V O LU M E 2 3 | N U M B E R 2 | S P RIN G 2 0 1 1

Journal of

APPLIED CORPORATE FINANCE

A MO RG A N S TA N L E Y P U B L I C AT I O N

In This Issue: Risk and Valuation

Downsides and DCF: Valuing Biased Cash Flow Forecasts
CARE/CEASA Roundtable on Managing Uncertainty and Risk

8
18

Richard S. Ruback, Harvard Business School

Panelists: Steve Galbraith, Maverick Capital; and
Neal Shear. Moderated by Trevor Harris, Columbia University

How the U.S. Army Analyzes and Copes with Uncertainty and Risk

34

Major Hugh Jones, U.S. Military Academy

Accounting for Sovereign Risk When Investing in Emerging Markets

41

V. Ravi Anshuman, Indian Institute of Management

Bangalore, John Martin, Baylor University, and Sheridan
Titman, University of Texas at Austin

Accounting for Risk and Return in Equity Valuation

50

Stephen Penman, Columbia Business School

Morgan Stanleys Risk-Reward Views: Unlocking the Full Potential of Fundamental

Analysis

59

Guy Weyns, Juan-Luis Perez, Barry Hurewitz, and

How Corporate Diversity and Size Influence Spinoffs and Other Breakups

69

Vlad Jenkins, Morgan Stanley

Gregory V. Milano, Steven C. Treadwell, and Frank Hopson,
Fortuna Advisors LLC

The Three-Factor Model: A Practitioners Guide

77

Javier Estrada, IESE Business School

Valuing Companies with Cash Flow@Risk

85

Franck Bancel, ESCP Europe, and Jacques Tierny, CFO,

Gemalto

Valuing an Early-Stage Biotechnology Investment as a Rainbow Option

Terminal Value, Accounting Numbers, and Inflation

94
104

Peter A. Brous, Seattle University

Gunther Friedl, Technische Universitt Mnchen, and
Bernhard Schwetzler, HHL Leipzig Graduate School
of Management

Comment on Terminal Value, Accounting Numbers, and

Inflation by Gunther Friedl and Bernhard Schwetzler

113

Michael H. Bradley, Duke University, and

Gregg A. Jarrell, University of Rochester

Valuing an Early-Stage Biotechnology Investment

as a Rainbow Option
by Peter A. Brous, Seattle University

n an article published in this journal in 2003,

Richard Shockley and three of his students
presented a detailed valuation of an early-stage
biotechnology investment using a binomial lattice
option pricing model.1 The article does a commendable job of
showing how investments with multiple stages can be treated
as compound sequential optionsthat is, a series of options
in which investment in one option provides the opportunity
to invest in the next in the series. The sequence continues
until the final decision to spend resources to produce and
distribute the product.
Any kind of investment opportunity that requires the
success of several costly developmental phases to create and
bring a product to market can be viewed as a sequential
compound option. Although such investments are especially
common in biotechnology, pharmaceuticals, technology, and
natural resources exploration, they are typically involved in
the development of most kinds of new products.
One of the simplifying assumptions in the analysis
presented by Shockley et al. is the possibility of capturing all
the risks associated with this type of investment in a single
measure, the standard deviation of the returns on the underlying asset. This assumption is standard when valuing either
financial or real options using either the Black-Scholes
Option Pricing Model or binomial option pricing models. But
as I argue in this article, this assumption becomes problematic
when analyzing an investment that faces idiosyncratic (or
firm-specific) risks such as uncertainty about the resolution of
some technology issue, FDA approval for a drug candidate, or
discovery of a natural resource. Such risks are also referred to
as unsystematic risks, and one of the most notable characteristics of such risks is their tendency to decrease over time until
they are eventually resolved. In this sense, the profile of such
risks is very different from that of systematic risksrisks
that are associated with uncertainty about the product market
and the broad economywhere the potential for extreme
outcomes typically increases with longer periods of time.
Because of this fundamental difference, the widespread
use of a single measurethe standard deviation of asset
returnsto capture both product market and technology

risks raises doubt about the reliability of the model. As Tom

Copeland and Vladimir Antikarov argue in their Practitioners Guide to Real Options,

1. R. Shockley, S. Curtis, J. Jafari, and K. Tibbs, The Option Value of an Early-Stage

Biotechnology Investment, Journal of Applied Corporate Finance, Vol. 15, No. 2 (Winter 2003), pp. 44-55.

