5th Annual Real Options Conference
2001 Los Angeles
Anderson School of Management, UCLA
Friday: Software Development and
Network Expansion Options | Valuing Oil Development
Investments | Keynote Address: Eduardo Schwartz | Valuing Natural Resource and Transportation Investments
| Competition and Strategic Investments | Practitioner Panel Discussion
Saturday: Switching Options | Information,
Incomplete Information, and Non-Traded Assets | Private
Information and Incentives | Modelling
and Numerical Analysis | Academic Panel Discussion
FRIDAY
Software Development
and Network Expansion Options
Chairperson: Blake Johnson (Stanford U.)
Application Service Provider (ASP) is a
recently emerged software delivery model under which an ASP hosts, manages
and delivers software as a service to customers via the Internet or a private
network. The ASP model offers benefits from cost savings, specialized
expertise, a faster time to market, and a reduced risk due to a lower capital
investment. Buying services also provides more financial and technological
flexibility than owning the technology in-house. However, customers who are
unsure about the value of ASP services and their demands, in terms of the
number of users and a usage level, may be reluctant to commit to ASP
contracts. Many customers are also concerned with security and loss of
control and performance, especially when the software becomes more critical
as the company grows. Thus, for the ASP industry to move forward, ASPs must
help customers cope with these uncertainties and risks. In this paper we
propose three real options based approaches. First, in order to deal with the
uncertainties of software value and usage level, we propose a usage-based
pricing structure with an option to switch to a flat subscription fee. This
arrangement allows ASPs to penetrate the low-usage market and to link their
revenue to the value customers receive from their software, while still
providing an upper bound on customers™ cost. Second, we evaluate an option to bring the software in-house after an initial period
of ASP-based access. In addition to allowing customers to manage the risk
created by uncertainty in their usage level, the number of users and the
value per usage that the software will provide, this option allows customers
to hedge against assuming an unacceptable level of security and performance
risk. Implicitly, the option to bring software in-house is a growth option
for a company to invest fully in the software should it become desirable to
do so. Lastly, we analyze an option to end an ASP contract prior to its
expiration, including when customers are under a minimum usage requirement
and a minimum time commitment. This exit option allows customers to manage
their software investments when they are unsure of the value of the software.
It also alleviates the risks of relying on certain technologies and service
providers, which may be especially valuable in a fast-changing technology
environment, and when dealing with early-stage ASPs. As a result, this option
illustrates the value of technological flexibility that ASP-based software
delivery offers relative to an in-house implementation. In order to study
these real options, we identify the three key underlying uncertainties to be
software value per usage, usage level and the number of users. We then employ
a mean reverting process with a time-varying mean and a time varying variance
to model these uncertainties and their correlations. This method allows us to
pattern after a software cycle, a learning effect, a company growth and the
correlations between software benefit and usage. Finally, we utilize a
Monte-Carlo simulation approach to approximate the option values and the
exercise thresholds of these real options.
Software development based on Commercial
Off-the-Shelf products is subject to multiple sources of uncertainty. One
potential source of uncertainty is the license costs of the COTS product used
in the system. The management of such uncertainty requires strategies that
effectively mitigate the underlying risk with minimal impact on the economic
value generated. This paper presents such a strategy, and shows how it can be
evaluated using a state-of-the-art financial valuation
technique, namely, real options analysis.
This paper provides a strong conceptual
framework to analyze investment decisions for network expansion problems.
Specifically, we develop an algorithm to find the optimal time to open a new
segment in an already existing network. Segment demand is uncertaint
and capital investments are high; hence, the investment decision is
non-trivial.
Due to the network environment, the decision to open a new segment cannot be
analyzed as an indipendent one: the network
externalities that arise both in the price and cost functions influence the
project value and the optimal investment policies. Furthermore, the inclusion
of a new segment also increases the value of the network as a whole, which
augments the benefits of expansion. Finally, future growth should also be
taken into account in the expansion model; that is, the option to open
further segments branching out from the current one not only adds strategic
value to the segment itself, but will in general also generate new network
effects.
The model starts by quantifying the effects that network structures have on
segment economics, including both price and cost. For a given network, a real
options approach is then taken to derive expressions for the optimal time to
add a segment, the option value of the expansion opportunities, and the
sensitivity of these results to the key parameters of the analysis. We show
that, for positive network externalities, an increase in the network size
both raises the option value and lowers the demand level at which it is
optimal to add the segment. Future growth options are then incorporated by
expanding the analysis to a sequential capacity expansion framework with
different underlying stochastic processes for each new segment.
