In addition, firms require related divisions to undertake reviews
in order to moderate biases that business plans may have. In addition, bias
is reduced with the routines such as detailed guidance and consultations
included in valuation process.

Accuracy of information: In the process of applying valuation techniques, a

firm iterates sensitivity analysis, data gathering and modeling in order to
assist informative decision-making. In risk analysis process, a firm provides
explicit guidelines about which topics managers should investigate in both
idiosyncratic and systematic components as well as Knightian uncertainties.
The framework for developing business logic also helps effective information
searches by helping employee to break down their issues into manageable units.

Interactional justice

Valuation process is also relevant to other dimensions of organizational

justice. By functioning as a communication tool, valuation generates
interpersonal justice. People become more respectful and acknowledge
proprietary information that other teams have during the discussion to
resolve uncertainties. In addition, since valuation process facilitates
information sharing, it ensures informational justice such as truthfulness
and justification, which is particularly important during high uncertainty.
Interpersonal and information justices constitute interactional justice.

Case Conclusion

This case investigates valuation process to assess new businesses under

Knightian uncertainty. Valuation process is an important element of internal
capital market because it evaluates projects and finally determines the
amount of investment. This paper reveals that valuation is more than applying
formulas. Valuation process covers quite broader subjects. Through valuation
process, firms can construct values as well as compute values when
uncertainty is high. First, the process guides managers to consider
corporate-level outlooks overcoming division-specific narrow views and to
enhance communication in organizations. Second, the valuation process offers

30
managers with the frameworks to formulate better strategies. Third, valuation
process also assists organizational learning by routinizing knowledge when
organization deals with Knightian uncertainty. Fourth, using valuation
process in broader manners adds values to the firm by affecting the managers’
schema and by achieving the formalism in the organization under high
uncertainty.

Risk analysis, an important component of valuation process, illustrates the

propositions more clearly. To illustrate, the standard classification of
risks into systematic and idiosyncratic parts does not capture real practices
of valuation. Firms may catalog risks upon strategic views in order to
understand the challenges they face and to manage qualitative risks better.
If such practice of risk analysis helps firms to create better strategies and
to design better organizations based on the strategies, simple binary and
financial notion of systematic and idiosyncratic risks may not be optimal at
all.

Case #2: High Uncertainty/ High Controversy

When both uncertainty and controversy of investment opportunities are high,

organizations not only should make effort to understand the opportunities and
to build expertise, but also need to legitimize it simultaneously. This is
challenging situation. How can one legitimize a project when he knows he
lacks expertise, and other people know he lacks expertise, and he knows other
people know he lacks expertise, ad infinitum?

In this situation, if a team proposes new business, it needs to communicate

clearly: (1) the value that proposed businesses add in excess of the
opportunity cost of using resources for other teams; (2) the research that
allow the organization to make informed judgment; and (3) the capability that
the team and the organization to do the businesses.

The firm is an energy company and the subsidiary of an Asian conglomerate. It

has organized a taskforce to find new business opportunities. The taskforce
recommends several new businesses such as logistics and the collection of the
Internet business. The proposed Internet business and logistics services
share common properties.

First, the firm has no experience in both of them. Thus, the new businesses
are perceived to be highly uncertain to the firm. Second, they are apparently
unrelated to the firm’s core businesses. Thus, the taskforce had to work hard
to prove that the projects are relevant to the firm. The taskforce argued
that the firm should view its retail network as the resource to be utilized
to diversify the firm toward logistics and the Internet. Third, financial
analysis is not important in both business plans. The plan for the Internet
commerce does not include financial analysis at all. The logistics plan
contains a financial model that mainly focuses on explaining various
scenarios. Interviewers argue that:

"... given the constraints in time and resources, we have to focus on

understanding market and developing strategies rather than working on
financial models. Even if we develop quantitative models, people will
keep asking underlying logic and be interested only in business ideas
and strategies. They will not believe math... If they believe math or
financial model, we would have been much comfortable. It is easy to
build models..."

31
Fourth, both plans emphasize that the initial investment is small and that
investments will follow some road maps. Same to real option argument, the
taskforce proposes that the initial small investment would provide the
learning opportunities about the true potential of the businesses. Main
investments follow contingent on the outcome of initial investments. Such
real option argument constitutes main logic to legitimize the projects
together with resource-based view (RBV, Wernerfelt 1995). The taskforce uses
RBV argument to propose retail network as a strategic factor toward
diversification. Real option argument is for mitigating concerns about risks
and for convincing internal competitors to allow shifting internal resources
to the projects. Because the proposed projects require only small investment
in the beginning, competing teams would have fewer objections. RBV argument
focuses on suggesting the close relation between the projects and the main
businesses of the firm through retail networks. Showing the tight relation
makes the projects more legitimate and understandable. Thus, communication is
important part of assessing new businesses.

For simplicity, I will focus on analyzing the investment plan about the
logistics service.

Background

Two reasons exist why the firm wants to find new businesses. First, the
chairperson of the conglomerate asks subsidiaries to find new businesses. The
chairperson has the vision to make each subsidiary of the conglomerate to
become global players in near future. To do so, he presumes that starting new
businesses is the most effective way. The chairperson introduced a
competition among subsidiaries to find new business. Second, in relation to
the first, other subsidiaries want to use the firm’s key resources, i.e.
retail network, to execute their new businesses. Nevertheless, the firm
prefers starting own businesses to renting the networks. The firm worries
that other subsidiaries would take all the glories from the success while the
firm should work only behind to handle less prestigious tasks. In addition,
sharing key resources may work against the firm when corporate reorganization
occurs or at least when the performances of top managers are evaluated.
Following is a quote from a top manager of the firm.

"It is funny that the other subsidiaries propose new businesses upon
our networks and assets. Some neither ask our permission beforehand nor
even seek our opinions. We need our projects."