2. T. Copeland and V. Antikorov, Real Options: A Practitioners Guide (New York, NY:
Texere 2001), page 270.

BI

94

Journal of Applied Corporate Finance Volume 23 Number 2

We need to build an event tree that reflects the actual resolution of the uncertainty over time so that we can get optimal
execution of the available real options and correct ROA valuation. The way to do this is to keep the major uncertainties separate
and to model the interaction and effect on the projects value
explicitly.2
In the pages that follow, we use the same business case
analyzed by Shockley et al. (hereafter referred to as Shockley)
to demonstrate how to value this early-stage biotechnology
investment by separately modeling the two types of risks:
technology and product market. An option that has two
distinct kinds of risk that develop differently over time is
known as a rainbow option. The key adjustment to the
option pricing model required to value such an option is that,
in place of the standard binomial option pricing model with
two outcomes at each point in time, we employ a quadranomial option pricing model with four outcomes at each point
in time. Instead of the outcomes being either an uptick or a
downtick in the product market, we now have four possibilities: an upturn in the economy combined with a technology
success; a downturn with technology success; an upturn with
technology failure, and a downturn with technology failure.
Whats more, by distinguishing technology risks from product
market risks and allowing them to develop differently over
time, our analysis leads to a different decision about the initial
investment than the one produced by Shockleys model.
Description of the Investment Opportunity
Wahoo Genomics (not its real name), a privately held biotechnology firm, has developed a coat protein that will be used
in treating viral outbreaks in farm livestock. The decision at
hand, in April of 2002, is whether the firm should spend $2
million to start the preclinical trials. The series of sequential investments necessary to get this product to market are
depicted in Figure 1.

A Morgan Stanley Publication Spring 2011

Figure 1 Timeline of the Various Test Phases

Apr-02

Oct-02

Apr-03

Oct-03

Apr-04

Oct-04

Apr-05

Oct-05

Apr-06

Oct-06

Apr-07

Preclinical Testing
$2 million
18 months
5%

INADA
$1 million
6 months
75%

Field Trials
$5 million
24 months
25%

NADA
$0.5 million
12 months
75%

Market
$10 million

This investment opportunity is a classic example of a

sequential compound option. Initially, the firm must decide
whether to spend $2 million on preclinical testing that will
take 18 months to complete, with an assumed probability of
success of 5%. If successful, management has the option to
spend $1 million to complete the process for an Investigational
New Animal Drug Application (INADA). This process will
take six months to complete, and has a 75% chance of being
approved for the INADA. If the INADA is granted, the firm
has the option to spend $5 million to complete field trials over
the next 24 monthsand it is assumed there is a 25% chance
that the data collected from the field trials will allow them to
apply for a New Animal Drug Application (NADA). Given the
success of the field trials, the firm has the option of applying for
the NADA at a cost of $0.5 million. The application process
will take one year to receive approval with an assumed success
rate of 75%. As noted by Shockley, there is only a 0.7% (.05 x
.75 x .25 x.75) probability that the firm will be successful in all
four phases, creating the opportunity to spend $10 million to
exercise the final option to take this product to market.
Shockleys figures 2 and 3 (not presented here) lay out the
assumptions and cash flow estimations of the product, conditional upon the product being taken to market. The projected
after-tax cash flows start at $0.87 million in year 1 (ending
in April 2008) and grow to $14.71 million in year 10 (April
2017). Assuming a -20% long-term growth rate for the cash
Journal of Applied Corporate Finance Volume 23 Number 2

flows after 2017 and a risk-adjusted discount rate of 11%/year,

the current value (April 2002) of the future cash inflows generated by taking this new protein to market is $29.77 million. To
estimate the static NPV of this investment, one must consider
the five possible outcomes, the cumulative probabilities of
each outcome occurring, and their respective NPVs should the
outcomes occur. Figure 2 (of this paper) provides the inputs
used to estimate the static expected NPV of this investment.
The cumulative probability for the four failed outcomes
equals the probability of success for previous hurdles times the
probability of the outcome failing. For example, the cumulative
probability that the field trials will end in failure will require the
Preclinical Testing and the INADA to be successful before the
field trials failan outcome that has a 2.8% probability (.05 x
.75 x .75). The NPV of each outcome associated with a failure
is simply the present value of all of the costs incurred to get
to that point of failure. For example, if the field trials fail, the
company would have spent $2 million in April of 2002 for the
preclinical trials, $1 million in October of 2003 for INADA,
and $5 million in April of 2004 for the field trials, yielding an
expected NPV of -$6.91 million (NPV = -$2 - $1/(1.111.5) $5/(1.112)). Finally, the sum of the expected NPV is a negative
$2.03 million, which is the so-called static expected NPV of
taking on this investment opportunity.
Based on this number, Wahoo Geonomics would not
make the initial investment to conduct the preclinical
A Morgan Stanley Publication Spring 2011