Valuing Oil
Development Investments
Chairperson: Robert McDonald (Northwestern U.)
J. McCormack (SVP Stern Stewart),
D. Calistrate and G. Sick (U. of Calgary), Applying
Real Options to Assessing Proven Undeveloped Petroleum Reserves
The oil industry was among the first of
the large industries both to adopt discounted cash flow methods in valuing
assets and projects. Discounted cash flow (DCF) tools are fundamental to
engineering and financial analysis in the oil industry. They are well
understood by managers and generally provide accurate valuations of developed
hydrocarbon reserves. Unfortunately, DCF techniques systematically undervalue
undeveloped reserves. Moreover, they may encourage premature development of
certain reserves, and may also fail to identify important risk management
opportunities.
Managers in the oil industry have long been aware that the market value of
individual oil properties, not to mention entire E&P companies, is
usually greater than the value of their discounted cash flows. This is
particularly true in cases where there are significant quantities of
undeveloped reserves. For this reason, E&P managers have often been
willing to pay a premium above a DCF value for some undefinable
"upside" associated with undeveloped reserves.
Unfortunately, the analytical discipline usually imposed by the DCF method is
lost when managers value properties or companies on the basis of
rules-of-thumb or simple intuition.
Real option models address these shortcomings. Though more complex than
traditional DCF analysis, real option models provide a far more complete
picture of not only reserve values but also the drivers of that value. Proven
undeveloped reserves (PUDs) lend themselves to real
option analysis because owners of PUDs have the
right, but not the obligation, to develop those reserves in the future, so
that the total value of a PUD includes both a DCF
value plus some additional option or "volatility" value. There is a
sound economic reason to assess undeveloped reserves at more than their DCF
value, and real options
models provide the means to do so.
Consider an undeveloped oilfield with
uncertainty about the size and quality of its reserves. There are some
alternatives to invest in information to reduce the risk and to reveal some
characteristics of the reserve. This paper presents an evolutionary real
options model of optimization under uncertainty with genetic algorithms
and Monte Carlo simulation, to select the best alternative of
investment in information.
There is a legal time to expiration of the option to start the investment for
the oilfield development. The model considers both the technical
uncertainties revealed by the information and the market uncertainty using
two different stochastic processes for the oil prices, which are simulated. Monte Carlo simulations evaluate the decision rule curves generated in the evolutionary process.
The process evolves toward a near optimum solution, giving the real option
value and the optimal decision rule. The evolutionary programming under
uncertainty was performed in C++ environment with good results and a
description of
the programming procedure is provided.
Albino L. d´Almeida,
Ivo F. Lopez and Marco A.G. Dias (Petrobras, Brasil), Oil Drilling Rig Fleet Decisions
The oil drilling activity is complex,
involving onshore and offshore units, for perforation and production, of
great or small size and with extremely varying degrees of technological
complexity. It is a risk business due to work with high costs and
remuneration highly variable in time, function of market conditions (oil
barrel price and availability of equivalent rigs). Also, it has significant
weight in the composition of exploitation and development costs of a field.
The Real Options Theory considers the technical and economic uncertainty, the
flexibility in the managerial decision-making, the irreversibility (total or
partial) of investments and the waiting value, that is, to wait for better
conditions or new information.
This work presents a study for the oil drilling rig owner to decide among
options (operation, temporary suspension, exit from the business) function of
freight charge values (day rate). Operational, acquisition, suspension,
maintenance, reactivation and abandonment costs are considered; and also,
volatility, dividend rates and capital attractiveness rates. Thus, it is
possible to determine the adequate composition of the oil drilling rig fleet,
departing from the determination of day-rate ranges for each type of drilling
rig and to make an optimum decision (abandonment, temporary suspension,
operation, reactivation or reentry/expansion). It is reached the most
adequate rig fleet composition, from the determination of the optimal
switches - the threshold points - among the alternatives (entry, operation,
suspension, exit, reactivation) and the
correspondent day-rate intervals to each rig type. Two sets of complex
partial differential equations are generated to represent the policy options
and general conditions and they are solved by numerical methods. Finally a
sensitivity analysis is performed varying the values of the previous deterministic
variables and showing their effect on both the optimal decision rule and the
value of rig considering the embedded options.
Luncheon Keynote Address
by Eduardo Schwartz (University of
California at Los Angeles)
Rational Pricing of Internet Companies
Professor Schwartz is the California
Professor of Real Estate and Professor of Finance at UCLA's Anderson Graduate
School of Management. Previously he taught at the University of British Columbia and was visiting at the London Business School and the University of California at Berkeley. He received a Masters and a Ph.D. in Finance from
the University of British Columbia.