Controversy and resolution

The project is highly controversial. The taskforce admits that people always
challenge why an energy company wants to engage in logistics business. People
have argued that it is not the core business, and even if it is promising
business, other subsidiaries could do the business better. In addition, the
chairperson of the conglomerate has emphasized using intangible assets for
new businesses. Logistics service uses intangible assets only little, but
involve large capital investment and human resources.

To mitigate those concerns, the taskforce writes business plan to communicate

its strategies about logistics business in the following way.

"First, the business uses our intangible asset heavily, but minimizes
capital investments. It uses our retail networks to access local
customers and to build pick-up center. It forms the alliance networks

32
 with selected competitors and the Internet commerce sites to reach
wider market with small risk taking. Second, the retail network in
combination with our new businesses will create customized experience
for customers. In particular, the Internet business we suggest will
create mutual benefits to both partners and us, and offer comprehensive
services to customers. Third, other subsidiaries will be our first
customer. Thus, our distinct services can create synergies at the level
of our group (conglomerate)."

While RBV and real option arguments have been useful, the taskforce could not
convince top management team successfully. HQ of the conglomerate halts the
project just before implementation.

Uncertainty and resolution

The taskforce views the project full of qualitative uncertainty. First, the
firm has no experience on the business. Second, the logistics business is
dependent on how the Internet industry evolves. The more the Internet
commerce becomes prevalent, the more logistics demands increase. Everyone
agrees that the Internet commerce will become large. However, it is not
certain how quickly it would expand. In addition, the dynamics in domestic
Internet commerce sector add uncertainty. As the price competition among the
large Internet shopping sites intensifies, the profit margin of logistics
services are likely to shrink further due to the pressures from the online
malls. Third, many firms consider entering the market, creating uncertainty
in the extent of future competition. Fourth, the cost of business depends on
what logistics system the firm chooses and how to deal with logistic
complexities. This complicates evaluating the business. The taskforce
recommends combining the advantages of hub-spoke and point-to-point system.
However, it is not certain how they relate the hybrid model with the firm’s
retail networks. Fifth, a few portals dominate domestic Internet landscapes.
The behaviors of them will determine the fate of both internet and logistics
business.

In order to address the uncertainties, the business plan describes market,

operation strategies and investment schedule. Next section explains the
contents of business documents in detail.

Structure of business documents

I have investigated business plans and discussion documents. They comprise

the contents about business concepts, market, competitor, capability,
marketing strategy, operating strategy and financial analysis.

Business concept: The business presumes that the growth of the Internet
economy will make logistics services also grow. The firm plans to utilize the
retail network to implement logistics services. The retail network can serve
as pick-up centers and the platform to serve local needs. The business
attempt to address logistics needs and provides distinct services in
combination with the Internet commerce activities. The business also brokers
between alliance online services and customers through both physical
logistics and the links in web portal.

Market analysis: This section occupies the largest parts in the business
plans. Market analysis focuses on the growth of the Internet/ mobile business
and the demand of the logistics services. The taskforce decomposes logistics
services into three subsets: business-to-business (B2B), business-to-customer

33
(B2C), and customer-to-customer (C2C). Each subset is related with the growth
of the Internet business to estimate the future size. The taskforce finds
that B2C is most promising. While the level of competition becomes higher,
the taskforce argues that the Internet/ Mobile commerce can change market
substantially. Niche players that combine online and offline services are
examples. The data sources for the market analysis are diverse, such as
brokerage firm’s analyst reports, government documents, interviews and survey.

Competitor analysis: The competitor analysis follows several steps. First,

the taskforce categorizes competitors into existing, niche and potential
competitors. Second, it analyzes the competitors in terms of strategic goals,
competitive advantages, performances, financial strength, product/ services
and business systems in order to derive the implication to the firm. This
analysis yields the list of significant existing/ potential competitors.
Third, the taskforce analyzes the (expected) attributes of their services and
products. This yields what position the firm should take. Then, the taskforce
analyzes the situation when the firm will take the position. Fourth, the
taskforce derives alliance strategies the firm needs to win competition and
to share risks. It also applies elementary game theories to describe expected
market dynamics. While such competitor analysis static, it covers various
topics comprehensively.

Capability analysis: The taskforce identifies five required capabilities:

infrastructure, sales network, system to manage network, human capital and
brand. The firm has good resources for sales network and brand, moderate
resources for system and human capital, but none for infrastructure. The
taskforce recommends several strategies to transform the firm’s intangible
resources to core competence of the business in order to offer differentiated
services and operation. For instance, the taskforce suggests using customer
intelligence accumulated in other subsidiaries and formulating strategies in
joint with them. The taskforce also provides the extensive list of intangible
resources in five categories that the firm can use for the new businesses:
knowledge, people, relationship, reputation and synergy with other
subsidiaries. It also lists the required resources that the firm does not
have yet. It provides the list of the resources to acquire in short-/ medium-
/ long-term and the related acquisition plans for the resources.

Marketing strategy: The business plan also deals with marketing strategies in
detail. It is a bit surprising to see such comprehensive marketing plans when
the products/ services to offer are ambiguous and much uncertainty exists. I
believe the development of such detailed marketing strategy is the part of
research to understand the market. We will observe similar intention in the
financial model the taskforce develops.

The marketing strategy starts from customer segmentation. There are four
major customer segments and several minor segments below them. Segment
profiles are also provided in detail. The four major segments are premium,
professionals, service centric and networks. The marketing strategy matches
products/ services to appropriate customer segments. It also analyzes the
firm’s capability and relative position against competitors and alliance
strategies in each segment. Those customer segmentation and profiles are
based on customer survey and alliance survey/ interviews. Next illustrates
some of survey questions.

34
To customers
If xxx types of services are offered, how much are you likely to use them?
(5-Likert scale)
In which services do you feel most uncomfortable while dong online
shopping? (1) price, (2) delivery, (3) purchase, (4) return.

To alliances
Do you offer delivery option to customers? If you do not, what is the reason?
If customers increase by 10% due to new logistics options, how much do you
willing to pay for the service?
Are you interested in xxx types of services?