95

Figure 2 Static NPV Analysis: April 2002

Cumulative
Probability

Probability

of Success

of Outcome

NPV

NPV

5.0%

95.00%

-$2.00

-$1.90

Failure of INADA

75.0%

1.25%

-$2.86

-$0.04

Failure of Field Trials

25.0%

2.81%

-$6.91

-$0.19

Failure of NADA

75.0%

0.23%

-$7.24

-$0.02

0.70%

$16.57

Outcome
Failure of Preclinical Testing

Taking the Product to Market

Sum

Expected

$0.12

100.00%

-$2.03

Figure 3 Value Tree for Underlying Asset

Apr-02

Oct-02

Apr-03

Oct-03

Apr-04

Oct-04

Apr-05

Oct-05

Apr-06

Oct-06

Apr-07
35052.05

17283.07
8521.74
4201.80
2071.78
1021.53
503.68
248.35
122.45
60.38
29.77

29.77
14.68

503.68

122.45
60.38

7.24

122.45

29.77

7.24

29.77

7.24

1.76

29.77
14.68

7.24
3.57

1.76
0.87

122.45
60.38

14.68

3.57

503.68
248.35

60.38

14.68

3.57

503.68

122.45

29.77

2071.78
1021.53

248.35

60.38

14.68

2071.78
1021.53

248.35

8521.74
4201.80

7.24
3.57

1.76
0.87

0.43

1.76
0.87

0.43
0.21

0.43
0.21

0.10

0.10
0.05
0.03

trials. But what this calculation ignores is the reality that

management does not have to decide on making this series
of investments today, but can wait and incorporate any new
information into their investment decisions at the point in
time when it becomes necessary. A revised assessment can
be made based not only on whether they succeeded in the
previous test phase, but also on their perception of the value
of taking the product to market at that point in time.
Valuation Approach Applying a Binomial Option
Pricing Model
Because the benefits of waiting and learning are not reflected
in the static expected NPV, Shockley suggests valuing this
96

Journal of Applied Corporate Finance Volume 23 Number 2

early-stage biotechnology investment as a compound sequential option. To value this phased investment opportunity
using a binomial option pricing model, a value tree for the
underlying asset must be developed. Shown in Figure 3, the
value tree for the underlying asset models the potential movement in the value of the underlying asset over time and is
generated based on the present value of the expected future
cash inflows of $29.77 million and an assumed volatility of
the underlying assets returns of 100% per year.
Given the development of the value tree for the underlying asset, the value tree for the sequential compound option is
developed based on the value of the underlying asset and the
various exercise prices (as shown in Figure 4). More specifiA Morgan Stanley Publication Spring 2011

Figure 4 Value Tree for Sequential Compound Option

Preclinical Testing
Apr-02

Oct-02

INADA
Apr-03

Oct-03

Field Trials
Apr-04

Oct-04

Apr-05

NADA
Oct-05

Apr-06

Oct-06

Apr-07
35042.05

17273.31
8512.22
4192.20
2061.76

2062.30
1012.25
489.80
234.41
48.12

19.81

5.41
1.83

51.32

0.00
0.00

8.18

0.00

0.59
0.20

0.07

19.77
6.68

1.76

3.14
1.19

112.45
50.62

20.88

22.56
9.63

493.68
238.59

112.44

113.37

18.96

19.72
7.83

238.58

52.71

2061.78
1011.77

493.67

494.15

109.52

109.92
49.73

1011.75

239.34

8511.74
4192.05

0.00
0.00

0.00
0.00

0.00

0.00
0.00

0.00
0.00

0.00
0.00

0.00

0.00
0.00
0.00

Result: Spend $2.0 million on the preclinical testing

Bold numbers suggest nodes in which options are excercised.

cally, the option value tree reflects the value of the option at
each node based on maximization of the value if the option
is exercised and the value if the option is not exercised.
The option value tree is generated by starting at the end
nodes (April 2007) and determining the larger of two values:
the value of exercising the option and taking the product to
market (option value = value of underlying asset minus exercise
price), or of not exercising and letting the option expire (option
value = 0). If the value of the cash inflows from taking the
product to market is greater than the cost of taking the product
to market ($10 million), the firm should exercise the option
and take the product to market. As presented in Figure 4, the
option to take the product to market will be exercised at the
top six end nodes but not at the bottom five nodes.
The specific equation used to value the sequential
compound option in prior nodes depends on whether the
node occurs at a point in time when a test phase is expected
to be resolved. If the node does occur when a test phase is
expected to be resolved and if the phase has been completed
successfully, the decision needs to be made whether to
exercise the option to invest in the next test phase. The
option value is determined by the maximum value of either
exercising the option to invest in the next test phase (option
Journal of Applied Corporate Finance Volume 23 Number 2

value = value of the option to invest minus exercise price),

or not exercising the option (option value = 0). To value the
option to invest at that point in time, we must calculate
the certainty equivalent value of the two uncertain future
option values.
For example, to estimate the option value at the top node
at the end of April 2006, we examine the value if the firm
exercises its option to spend $500,000 to submit a NADA and
the value if the firm does not exercise this option. If the firm
does exercise this option, they obtain either the value of this
rainbow option next period if there is an uptick ($17.3 billion)
or the value if there is a downtick ($4.2 billion). The certainty
equivalent value of these two future uncertain option values
in April of 2006 is $8.5 billion, and is calculated using the
following equation:
CEQ option value = {p*(Cu) + (1-p)*Cd} / (1 + Rf) (1)
CEQ Option Value = {.4472*$17,273.31 m +
(1-.4472)*$4,192.05 m} / (1 + .025) = $8,512.22
where;
CEQ is the certainty equivalent;
p is the risk-neutral probability of an uptick (p =
(Rf d) / (u d));
1 p is the risk-neutral probability of a downtick;
A Morgan Stanley Publication Spring 2011