Professor Schwartz has been President of
the American Finance Association and the Western Finance Association. He has
been associate editor for more than a dozen journals, including the Journal
of Finance, the Journal of Financial Economics and JFQA,
and is a Research Associate of the National Bureau of Economic Research.
Professor Schwartz has made significant
contributions on various dimensions in asset and securities pricing, such as
interest rate models, asset allocation issues, evaluating natural resource investments, pricing Internet companies, and
the stochastic behaviour of commodity prices. Dr.
Schwartz (with Michael Brennan) has been one of the early pioneers in real
options, starting with the classic work on the valuation of a mine, to
commodity claims, R&D and Internet company valuation and other
contributions.
He won a number of awards for teaching excellence and for the quality of his
published work.
Dr. Schwartz has also been a consultant to governmental agencies, banks,
investment banks and industrial corporations.
Valuing Natural
Resource and Transportation Investments
Chairperson: Michael Brennan (UCLA)
Katia Rocha, Ajax R.B. Moreira, Leonardo Carvalho and Eustàquio Reis
(IPEA/DIMAC, Brasil), The Option Value of Forest Concessions in
Amazon Reserves
The Brazilian government is now planning
to implement natural forest concessions for timber extraction. In addition to
the legal requirements imposed on the management of concessions (minimum
reserves, maximum extraction rates, etc.), the value of concessions is
closely linked with uncertainties in estimates of the volume of commercial
logs within the concession area and on future timber prices.
This paper proposes a method to appraise the value of forest concessions
based on the real option theory (ROT). By combining the hypothesis of
uncertainty in the volume of logs in a concession, logs prices modeled as a
mean-reverting stochastic process, and applying inter-temporal maximization
of profits, the method provides a more realistic estimate of the market value
of concessions than does Net Present Value (NPV), which does not take these
uncertainties into account.
Comparison between estimates using NPV and ROT shows that the latter are
systematically higher. For the base case, the concession value using ROT is
153% higher. Since forest concessions are public resources, differences of that
magnitude cannot be neglected. The paper also proposes methods to estimate
the probability distribution of logging volumes in concession areas along
with future prices. The volume distribution is specified in a spatial model
as a function of geographic characteristics of the area as well of the
neighboring areas.
Modern asset pricing (MAP; commonly known
as real options valuation) has been used as an alternative to discounted cash
flow (DCF) methods in the mining industry to improve the representation of
project structure within project valuation models. Previous mining
applications of MAP have tended to treat the ore deposit as a homogenous
entity as opposed to a heterogenous one in which
the deposit can be subdivided into zones differentiated by size, quality and
location. This is an inadequate approach for some mining applications because
management may implement operating strategies that capitalize on geological
structure such as selective zone closure in response to low mineral prices.
This paper introduces a project structure model that reflects the heterogenous nature of mineral deposits by representing
the project as a real asset portfolio in which each zone represents a
portfolio asset. The project is operated in discrete intervals by choosing,
at the start of each interval, an operating mode from a set of competing
operating modes. Each mode specifies the combinations of zones that will be
active and the amount of project capacity that is built, abandoned or
temporarily closed. A two-zone mining example is used to demonstrate the
proposed model and show how operating strategies that capitalize on
geological structure can add value.
Pricing and risk management of
commodity-contingent assets requires an adequate specification and estimation
of the risk-adjusted underlying stochastic commodity prices. Recent efforts
include Gibson, R, Schwartz, E. S. (1990), Schwartz, E. S. (1997), Schwartz,
E. S., Smith, J.E. (2001), and Cortazar et al (2000) among many others. A
shared attribute of all of them is their reliance only on linear payout
assets (futures and sometimes swaps) for estimation purposes.
The benefit of using futures prices is that they trade in a relatively deep
market. On the other hand the drawback of this approach is that some process
parameters (i.e. volatility) may be poorly estimated, because they do not
have a strong effect on futures prices. This paper explores the use of option
prices (in addition to futures prices) to estimate commodity stochastic
prices and discusses preliminary evidence on the behavior of the proposed
models for valuing option-like assets.