The fieldwork on customer and alliances becomes the basis to identify

customer needs and characteristics, which constitute marketing strategies.

Operation strategy: Operation strategy and market analysis are the two most
highlighted sections in the business plans. Operation strategy is composed
with five parts: road map, model, products, profiting and management.

• Road map: The investment occurs in three stages as (1) Establishing

market presence, (2) Nation-wide coverage and (3) Full service network.
The key message is that the initial investment will be small and that
it is easy to exit when business does not go well. In the Establishing
market stage, the firm builds alliance relationship, on and off-line
customer database and B2C basic services. The goal of this period is to
identify competitive positioning. In the Nation-wide coverage stage, it
establishes deep local and broad nationwide network together. In the
Full service network stage, it becomes a player with full service
network. In this stage, it builds end-to-end system, starts global
services and has solid market leadership. The business plan includes
detailed spreadsheet that lists various activities and their time lines
at every month for next five years.

• Model: The business intends to maximize the usage of the intangible

assets of the firm. For example, it will use customer information
present in other subsidiaries. While fully utilizing retail network, it
outsources the functions that require heavy capital investment or labor
forces. Retail network does central role to collect customer
information and recruiting. Customer service is the most important
differentiating factor. The business starts from capital city. Other
areas are serviced through outsourcing. The target city will be
decomposed into four areas in which hubs exist. In order to function as
the broker between vendors and customers, it is important to have
payment system.

• Product: The business will have its sister online shopping sites. Given
the focus on local information, the sites will be tailored to locals/
community characteristics. Such local services and the presence of
retail network will create synergy. Local vendors and communities
should be absorbed in alliance network. Therefore, the products have
integrated services of commerce, contents and community.

• Profiting: One vehicle should perform 60 services per day. The demand
from subsidiaries will be our initial revenue source. However, the

35
 business has other diverse sources of profit. For example, the business
can provide to alliance the services such as payment, return handling,
producer-customer match and advertisement.

• Management: Brand management is important. The business will use the

umbrella brand of the conglomerate, but should create distinct identity
through creative services. Initial marketing is important. Also, it is
important to control alliances to maintain the firm’s prestigious brand.
Uniform and vehicles design should communicate single message in line
with conglomerate brand. Localized promotion activity is important.
Marketing activities will target the customers visiting the firm’s
retail network. The firm should convince alliances that they will
expand their services without incurring extra costs much. For example,
extra orders come from online services that the firm operates.
Alliances can do advertisement easily based on the firm’s network and
customer information. Delivery services will differentiate the
alliances from competitors to grant them with competitive advantage.

Financial analysis: Financial model is organizational one pretending to be

SCBM. Financial model produces several market insights. The focus of the
model is not to compute NPV for investment decision-making, but to simulate
strategies and the understanding about market. The output of financial model
is the set of recommendations. First, the financial model finds that economy
of scale is important for hub model, while pricing and cost leadership are
important for end model. End model is built on the close contact between
retail network and customers. Second, it finds that the firm should ensure
end model to work as efficient as the model of incumbent players while
providing some unique services based on firm’s intangibles. It recommends
starting the end model at initial stage. Third, vehicles should process at
least 60 services per day. In order to do that, it is not enough to rely on
the demand from subsidiaries only. The firm should offer both B2B and B2C
services to find additional demands. Fourth, financial model suggest charging
1.5 dollars per box. Fifth, it is important to start hub mode in later stage
with economy of scale.

The financial model derives the recommendations based on scenario analysis.

The first scenario is positive one in which the competition among market
players are moderate, and the firm is successful to maintain operational
efficiencies. The second scenario is negative in which competition is intense,
and the firm loses the efficiencies in the operation of vehicles. In this
case, the firm cannot generate profits and needs to position it as niche
player. The niche play is to use retail network as pick-up center because
this service is expected to make profit even in negative situation. The third
scenario is moderate in which the competition is at current level, and
operational efficiency is at industry average. In this case, it is important
to acquire certain market share, and to differentiate the firm from
competitors with broad services.

The financial model also includes sensitivity analysis on several factors.

The first factor is vehicle efficiency. Vehicle efficiency is found to be
highly important affecting all services. The second factor is market price.
Because of differences in price sensitivity, the revenues of some services
are affected, but not all. This analysis also reveals the importance of
unique services. The third factor is market share. The change in market share
does not affect profitability ratios much, but the firm will lose competitive
advantage due to economy of scale in the end. Sensitivity analysis concludes
that the firm should focus on vehicle efficiencies to succeed in new business.

36
Both scenario and sensitivity analysis clearly show that the financial
analysis is an organizational model. The purpose of financial analysis is not
to evaluate investment opportunities and to decide whether to invest or not.
Instead, financial analysis is designed to identify what strategic variables
derive the performance and in consequence, what strategies the firm should
take. The financial model also communicates and constructs the structure of
projects.

Clash of ideas

The project team uses RBV and real option argument to support the new
businesses. RBV argument is to rationalize the diversification strategy using
the resource of the firm. Real option argument is to alleviate the concerns
about the uncertainties in the project stressing the learning and the
easiness in exiting from the project.

In comparison, opposing groups tend to use the argument of transaction cost

economics and core competencies. While they admit that the resource of the
firm is valuable for new businesses, they also point out that better
arrangement can exist. The firm may not be the best organization to perform
the project given its capability, experience and core competencies. In this
situation, the conglomerate, other subsidiaries and even the firm would
obtain more benefit if the firm could lend the resource to other
organizations and just share the profit from new projects.