97

Figure 5 Evolution of Risk Profile Technology Product/Project

Product/Market Risk
(Correlated with Market
Beta 0)

Technology Risk
(Independent of Market
Beta = 0)

Product/Market Risk

Research & Development

Phase 1

Phase 2

Phase 3

Rollout

Source: David Dufendach (2000).

Cu is the value of the call option if the value of the

underlying asset moves upward;
Cd is the value of the call option if the value of the
underlying asset moves downward;
Rf is the six month yield on a five-year Treasury Bond
in April 2002;
u is exponential raised to the power of the standard
deviation times square root of 1 divided by the number of
periods in a year (u = esd(sqrt(1/t)));
d is the inverse of u (d = 1/u).
In this case, the value of the option at the top node at the end
of April 2006 is maximized if the firm exercises the option to
spend $500,000 for the NADA:
CApril 2006 top node = Max [$8,512.22 m - $0.50 m, 0]
= $8.51 billion
At earlier nodes where the resolution of a specific test
phase does not occurfor example the nodes at the end of
October 2006the value of the option is simply equal to the
certainty equivalent value of the two uncertain option values
found at the subsequent nodes (calculated using equation 1).
The results of this model suggest that if the firm invests the
$2.0 million for preclinical testing, they would be acquiring
an option to invest worth $21.8 million (NPV with flexibility
= $19.81 million), and therefore the firm should invest.
Although this appears to be an excellent example of
the valuation of a sequential compound option using a
binomial model, the net result, to invest, seems unintuitive
98

Journal of Applied Corporate Finance Volume 23 Number 2

given the characteristics of the investment. The output of

the model suggests that the firm should invest $2 million,
with expected additional investments required if things go
well, with the aim of benefiting from a new product that is
expected to create future cash flows valued at $29.8 million,
even when there is only a 0.7% probability that the product will
be taken to market. In my opinion, most practitioners would
choose not to invest in this project, given the low probability
of taking the product to market and the expected value of
taking the product to market. And in the pages that follow,
I will argue that the models decision to invest is the result
of using an inappropriate method of incorporating risk into
the model.
In the binomial option pricing model presented by
Shockley, all of the risks associated with the value of the
underlying asset (both product market and technology) are
incorporated by applying the standard deviation of asset
returns, which is assumed to be constant over the life of
the investment. This assumption is applied in the creation
of both the value tree of the underlying asset (through the
size of the upticks and downticks in value) and the value
tree for the option (through the risk-neutral probabilities).
Wahoo Genomics investment opportunity, like most
phased investments, is subject to both standard product
market riskssuch as uncertain supply and demand for the
product under considerationand technology risks, notably
uncertainty about the firms ability to pass a series of tests
A Morgan Stanley Publication Spring 2011

Figure 6 Value Tree for Underlying Asset (SD = 50%/year)

Apr-02

Oct-02

Apr-03

Oct-03

Apr-04

Oct-04

Apr-05

Oct-05

Apr-06

Oct-06

Apr-07
1021.53

717.31
503.68
353.68
248.35
174.39
122.45
85.99
60.38
42.40
29.77

29.77
20.90

122.45

60.38
42.40

14.68

60.38

29.77

14.68

29.77

14.68

7.24

29.77
20.90

14.68
10.31

7.24
5.08

60.38
42.40

20.90

10.31

122.45
85.99

42.40

20.90

10.31

122.45

60.38

29.77

248.35
174.39

85.99

42.40

20.90

248.35
174.39

85.99

503.68
353.68

14.68
10.31

7.24
5.08

3.57

7.24
5.08

3.57
2.51

3.57
2.51

1.76

1.76
1.24
0.87

in the development process that are necessary to allow them

to bring the product to market.
As stated earlier, one important difference between these
two types of risk is that product market risks are systematic,
with the main source of uncertainty increasing over time (that
is to say, whatever the strength of product markets and the
general economy today, it is likely to be different in the future,
in some cases better, in some worse). By contrast, technology
risks, as illustrated in Figure 5, get resolved over time.
As a consequence, the use of a single constant measure of
uncertainty to reflect both of these types of risks is problematic. The purpose of this paper is to present an alternative
quadranomial option pricing model. This model incorporates
product market risks and technology risks separately, thereby
allowing them to develop differently over time.
Valuation Approach Applying a Quadranomial Option
Pricing Model
To apply this quadranomial model, we need a revised measure
of the standard deviation of the underlying assets returns
that reflects only product market risks and not technology
risks. Given the lack of correlation between product market
risks and technology risks, it is safe to assume that the revised
measure of the standard deviation should be significantly
lower than the 100% assumed when considering both types
of risk.
Journal of Applied Corporate Finance Volume 23 Number 2