We analyze multi-period harvest problems
for a renewable resource under biological uncertainty when harvesting is
size-dependant. First, we show that the decision to harvest can be modeled as
a real option and we derive analytical expressions for the value of the
resource stock and the mean time between harvests, with and without
uncertainty. We then illustrate numerically how uncertainty affects the
decision to harvest: when uncertainty increases from zero, the amount
harvested and the stock biomass at harvest first increase, and then decrease
because of the risk of extinction when uncertainty is high enough. This paper
is a first step towards defining sustainable harvesting rules under
uncertainty.
Investments in R&D for product
application extensions and in infrastructure service enhancements have
interesting similarities and can be analyzed in a common template. They
typically require multiyear investments in outcomes exposed to several
sources of uncertainty. Technical risks, market size and acceptance, and
actual capital requirements make these investments difficult, if not
impossible, to evaluate with a static discounted cash flow analysis.
Further complicating the initial investment decision, as well as the timing
of intermediate investments, is the potential for a competitor's entry that
takes market share or even eliminates demand for the product or service.
This article describes a Real Options
analysis of shared options on extensions and enhancements in the energy
service industry. We illustrate the analysis framework, the data requirements
and the insight gained from Real Options. The material focuses on extracting
managerial insight from a Real Options assessment of incremental investment
programs, whose timing and success may be impacted by external market forces.
Han Smit (Erasmus U.)
and Lenos Trigeorgis (U. Cyprus and U. Chicago), R&D Option Strategies
Bart Lambrecht
(U. Cambridge, UK), The Timing
and Terms of Mergers, Stock Offers and Cash Offers
This paper presents a real options model
for the timing and terms of mergers, stock offers and cash offers under
product demand uncertainty and complete information. The timing of mergers is
shown to be globally efficient. Bidding causes stock offers to happen
inefficiently late and pure cash offers are even more inefficient because
bidding is combined with the fact that the acquiree
does not hold a stake in the restructured company. The acquiree
prefers stock offers, whereas the acquirer prefers mergers or cash offers.
Whether restructurings happen in a rising or falling product demand movement
depends on how the restructuring alters the demand elasticity of the firms'
market capitalization. restructurings leading to an
increase (decrease) in the firms' demand elasticity happen in a rising
(falling) product market and are expansive (contractive) in nature. The paper
supports the hypothesis that takeover activity is driven by economic shocks.
Panel Discussion: Challenges and Future Prospects I
(Practitioner Perspectives from Consulting Firms)
Moderator: Martha Amram (independent consultant)
Panelists include:
Ari Axelrod
(Boston Consulting Group)
Stephen Black (PA Consulting)
Adam Borison (ADA/PricewaterhouseCoopers)
Raul Guerrero (Accenture)
Alberto Micalizzi (Real Options Group)
Remy Schosmann (Ernst & Young)
SATURDAY
Switching Options
in Manufacturing, Supplier Contracts and Shipping Chairperson: Alex
Triantis (U. Maryland)
Nahoya Takezawa (International U. Japan), The Option
to Switch Scheduling Priority in Manufacturing
Consider a group of product devices such
as DRAM's of different specifications while such products can be purchased
from the market at the spot price that fluctuates over time.
Because of this, a final product-maker often makes a contract with a supplier
where delivery dates and prices of such devices are pre-determined. In this
situation, the supplier typically ignores the price volatilities and
organizes its production activities based on the delivery dates. A prevalent software package using MRP for example cannot take the price
volatilities into account explicitly. In this paper, a theoretical framework
is exhibited where the supplier determines its production schedule based on
not only the delivery date and the pre-determined prices but also the price
volatilities of the devices.
Using a real options framework, we value
and analyze supply contracts characterized by exchange rate uncertainty,
order quantity flexibility and supplier-switching options. Analogous to the
portfolio optimization framework, our framework analyzes the incentives that
the suppliers face in accepting order level flexibility. The resulting
tradeoff (for the supplier) is a balance between greater volatility in the
supply schedule and the prices that the producer pays. In this context, we
explicitly model how flexibility can be beneficial to both the producer and
multiple suppliers. In other words, how a contract with quantity flexibility
can be Pareto optimal, with neither the producer nor the suppliers being
worse off as compared to inflexible contracts. This implies that option to
switch between suppliers is not costless, thus resulting in the producer
having to compensate suppliers in a manner consistent with a profit
maximization objective for all parties.
Several entry-exit models
under price uncertainty are discussed by a new markup approach to investment,
starting with the classical model by Dixit (1989).
The markup approach, introduced by Dixit et al.