The project team does not believe the idea. First, with the separation of
ownership and usage of the resources, the users may abuse the resources. At
least, it is possible that the alternative usage reduces the value of the
resource for the main usage. The firm bears the responsibility of maintenance
as the owner of the resource. Such misalignment of incentives between
ownership and usage would create more problems for the firm. Second, it is
difficult to measure the contribution of various groups to the outcomes of
new businesses. Since it is difficult to measure the contribution, it is hard
to design the compensation and other internal contracts. Third, such internal
arrangement is generally not enforceable as the legal contracts in markets
are. The parent company can change the corporate structure or top management
team anytime. Such situation makes internal arrangement irrelevant to the
firm. More seriously, the consequences of the projects can trigger the
changes in organizational structure. For instance, if the new business upon
the resource becomes successful, the people in the new business could take
over the important position in the firm as well as the ownership on the asset.
In the event of corporate restructuring, the current management of the firm
may completely lose the control over the resources. Firm is a device to
address hold-up problems. This case shows that hold-up problem can be even
more serious in a firm due to the lack of enforcement devices.

In addition, property right argument has been also used, although not so
frequently as RBV or real option, to deal with the controversy. The firm is
indispensable to the resource. An agent i is indispensable to an asset a if,
without agent i in a coalition, asset a has no effect on the marginal product
of investment for the members of that coalition. Hart and Moore (1990) show
that if an agent is indispensable to an asset, then he should own it. The
firm is indispensable to the resource because, if the resource has no value
as the retail network of the firm, the resource would lose the value for all
other opportunities. The firm could have argued that the strong ownership
should belong to the firm. Ownership means the residual right, i.e. the right

37
to decide the use of assets except to the initial specification about the
usage. The specialized definition of the residual right is the ability to
exclude others from the usage of the resource. Thus, the firm could have
argued that it should have the rights to exclude other parts from use.

Another relevant perspective is dominant logic view for diversification

(Prahalad & Bettis 1986). In dominant logic view, what matters is the
cognition of top managers in diversification. In that sense, the firm may
relate logistics service with energy businesses. This is exactly what
occurred in the taskforce.

In sum, these various perspectives suggest interesting implications. The

debate in academia about firm boundary exists in business practices as
ideologies. The managers select appropriate firm theories to conduct or to
defend their projects. Surprisingly, they use the various firm theories even
when they have never learned the theories such as TCE, property right theory
of firms or RBV through formal education. The managers become to bear the
theories from their experience, intuition and most importantly in the middle
of controversy and struggle over resources.

Conclusion
Organizational Capital Budgeting Model (OCBM) is a general theory of capital
budgeting that admits the traditional financial capital budgeting model as a
special theory and that understands capital budgeting as an organizational
phenomenon. Therefore, OCBM not only broadens the traditional model, but also
explains the behaviors of firms using quasi/non-financial version of capital
budgeting. I demonstrate the validity of OCBM with qualitative studies. The
ethnography about Asian conglomerates is carefully constructed. Asian
conglomerates are important dataset to study organizational decision making
because of their size, scope, controversial behaviors and global presence.
The ethnography supports and enriches OCBM. OCBM in turn contextualizes the
ethnographic data.

OCBM can generate rich managerial implications because it is both descriptive

and normative model. OCBM is a descriptive model, but appears as normative to
individual decision makers. I develop OCBM to describe observed capital
budgeting practices about which existing models offer only limited
explanation. I enrich OCBM with multiple qualitative studies. The iterative
process between theory and qualitative data completes OCBM. Thus, OCBM is
descriptive. In addition, since I frame OCBM as an optimization problem that
managers face, OCBM is a normative model.

I assume that investment decisions are rational on average. Then, with enough
data, irrational behaviors or mistakes in valuation cancel out because
average rational actors do not make mistakes systematically. Since OCBM is
the abstraction of data to find commonalities, it discusses average behaviors.
Thus, it is rational model. In turn, when we observe individual cases with
the lens of OCBM, we may observe aberrant behaviors in capital budgeting.
Then, OCBM can prescribe firms how to conduct valuation.

The rationality of OCBM needs further discussion. OCBM models the average
behavior through contextualization, not through statistical technique. I
cannot apply law of large numbers with the dozens of qualitative studies I
have conducted. In addition, this paper includes only subset of my
qualitative studies. Instead, in ethnographic data, I narrate in detail what

38
interviewees say about the rationale of practices and what I find in archival
data about how organizational environment leads to the observed pattern.
Comparing what such data tell and what theories tell, I can abstract the
findings to discuss what is rational or not. Of course, readers can disagree
about my interpretation and abstraction, and I do not argue that only my
interpretation is true. What I can present as an alternative is to delineate
this whole process, so that future researches propose disagreement and
attempt alternative interpretation and theories. I believe this is typical
academic discourse whether data are qualitative or quantitative. Many papers
illustrate how same data leads to different interpretation depending on
theories or econometric models.

OCBM argues that it is justifiable for managers to include various factors

such as financial, strategic, and organizational/ process factors. Business
plans include those factors. Therefore, we can argue that many firms make
investment decision on business plan, not only on NPV or on SCBM. The use of
business plan does not exclude the possibility that SCBM is still the only
important criteria in decision-making. Business plans are flexible enough to
include SCBM and OCBM in various ways.

First, business plan may include SCBM only. I do not recommend this method.
This case is rare. None of my qualitative data show the case in which
managers make decision only on excel file that includes SCBM. Second,
business plan may include many non-financial discussions, but all converge
discussing the assumptions and outputs of SCBM. Thus, the output number of
SCBM is the most dominant criteria in business plan. This method can be used
for monitoring or repetitive investment. It is also similar to stock
analysts' report. Stock analysts recommend the strategies to invest in stocks.
The recommendation is based on target price, which is usually NPV, multiples,
discounted earnings, etc. In order to support the assumptions of calculation,
stock analysts tend to explain industry trends and company policies in detail.
Third, business plan can discuss many non-financial factors with different
importance allocated. Financial models are one of the many important criteria
in decision-making, but not everything. Sometimes, financial models are used
to communicate or confirm intuitions. I observe this case most often. Fourth,
business plan may not include any quantitative discussion at all. If
uncertainty is very high, managers may have to make too many assumptions or
to spend too much time on quantitative analysis. In this case, financial
analysis may not be a cost efficient way of decision-making.