Figure 6 presents a revised value tree for the underlying

asset that assumes the standard deviation of the underlying
assets returns is 50%/year because this measure of uncertainty now reflects only the product market risks. And
applying a lower standard deviation for the underlying asset
yields less extreme potential future values of the underlying
asset. The minimum and maximum values with a standard
deviation of 50% are $0.87 million and $1.02 billion, respectively, as compared to minimum and maximum values of
$0.03 million and $35.05 billion when assuming a standard
deviation of 100%.
A second, more subtle change in the value tree for the
underlying asset applies to nodes that represent a point in
time when a test phase is completed. In our approach, each
node yields four possible outcomes instead of two as under
the Shockley approach. The value of the underlying asset is
affected not only by an uptick or downtick in the product
market, but also by its success in dealing with technology
risk. These nodes expand to include four possibilities for the
value of the underlying asset: two cases where the technology
succeeds, one in combination with an uptick in the product
market and one with a downtick; and two cases representing
the situation in which the technology fails and the product
market either improves or worsens. Because the two nodes
in which the technology fails result in asset values of zero,
these nodes can be left off the value tree for the underlying
A Morgan Stanley Publication Spring 2011

99

Figure 7 Value Tree for Sequential Compound Rainbow Option

Preclinical Testing
Apr-02

Oct-02

INADA
Apr-03

Oct-03

Field Trials
Apr-04

Oct-04

Apr-05

NADA
Oct-05

Apr-06

Oct-06

Apr-07
1011.53

530.66
370.12
64.45
178.62

44.75
30.92
16.23
6.17
0.51

0.00

0.01
0.00

6.09

0.00
0.00

2.14

0.00

0.46

4.68
1.53

0.17

0.21
0.10

19.77
8.36

3.97

1.18
0.63

50.38
24.48

14.69

3.81
2.30

112.45
57.17

37.65

9.50

0.00
0.00

0.00

14.26

6.20

238.35
123.47

84.20

21.14

4.60

0.15
0.07

30.83

14.35

493.68
257.94

0.02
0.01

0.00
0.00

0.00
0.00

0.00
0.00

0.00

0.00
0.00

Result: Do not spend $2.0 million on the preclinical testing

0.00

Bold nodes are situations when the firm would exercise their option

assetbut they must be accounted for when determining

the option value tree that is based on the value tree of the
underlying asset.
Given the revised value tree for the underlying asset, we
can then generate a revised option value tree. The value of the
option tree at the end nodes are calculated using the same
equation as presented earlier when discussing the Shockley approach. The options value is the greater of the value
of exercising the option and taking the product to market
(option value = value of underlying asset minus exercise price)
and the value associated with letting the option expire (option
value = 0). If the value of the cash inflows from taking the
product to market is greater than the cost of so doing, the
decision rule is to exercise the option and take the product to
market. As presented in Figure 7, the option to take product
to market in April of 2007 will be exercised at the top seven
end nodes, but not in the bottom four nodes. (And keep in
mind that there is another set of end nodes, not shown in
the figure, that reflect the failure of the NADA for which the
option value is zero.)
To calculate the value of the option at the earlier nodes,
one must determine whether or not a test phase is expected
to be completed at that point, and understand how to adjust
100

Journal of Applied Corporate Finance Volume 23 Number 2

the valuation equation based on the type of uncertainty being

resolved over the next period (which determines the size of
the discount for risk). For those nodes that do not precede the
ending of a test phase (including April 2002, October 2002,
April 2004, October 2004, April 2005, and April 2006), only
product market uncertainty will be resolved over the coming
six-month period, and so one should apply a discount that
reflects only the product market risk. In such cases, the value
of the option can be calculated as the certainty equivalent
value of the future uncertain option values using binomial
risk-neutral probabilities based solely on the product risk. In
equation form,
CEQ option value = {p*(Cu) + (1-p)*Cd} / (1 + Rf) (2)
where;
CEQ is the certainty equivalent;
p is the binomial risk-neutral probability reflecting an
uptick in the product market (p = (Rf d) / (u d));
1-p is the binomial risk-neutral probability reflecting a
downtick in the product market;
Cu is the value of the call option if the value of the
underlying asset moves upward;
Cd is the value of the call option if the value of the
underlying asset moves downward;
A Morgan Stanley Publication Spring 2011

Figure 8 When to Apply Binomial or Quadranomial Risk Neutral Probabilities

Date

Technology Resolution

Risk Neutral Probability

Apr-02

Binomial

Oct-02

Binomial

Apr-03

Quadranomial based on Pre-Clinical Trial probabilities

Oct-03

Pre-Clinical Trial

Quadranomial based on INADA probabilities

Apr-04

INADA

Binomial

Oct-04

Binomial

Apr-05

Binomial

Oct-05
Apr-06

Quadranomial based on Field Trial probabilities

Field Trial

Oct-06
Apr-07

Binomial
Quadranomial based on NADA probabilities

NADA

Rf is the six-month yield on a five-year Treasury Bond

in April 2002:
u is exponential raised to the power of the standard
deviation times square root of 1 divided by the number of
periods in a year (u = esd(sqrt(1/t)));
d is the inverse of u (d = 1/u).
For nodes that do precede the ending of a test phase (April
2003, October 2003, October 2005, and October 2006), the
project is subjected to both product market and technology
risk over the next six-month period, and therefore a discount
reflecting both of these risks is required. This discount is determined by applying quadranomial risk-neutral probabilities
that reflect the specific test phase that is being completed.
CEQ option value = {p1*(Cu) + (p2)*Cd + p3*0 +
p4*0} / (1 + Rf)
(3)
where;
CEQ is the certainty equivalent;
p1 is the quadranomial risk-neutral probability reflecting an uptick in the product market and technology success
(p1 = (Rf d) / (u d) * prob of tech success);
p2 is the quadranomial risk-neutral probability reflecting a downtick in the product market and technology
success (p2 = (1- p1) * prob of tech success);
p3 is the quadranomial risk-neutral probability reflecting an uptick in the product market and technology failure
(p3 = (Rf d) / (u d) * prob of tech failure);
p4 is the quadranomial risk-neutral probability reflecting a downtick in the product market and technology
failure (p4 = (1 p3) * prob of tech failure);
Cu is the value of the call option if the value of the
underlying asset moves upward;
Cd is the value of the call option if the value of the
underlying asset moves downward;
Rf is the six-month yield on a five-year Treasury Bond
in April 2002;
Journal of Applied Corporate Finance Volume 23 Number 2

N/A

u is exponential raised to the power of the standard

deviation times square root of 1 divided by the number of
periods in a year (u = esd(sqrt(1/t)));
d is the inverse of u (d = 1/u).
Figure 8 provides a guideline for determining whether a
binomial or a specific set of quadranomial risk-neutral probabilities are appropriate for each node in the future based on
the upcoming periods risk exposure. Note that these probabilities will vary depending on the particular test phase to be
completed and its associated probability of success. Figure 9
provides the various risk-neutral probabilities based on the
particular phase being completed.
For example, to calculate the value of the option at the
top node on October 2006, a maximization equation is not
necessary because a test phase is not being completed at this
point in time, and so there is no decision as to whether or
not to exercise an option. But because the project is subject
to both product market and technology risks over the next
six months, there will be four possible outcomes at the end of
this six-month period. The possible outcomes are as follows:
NADA is successful and there is an uptick in the product
market ($1.01 billion); NADA is successful and there is a
downtick in the product market ($494 million); NADA is
unsuccessful and there is an uptick in the product market
($0), and NADA is unsuccessful and there is a downtick in
the product market ($0). Therefore, to calculate the certainty
equivalent of these four potential future outcomes, we use
quadranomial risk-neutral probabilities.
The certainty equivalent option value for the top node at
the end of October of 2006 equals:
CEQ option value = {.3354*$1,011.53 m +
.4146*$493.68 + .1118*0 + .1382*0} / (1 + .025)
= $530.66 m
The value of the option at the top node in April of 2006
would be calculated using binomial risk- neutral probabilities
A Morgan Stanley Publication Spring 2011

101

Figure 9 Various Quadranomial Risk Neutral Probabilities

Prob of
Success

Tech Success
& Uptick
(p1)

Preclinical Testing

0.05

0.0224

0.0276

0.4248

0.5252

INADA

0.75

0.3354

0.4146

0.1118

0.1382

Field Trials

0.25

0.1118

0.1382

0.3354

0.4146

NADA

0.75

0.3354

0.4146

0.1118

0.1382

Tech Success
& Downtick
(p2)

Tech Failure
& Uptick
(p3)

Tech Failure
& Downtick
(p4)

Quadranomial Probabilities are estimated using the following equations:

p1 = prob of success * RNP of uptick (0.4472)
p2 = prob of success * RNP of downtick (0.5528)
p3 = (1-prob of success) * RNP of uptick (0.4472)
p4 = (1-prob of success) * RNP of downtick (0.5528)