Among the four ways aforementioned, managers can choose appropriate ones for
business plan to conduct valuation. The important considerations are
controversy and uncertainty. When both controversy and uncertainty are low,
managers regard the valuation process as pricing. When controversy is high,
managers can use valuation process to communicate. When uncertainty is high,
managers may view valuation as research process.

When valuation becomes communication, managers need to investigate audiences.

The mode of communication can be technical/ non-technical, top-down/ bottom-
up, tacit/ explicit, etc. Managers can choose appropriate mode in
consideration of target audiences and the relationship with them. Business
plan for the valuation should be understandable. It is also the means to
reduce or to coordinate the interests of sub-coalition. For instance,
business plans can be used as guidelines or instructions during
implementation phase. Managers need to tailor the valuation process and
business plan in order to fit them properly into the audience and
communication mode.

39
When valuation becomes research process, managers decide whether the process
should be exploratory/ descriptive, evaluative/ constructive, etc. I find
firms often disregard financial models when business plan should provide
exploratory contents or when firms need to design the architecture of
projects. In those cases, valuation process becomes similar to searching
process. Since the manner of searching affects the architecture of new
projects much, managers should keep in mind that searching and valuation
often become inseparable.

I consider the valuation process and business plan as the deliberate choice
of managers in value space. The design of business plan is similar to the
choice of valuation strategies (frames). The choice of projects is the choice
of investment opportunity (objects). In order to make this value space idea
explicit, I propose a template in what follows.

I suggest writing a report for each project. In each report, the first column
presents OCBM variables such as perceived value, accuracy, controversy and
uncertainty. Accuracy item includes gross accuracy and the required effort to
achieve the accuracy. Controversy items can be classified in terms of
audiences. Uncertainty items can be categorized with well-known strategy
frameworks such Porter's five forces, RBV, real option, etc.

The first row includes valuation strategies. They are financial criteria such
as NPV, strategic criteria such as synergy, learning opportunities, etc. and
process criteria such as valuation process, implementation process,
monitoring process, research methods and communication modes. The process
criteria are most diverse and are expected to vary substantially across firms.

In this value space, managers can fill the cells with numbers in accordance
with first column and row items. Once numbers are filled, managers can pick
only one cell, or select multiple cells in order to creatively integrate them
to construct business plan. Managers iterate this process for each project.
Managers can pick only one project or select multiple projects and combine
them in order to design new business architecture.

My suggestion is only one way to apply OCBM in selecting or constructing

projects. As my qualitative studies show, firms implement OCBM in very
creative and diverse manners. I encourage managers to be creative in applying
OCBM.

Appendix
Value space and existing literature

Similar concepts to value space appear in many theories in strategy and

organization. For instance, various theories distinguish endogenous and
exogenous variables. The value of endogenous variables may vary with the
exogenous variables. A range of outcomes occurs in response to how selected
endogenous and exogenous variables are matched. More generally, there are
internal and external variables. Internal variables are inside the boundary
of organizations, while external variables are from outside of organizations.
The interaction between internal and external variables produces outputs.

Since value space idea occurs in many theories, value space is a natural
benchmark from which I can discuss the commonalities and differences among

40
theories. Similarly, since value space is a central concept of OCBM, I can
find how OCBM inherits the intuition of various theories and how OCBM differ
from and contribute to literature. In the following literature review, I
discuss several theories with value space framework.

In Standard Capital Budgeting Model (SCBM), we can conceptually distinguish

between frames and objects. Nonetheless, since NPV is the optimal solution,
the distinction is irrelevant. The value space becomes one dimension in which
the frame axis degenerates into a singleton, i.e. NPV. When NPV meets an
investment opportunity, valuation occurs. Since any deviation from NPV is
suboptimal, SCBM imposes NPV to all firms to evaluate investment
opportunities in all situations. Thus, there is clear dichotomy between
frames and objects. Since OCBM rejects such dichotomy, it is clear why SCBM
is a special case of OCBM.

In behavioral decision theories, choice strategies are frames, and objects

are choice sets. The concept of constructive preference is particularly
relevant here. Bettman, Luce and Payne (1998) assert that preferences are
constructed at valuation and that the construction process is subject to both
human information processing system and task properties (Payne, Bettman and
Johnson 1992; Slovic 1995; Bettman, Luce and Payne 1998; Payne, Bettman and
Schkade 1999). The process of construction involves the selection of choice
strategies. Choice strategies include WADD (weighted adding), LEX
(lexicographic), SAT (satisficing), EBA (elimination-by-aspects), EQW (equal
weight), MCD (majority of confirming dimensions), FRQ (frequency of good
and/or bad features), CCM (componential context model) and RC (random choice).
Without the choice of choice strategies, the preference of an object cannot
be determined ex-ante. Thus, the behavior of selecting choice strategies is
termed as constructive preference.

In resource-based view (RBV, Wenerfelt 1995 for review), objects are

resources. Since the deployment of resources defines strategies, we can
regard strategies as frames. Boundedly rational firms make choice in
strategy-resource space in order to maximize performance. RBV has important
implication to OCBM. First, I can regard investment opportunities as the
valuable resources upon which firms can create sustainable competitive
advantage. Second, more importantly, we can view the choice of valuation
methods as the deliberate strategic choice. Well-chosen strategies may
include more than computing NPV of an investment opportunity. Then, the
apparent deviation from NPV may not be a mistake, but the strategic action to
balance the accomplishment of various strategic goals with the computation of
values. The comparison of RBV and OCBM also suggests substantial managerial
implication. Managers should understand valuation as a strategic choice
rather than mathematics.

In dynamic capabilities framework, frames are processes, and objects are

rapid technological changes. The processes include both managerial and
organizational ones such as “the way things are done in the firm, or what
might be referred to as its routines, or patterns of current practice and
learning” (Teece, Pisano & Shuen 1997). The ability of deploying process
under rapid technological changes determines the competitive advantage of
firms. Dynamic capabilities mean such deployment ability. Dynamic
capabilities framework and RBV suggest similar implication to OCBM. In
environment of rapid technological changes, investment opportunities may
change quickly. OCBM argues that the right choice of valuation strategies in
response to dynamic investment opportunities generate superior performance.
The ability to conduct proper valuation is a considerable dynamic capability.

41
In behavioral theory of firm (BTF) by Cyert and March (1963), frame is
routine, and object is problem. Under bounded rationality, firms rely on
routines to solve problems. It is hard to change routines. They change only
incrementally to resolve problems. Besides value space, BTF and OCBM share
important intuition such as controversy and uncertainty. For instance, both
regard organization as a collection of sub-coalitions.

Evolutionary economics (Nelson & Winter 1982) is an important descendent of

BTF. Thus, same to BTF, frame is routine, and object is problem. Evolutionary
economics assume that it is hard to change routine. Nelson and Winter
consider routine as genes in biological evolutionary theory. Routine
functions as memory, truce, target (control, replication and imitation),
skills and optimization. Thus, similar to OCBM, the routine, as a frame,
addresses controversy and uncertainty.

Organizational learning (e.g. Levitt & March 1988) literature has been
influenced by BTF and evolutionary economics. Similar to BTF and evolutionary
economics, frame is routine, and object is information. I view information as
a subset of problems in organizational learning literature. Organizational
learning is the routinization process in which organizations match routines
with information. Routines bear organizational knowledge. Since investment
opportunities appear as information to firms, valuation activities can be
regarded as learning the investment opportunities by matching valuation
routines and investment opportunities. Such learning can mean more than
computing values. As organizational learning argues, learning involves the
communication and storing of information, so that organization can retrieve
knowledge for later purpose. In addition, building absorptive capacity (Cohen
& Levinthal 1990) is an important purpose of current learning activity. Such
consideration for learning influences what valuation strategy a firm should
choose, and this is what OCBM investigates.

In organizational ecology (OE; Hannan & Freeman 1977), frame is

organizational forms, and object is environment. It determines the vital
rates of the population of organizations how fit an organizational form is to
environment. The vital rate is the rate of population change due to founding
and mortality. Organizational forms determine positions such as specialist in
a niche or generalists. Hsu and Hannan (2005) propose the identity-based
conceptualization of organizational form, which sees organizational forms as
types of identities. Organizational identities is the set of social codes/
rules that audience members hold. It is similar to genres of art works

OCBM suggests that valuation patterns can be an important element to

determine the identity of organization. First, investment occurs in all for-
profit organizations. Second, investment is important enough to determine the
vital rates of organizations. Third, valuation determines the success of the
investment. Thus, OCBM and organizational ecology can be closely related.

In neoinstitutioanl theory (Scott 2007, 3rd ed.), frame is institution/

institutional change, and object is institutional environment. An institution
is legitimate or illegitimate depending on how the institution rests on
institutional environment. The rise of an institution or institutional change
occurs through structuration process. Structuration process involves the
interaction of bottom-up and top-down processes. Top-down process denotes how
higher-level structures outline the manner lower-level actors operate.
Bottom-up process discusses how lower-level actors form the context in which
they operate. Structuration is an interesting framework. I will use

42
structuration framework to show how OCBM models and suggests testing the
interaction of structure and agent, one of the classic issues in sociology.

Next table summarizes the comparison between OCBM and other theories in
strategy and organization literature.

Value space in Commonalities in value space Differences

theories Frames Objects Outcome ∆ Frame
OCBM Valuation Investment Valuation/ Choice
strategies opportunities Goal
SCBM NPV types Investment Valuation Not efficient
opportunities
Behavioral Choice Choice sets Preference Choice
decision strategies
theories
Resource-based Deployment Resources Competitive Strategy
view advantage
Dynamic Process Technological Competitive Dynamic
capabilities change advantage capabilities
Transaction cost Transaction Transactions Efficiency Choice
economics modes
Behavioral Routines/ Problems Performance Hard
theory of the procedures
firm
Evolutionary Routines Problems Performance Hard
economics
Organizational Routines Information Learning Routinize
learning
Organizational Organizational Environment Vital rates No. New
ecology (OE) forms species.
OE: Hsu & Hannan Identities/ Environment Vital rates Observed
2005 Genre
Neoinstitutional Institution Institutional Legitimacy Structuration
theory (changes) environment

Methodology

Ethnography, comparison and contextualization constitute anthropological

triangle upon which I ground my research. Ethnographic triangle defines the
method to write ethnography from ethnographic data (Barnard & Spencer, 1996).
Also, I try to be dialectic in order to relate new theory with data by
iterating deductive and inductive logic. This iteration makes my research
both normative and descriptive.

More specifically, first, I start from the deductive logic upon the existing
theories about capital budgeting. I compare their predictions with well-known
empirical findings. This reveals inconsistencies between theories and
observations, which I call as empirical irregularities. Second, I hypothesize
that the empirical irregularities exist as a blank in the theoretical
structure of existing capital budgeting models. I confirm the theoretical
lacuna using alternative perspectives. Third, as an induction, the empirical
irregularities and conjectured theoretical weakness become the basis of new
theorizing about capital budgeting. Fourth, the new theory generates distinct
predictions from old theories through deductive reasoning. Fifth, at the same
time, I conduct ethnography. The ethnographic data is compared and

43
contextualized upon the new theory. Sixth, I update the theory through
inductive reasoning. Finally, I iterate the process.

This is a dialectic process similar to grounded theory (Glaser & Strauss 1967,
Strauss & Corbin 1990) and its variations such as Eisenhardt (1989) and
Eisenhardt & Graebner (2007). As a caveat, I caution against interpreting the
limitations and critiques about my methods as those about the grounded and
its variant methods. Next explains the step in detail.