Figure 10 Sensitivity Analysis Based on

Standard Deviation of the Underlying Asset
Standard
Deviation

Value of

the Option

Investment

Decision

50%

$33,417

Do Not Invest

60%

45,947

Do Not Invest

70%

57,893

Do Not Invest

80%

$69,356

Do Not Invest

90%

$80,413

Do Not Invest

100%

$90,885

Do Not Invest

that reflect only the risk associated with the product market.
And a maximization equation is needed to determine if the
firm should exercise the option to spend $500,000 for the
NADA:
CApril 2006 top node = Max[.4472*$530.66 m +
(1-.4472)*$257.94 m - $0.50 m, 0] = $370.12 m
This process continues until the value on April 2002 is
calculated based on the following equation:
CApril 2002 = Max[.4472*$0.07 m + (1-.4472)*$0.0 m - $2.0
m, $0 m] = $0.0 m
This result suggests that Wahoo Genomics should not invest
in the preclinical testing because the option they would
acquire would be worth $33,417 (.4472*$0.07 m), which is
less than the cost of purchasing the option ($2 million). This
result is inconsistent with Shockleys results using the traditional binomial option pricing model, which suggested that
the value of the option was worth $21.8 million, and that the
firm should invest in starting the preclinical testing.
102

Journal of Applied Corporate Finance Volume 23 Number 2

Sensitivity Analysis
In my valuation of this early-stage biotechnology investment,
I assumed that the 100% standard deviation used in the
Shockley analysis should be reduced to 50% because the volatility measure in my revised model considers only the product
market risk and not the technology risks. When applying
the lower volatility measure, the revised model determined
that the value of the investment decreased to such an extent
that the firm should not take the initial step of developing
this new drug, the opposite recommendation provided by
Shockleys analysis.
But since the assumed level of volatility of 50% was
arbitrarily determined, it would be interesting to demonstrate
the effects of alternative assumptions regarding the volatility
of the underlying asset. Figure 10 presents the values of the
investment opportunity for various assumptions about the
volatility of the underlying asset when applying the revised
model presented in this paper. As the results show, the net
result of the modelthat the firm should not start this project
by investing in the pre-clinical trialsis not affected by the
assumption about the volatility of the underlying asset.
Conclusion
This article presents an alternative approach to one proposed
in 2003 by Richard Shockley and three of his students for
valuing an early stage biotechnology investment. Shockley et
al. presented an approach that uses a binomial option pricing
model in which all the risks or uncertainty associated with
investment is captured in the assumed standard deviation
of the underlying assets returns. The output of Shockleys
model suggested that the firm should invest $2 million to
start preclinical testing for a product with only a 0.7% probA Morgan Stanley Publication Spring 2011

ability of successfully completing all of the necessary test

phases required to bring the product to market. The expected
cost of taking the product to market is $10.0 million, with the
expectation that the value of the cash flows derived from this
product would be $29.8 million. The results from the alternative method presented in this paper suggest that the firm
should not invest the $2.0 million because the expected value
of the option they would be purchasing is worth considerably
less than $2.0 million.
The key difference between the two approaches is that
the approach taken in these pages takes the two main sources
of riskuncertainty about product market and uncertainty
associated with passing test phasesand models them
separately. Whereas the Shockley approach assumes that the
combined risk or uncertainty remains constant over the life
of the project, this alternative approach incorporates measures
of risk that vary over time based on the probability of successfully completing a test phase. Because the probabilities of
successfully completing the various sequential test phases
differ, a constant measure of risk does not seem appropriate.

Journal of Applied Corporate Finance Volume 23 Number 2

The alternative quadranomial option pricing model

presented in this article is well suited to the valuation of
options on underlying assets that have two major types of
riskalso known as rainbow optionsof which the early
stage biotechnology investment is a prime example. But
investments with this combination of product market and
technology risks can be found in almost all investments that
require major successful research and development efforts
before the firm has the ability to take the product to market.
The message from this paper is that investments in new
technology, natural resource exploration, and almost all new
product development investments are subject to both product
market and technology risks, and that such risks should be
modeled separately when determining the value of the investment opportunity.
peter a. brous is the Dr. Khalil Dibee Endowed Chair in Finance at
the Albers School of Business & Economics Seattle University, and can
be reached at pbrous@seattleu.edu.