Analysis of lacuna in financial models

My research begins with analyzing the disruptions of the existing framework,

i.e. Standard capital budgeting model (SCBM). SCBM has been both normative
and descriptive framework for capital budgeting practices. SCBM is based on
financial and economic reasoning. My analysis about SCBM is two-folded. First,
I propose SCBM has limited ability in explaining observed capital budgeting
practices. When SCBM tries to explain practices, it requires disregarding
deviations as uneducated or insignificant routines. However, I argue they are
integral parts of capital budgeting. Second, SCBM exhibits the logical lacuna
at theoretical level. While capital budgeting is an organizational phenomenon
in internal capital market, SCBM has not incorporated organization theories
or modeled organizational contexts seriously. The two arguments are
associated with each other. The empirical irregularities appear as the lacuna
in theoretical space. In turn, the irregularities realize the theoretical
weakness. This constitutes my research question and requires new solution.

Construction of OCBM

In order to address the issues identified in the first step, I propose a

solution at theoretical level instead of ignoring or reclassifying observed
capital budgeting practices into the frameworks of SCBM. Organizational
Capital Budgeting Model (OCBM) is my proposal. First, in order to address the
theoretical lacuna of SCBM, I try to fill it with organization theories.
Capital budgeting happens in internal capital market, which is an
organization. Thus, I conjecture capital budgeting theory should be embedded
in broader organization literature. Second, as for empirical irregularities,
I trace what underlying variables may influence the extent of empirical
irregularities. I summarize the identified variables into one internal and
one external variables of organization: controversy and uncertainty, which is
qualitative and nonnumeric risk (Knight 1921, Keynes 1921). Since the two
variables have been important in organization theories and sociology, they
fit my plan of introducing organization theories into capital budgeting in
order to build OCBM.

Ethnography

This thesis puts much emphasis on detailed descriptions. While SCBM has
limitations, this does not render OCBM valid. I select ethnography as
empirical strategy. I suggest ethnography is valid to review theories
including OCBM and to complete theorizing OCBM. Several reasons exist. First,
existing theories lack the space to incorporate field data which arises as
chunks of diverse information. It is serious limitation because the empirical
irregularities against SCBM appear as episodes, for example. See my
motivating examples in appendix.

Second, the deviations from NPV, the representative SCBM, are understood as
the phenomenon marginal, wrong, exhibiting lack of understanding, confined to

44
small firms, occurring among uneducated managers, etc. (Brealy, Myers & Allen
2005, Graham & Harvey 2001). The phenomenon may not be marginal. Instead, I
conjecture they are the important foundations to apply capital budgeting
techniques rigorously and appear more often when capital budgeting is
significant activities. To trace such conjecture, ethnography is the most
proper option.

Third, detail-oriented qualitative studies contrast highly general financial

frameworks. Thus, the discovered particularities can be the basis of new
models to overcome traditional financial models. Quantitative methods have
limitations to reveal the hidden motives and assumptions of managers when
they conduct capital budgeting.

Fourth, I propose that the use of qualitative studies has relation with the
progresses and regresses of firm theory. As discussed in forwards, capital
budgeting occupies an important position in firm theory. Firm theory has not
developed in a smooth and cumulative way. Instead it has experiences abrupt
development, periods of retrogress and resuscitation. I argue the use of
qualitative studies can partially explain the dynamic history of firm theory.

Since the comparison with theories is an important objective of ethnography,

it is important to provide unbiased description of practices and to reveal
the limitations of theories even including OCBM. The extents of revealed
limitations in theories tell whether ethnographic data just confirm theories,
fill in the blanks in theories or becomes foundations to produce new theories.
In my case, ethnography reveals and fills in the blanks in OCBM, but becomes
the foundation to overcome SCBM, positioning OCBM as a generalization of SCBM.

Contextualization

OCBM helps me to interpret and abstract the ethnographic data. Also, the data
reveals certain problems of theories and allow me to rebuild OCBM. This
process achieves the persuasiveness of my research. This technique of
contextualization instructs how to advance OCBM, how to achieve fundamental
shift from SCBM and how to suggest OCBM as a significant generalization of
SCBM using ethnographic methodology. Thus, this process produces the refined
version of OCBM and prescriptive arguments. Further development occurs in the
subsequent iteration processes.

Iteration

I develop OCBM believing that it provides useful references to understand

practices. The ethnography challenges the original OCBM with unexpected
findings. I interpret the challenges as the request for the flexible
theoretical structure instead of the outright failures in OCBM and SCBM. This
iterative theorizing process eventually enriches OCBM by introducing new
variables such as communication, research and concepts such as exploratation.
The upgraded OCBM can not only explain the ethnography better, but also
suggest the possibility of application beyond the ethnographic data. I
suggest such challenges may occur to other researches. Thus my experience of
theorizing for flexible structure can be useful to them.

References
Aryee, S., Chen, Z. X., & Budhwar, P. S. (2004). Exchange fairness and
employee performance: An examination of the relationship between

45
 organizational politics and procedural justice. Organizational Behavior
and Human Decision Processes, 94(1), 1-14.
Barnard, A., & Spencer, J. (1996). Encyclopedia of Social and Cultural
Anthropology. Routledge.
Berger, P. L., & Luckmann, T. (1966). The Social Construction of Reality: A
Treatise in the Sociology of Knowledge. New York.
Bettman, J. R., Luce, M. F., & Payne, J. W. (1998). Constructive Consumer
Choice Processes. Journal of Consumer Research, 25(3), 187-217.
Bower, J. L. (1970). Managing the Resource Allocation Process: A Study of
Corporate Planning and Investment. Harvard Business School Press.
Brealey, R. A., & Myers, S. C. (2002). Brealey & Myers on Corporate Finance:
Capital Investment and Valuation (1st ed., p. 500). McGraw-Hill.
Brealey, R. A., Myers, S. C., & Allen, F. (2005). Principles of Corporate
Finance (8th ed., p. 1028). McGraw-Hill/Irwin.
Cohen, W. M., & Levinthal, D. A. (1990). Absorptive Capacity: A New
Perspective on Learning and Innovation. Administrative Science Quarterly,
35(1).
Cohen-Charash, Y., & Spector, P. E. (2001). The Role of Justice in
Organizations: A Meta-Analysis. Organizational Behavior and Human Decision
Processes, 86(2), 278-321.
Colquitt, J. A. (2001). On the dimensionality of organizational justice: a
construct validation of a measure. J Appl Psychol, 86(3), 386-400.
Cropanzano, R., Howes, J. C., Grandey, A. A., & Toth, P. (1997). The
relationship of organizational politics and support to work behaviors,
attitudes, and stress. Journal of Organizational Behavior, 18(2), 159-180.
Cyert, R. M., & March, J. G. (1963). A Behavioral Theory ofthe Firm.
Englewood Cliffs, 1, 963.
Eisenhardt, K. M. (1989). Building theories from case study research. Academy
of Management Review, 14(4), 532-550.
Eisenhardt, K. M., & Graebner, M. E. (2007). Theory Building From Cases:
Opportunities and Challenges. The Academy of Management Journal (AMJ),
50(1), 25-32.
Glaser, B. G., & Strauss, A. L. (1967). The Discovery of Grounded Theory:
Strategies for Qualitative Research. Aldine Transaction.
Gordon, R. J. (2000). Does the “New Economy” Measure up to the Great
Inventions of the Past? Journal of Economic Perspectives, 14(4), 49-74.
Graham, J. R., & Harvey, C. R. (2001). The theory and practice of corporate
finance: evidence from the field. Journal of Financial Economics, 60(2-3),
187-243.
Granovetter, M. (1985). Economic Action and Social Structure: The Problem of
Embeddedness. American Journal of Sociology, 91(3), 481.
Hannan, M. T., & Freeman, J. (1977). The Population Ecology of Organizations.
American Journal of Sociology, 82(5), 929.
Hart, O., & Moore, J. (1990). Property Rights and the Nature of the Firm.
Journal of Political Economy, 98(6), 1119.
Hobbes, T. (1881). Leviathan. Cambridge University Press New York.
Hsu, G., & Hannan, M. T. (2005). Identities, Genres, and Organizational Forms.
Organization Science, 16(5), 474-490.
Janicik, G. A., & Larrick, R. P. (2005). Social network schemas and the
learning of incomplete networks. Journal of Personality and Social
Psychology, 88(2), 348-364.
Katz, R. (1982). The effects of group longevity on project communication and
performance. Administrative Science Quarterly, 27(1), 81-104.
Keynes, J. M. (1921). A Treatise on Probability. Cosimo Classics.
Knight, F. H. (1921). Risk, Uncertainty and Prot. New York: Kelley.

46
Kogut, B., & Zander, U. (1992). Knowledge of the firm, combinative
capabilities, and the replication of technology. Organization Science,
3(3), 383-397.
Leventhal, G. S. (1976). What Should Be Done with Equity Theory? New
Approaches to the Study of Fairness in Social Relationships. .
Leventhal, G. S., Karuza, J., & Fry, W. R. (1980). Beyond fairness: A theory
of allocation preferences. Justice and social interaction, 3, 167–218.
Levitt, B., & March, J. G. (1988). Organizational Learning. Annual Reviews in
Sociology, 14(1), 319-338.
Lind, E. A., & Earley, P. C. (1992). Procedural Justice and Culture.
International Journal of Psychology, 27(2), 227-242.
Lind, E. A., & Tyler, T. R. (1988). The Social Psychology of Procedural
Justice. Plenum Pub Corp.
Nelson, R. R., & Winter, S. G. (1982). An Evolutionary Theory of Economic
Change. Harvard University Press.
Norman, D. A. (1983). Some observations on mental models. Mental Models, 7-14.
Payne, J. W., Bettman, J. R., & Johnson, E. J. (1992). Behavioral Decision
Research: A Constructive Processing Perspective. Annual Reviews in
Psychology, 43(1), 87-131.
Payne, J. W., Bettman, J. R., & Schkade, D. A. (1999). Measuring Constructed
Preferences: Towards a Building Code. Journal of Risk and Uncertainty,
19(1), 243-270.
Porter, M. E. (1985). Competitive advantage: creating and sustaining superior
performance. Free Press New York.
Prahalad, C. K., & Bettis, R. A. (1986). The dominant logic: a new linkage
between diversity and performance. Strategic Management Journal, 7(6),
485-501.
Rawls, J. (1971). A Theory ofJustice. Cambridge, Mass, 1, 33.
Scott, W. R. (2007). Institutions and Organizations: Ideas and Interests
(Third Edition., p. 280). Sage Publications, Inc.
Simon, H. A. (1945). Administrative Behavior: A Study of Decision-Making
Processes in Administrative Organization. New York.
Slovic, P. (1995). The construction of preference. American Psychologist,
50(5), 364-371.
Stein, J. C. (1997). Internal Capital Markets and the Competition for
Corporate Resources. Journal of Finance, 52(1), 111-133.
Strauss, A. L., & Corbin, J. M. (1990). Basics of qualitative research. Sage
Publications Newbury Park, Calif.
Teece, D. J., Pisano, G., & Shuen, A. (1997). Dynamic Capabilities and
Strategic Management. Strategic Management Journal, 18(7), 509-533.
Vigoda, E. (2001). Reactions to organizational politics: A cross-cultural
examination in Israel and Britain. Human Relations, 54(11), 1483.
Wegner, D. M. (1987). Transactive memory: A contemporary analysis of the
group mind. Theories of group behavior, 185-208.
Weick, K. E. (2001). Making Sense of the Organization. Blackwell Publishers.
Wernerfelt, B. (1995). The resource-based view of the firm: ten years after.
Strategic Management Journal, 16(3), 171-174.
Williamson, O. E. (1975). Markets and Hierarchies. New York.
Witt, L. A., Kacmar, K. M., Carlson, D. S., & Zivnuska, S. (2002).
Interactive effects of personality and organizational politics on
contextual performance. Journal of Organizational Behavior, 23(8), 911-926.

47