A Morgan Stanley Publication Spring 2011

103

Journal of Applied Corporate Finance (ISSN 1078-1196 [print], ISSN

1745-6622 [online]) is published quarterly, on behalf of Morgan Stanley by
Wiley Subscription Services, Inc., a Wiley Company, 111 River St., Hoboken,
NJ 07030-5774. Postmaster: Send all address changes to JOURNAL OF
APPLIED CORPORATE FINANCE Journal Customer Services, John Wiley &
Sons Inc., 350 Main St., Malden, MA 02148-5020.
Information for Subscribers Journal of Applied Corporate Finance is published in four issues per year. Institutional subscription prices for 2011 are:
Print & Online: US$441 (US), US$529 (Rest of World), 343 (Europe),
271 (UK). Commercial subscription prices for 2010 are: Print & Online:
US$590 (US), US$703 (Rest of World), 455 (Europe), 359 (UK).
Individual subscription prices for 2010 are: Print & Online: US$105 (US),
59 (Rest of World), 88 (Europe), 59 (UK). Student subscription prices for 2011 are: Print & Online: US$37 (US), 21 (Rest of World), 32
(Europe), 21 (UK).
Prices are exclusive of tax. Australian GST, Canadian GST and European
VAT will be applied at the appropriate rates. For more information on current tax rates, please go to www.wileyonlinelibrary.com/tax-vat. The institutional price includes online access to the current and all online back files to
January 1st 2007, where available. For other pricing options, including
access information and terms and conditions, please visit www.wileyonlinelibrary.com/access
Journal Customer Services: For ordering information, claims and any inquiry
concerning your journal subscription please go to www.wileycustomerhelp.
com/ask or contact your nearest office.
Americas: Email: cs-journals@wiley.com; Tel: +1 781 388 8598 or
+1 800 835 6770 (toll free in the USA & Canada).
Europe, Middle East and Africa: Email: cs-journals@wiley.com;
Tel: +44 (0) 1865 778315.
Asia Pacific: Email: cs-journals@wiley.com; Tel: +65 6511 8000.
Japan: For Japanese speaking support, Email: cs-japan@wiley.com;
Tel: +65 6511 8010 or Tel (toll-free): 005 316 50 480.
Visit our Online Customer Get-Help available in 6 languages at
www.wileycustomerhelp.com
Production Editor: Joshua Gannon (email:jacf@wiley.com).
Delivery Terms and Legal Title Where the subscription price includes print
issues and delivery is to the recipients address, delivery terms are Delivered
Duty Unpaid (DDU); the recipient is responsible for paying any import duty or
taxes. Title to all issues transfers FOB our shipping point, freight prepaid. We
will endeavour to fulfil claims for missing or damaged copies within six months
of publication, within our reasonable discretion and subject to availability.
Back Issues Single issues from current and recent volumes are available at
the current single issue price from cs-journals@wiley.com. Earlier issues may
be obtained from Periodicals Service Company, 11 Main Street, Germantown, NY 12526, USA. Tel: +1 518 537 4700, Fax: +1 518 537 5899,
Email: psc@periodicals.com

This journal is available online at Wiley Online Library. Visit www.wileyonlinelibrary.com to search the articles and register for table of contents e-mail
alerts.
Access to this journal is available free online within institutions in the developing world through the AGORA initiative with the FAO, the HINARI initiative
with the WHO and the OARE initiative with UNEP. For information, visit
www.aginternetwork.org, www.healthinternetwork.org, www.healthinternetwork.org, www.oarescience.org, www.oarescience.org
Wileys Corporate Citizenship initiative seeks to address the environmental,
social, economic, and ethical challenges faced in our business and which are
important to our diverse stakeholder groups. We have made a long-term commitment to standardize and improve our efforts around the world to reduce
our carbon footprint. Follow our progress at www.wiley.com/go/citizenship
Abstracting and Indexing Services
The Journal is indexed by Accounting and Tax Index, Emerald Management
Reviews (Online Edition), Environmental Science and Pollution Management,
Risk Abstracts (Online Edition), and Banking Information Index.
Disclaimer The Publisher, Morgan Stanley, its affiliates, and the Editor
cannot be held responsible for errors or any consequences arising from
the use of information contained in this journal. The views and opinions
expressed in this journal do not necessarily represent those of the
Publisher, Morgan Stanley, its affiliates, and Editor, neither does the publication of advertisements constitute any endorsement by the Publisher,
Morgan Stanley, its affiliates, and Editor of the products advertised. No person
should purchase or sell any security or asset in reliance on any information in
this journal.
Morgan Stanley is a full-service financial services company active in
the securities, investment management, and credit services businesses.
Morgan Stanley may have and may seek to have business relationships with
any person or company named in this journal.
Copyright 2011 Morgan Stanley. All rights reserved. No part of this publication may be reproduced, stored or transmitted in any form or by any means
without the prior permission in writing from the copyright holder. Authorization to photocopy items for internal and personal use is granted by the copyright holder for libraries and other users registered with their local Reproduction Rights Organization (RRO), e.g. Copyright Clearance Center (CCC), 222
Rosewood Drive, Danvers, MA 01923, USA (www.copyright.com), provided
the appropriate fee is paid directly to the RRO. This consent does not extend
to other kinds of copying such as copying for general distribution, for advertising or promotional purposes, for creating new collective works or for resale.
Special requests should be addressed to: permissionsuk@wiley.com.
This journal is printed on acid-free paper.