(1999), enables us to state the expected value of the firm in the entry-exit
model as a function of a chosen pair of entry and exit trigger prices. The
optimal policy appears by maximizing the value function with respect to the
trigger prices. Extensions being discussed include endogenous production
costs, diminishing production capacity over time, limits to the number of
available switches, and various models with scrapping decisions and
investment lags. The main new extension allows for an investment lag in the
entry-exit-scrapping model by Dixit (1988).
Implications of the investment lag are investigated by use of experimental
data and empirical data from shipping. We also correct some results on
investment lags from Bar-Ilan and Strange (1996).
Theoretical Issues I:
Options on Information, Incomplete Information, and Non-Traded Assets
Chairperson: Michael Brennan (UCLA)
Andrew Chen (Southern Methodist U.),
James Conover and John Kensinger (U. North Texas), Virtual
Options: Evaluating Options on Information
With virtual options the underlying assets
are information and the rules governing exercise are based on the realities
of the information realm (infosphere).Virtual
options can be modeled as options to "purchase "information items
by paying the cost of the information operations involved.Virtual
options arise at several stages of value creation.
The initial stage involves observation of physical phenomena with
accompanying data capture.The next refinement is to
organize the data into structured databases.Then
information is selected from storage and synthesizing it into an information
product (such as a report,article,or design
specifications for a product to be fabricated in the physical realm).Then the
information product is presented to the user via an efficient interface that
does not require the user to be a field expert.Virtual
options are similar in concept to real options but substantially different in
their details,since real options have physical
objects as the underlying assets and the rules governing exercise are based
on the realities of the physical world.Also,while
exercising a financial option typically kills the option, virtual options may
include multiple exercise.Virtual options may
involve high volatility or jump processes as well,further
enhancing their value.Application of option pricing
theory to real options has yielded worthwhile tools for disciplined decision making,and the potential also is great for worthwhile
decision support tools based upon virtual options.This
paper extends several important real option applications into the information
realm,including jump process models and models for
valuing options to synthesize any of n information items into any of m
output products.
In this paper we discuss a real options paradox of managerial intervention directed
towards learning and information acquisition: since options are in general
increasing functions of volatility whereas learning reduces uncertainty, why
would we want to learn? Examining real options with (costly) learning and
path-dependency, we show that conditioning of information and optimal timing
of learning leads to superior decision-making and enhances real option value.
A recent topical problem is how to deal with claims on `non-traded' assets. A
natural approach is to choose another similar asset or index which is traded
to use for hedging purposes.
To model this situation, we introduce a second non-traded log Brownian asset into
the well known Merton investment model with power-law utility. The investor
has an option on units of the non-traded asset and the question is how to
price and hedge this random payoff. The presence of the second Brownian
motion means that we are in the situation of incomplete markets. Employing
utility maximisation and duality methods we obtain
an approximation to the optimal hedge and reservation price. These are
computed for some example options and the results compared to those using
exponential utility.
Theoretical Issues II:
Options with Private Information and Incentives
Chairperson: Lenos
Trigeorgis
(U. Cyprus and U. Chicago)
The paper presents a model for the
valuation of marginal offshore oil fields that are geographically close to an
existing major oil production installation. The correct development strategy
for these satellite oil fields is given by optimally exercising American real
options to develop the fields. The most cost efficient way of developing the
fields is to connect the marginal fields, e.g. via a sub-sea development, to
the production unit of the main field. A problem arises by the fact that the
ownership structures of the satellites are not identical for all fields.
Hence, the order and timing of the development of the fields are a question
of negotiations between the licensees. A non-efficient outcome of the
negotiations reduces the total value of the oil fields.
An investor owns a right to invest in a
project that generates positive cash flows when the investment is undertaken.
Both the value of the future cash flows and the investment cost follow
stochastic processes. Thus, the investment project takes the form of an
exchange option of American type. In the paper we analyze this investment
project when the investor needs an agent to undertake the investment of the
project, and the agent has private information about the investment cost.
In the first part of the paper we assume that there is only one agent having
private information, and the problem is analyzed within a principal-agent
framework. The investor's problem is to optimize the compensation to the
agent. To induce the agent to make the preferred investment decision, the
investor needs to leave the agent some information rent.
In the second part we extend the model by assuming that n agents
compete about the contract. Each agent has private information, and the
competition is organized as an auction. We discuss how competition reduces
the information rent and the inefficiency of the chosen investment strategy.
T. Cottrell (U. Calgary, Canada), Incentive Contracts in the Presence of Real
Options
Theoretical Issues III: Modelling
and Numerical Analysis
Chairperson: Gordon
Sick (U. Calgary)
In this paper we produce a formula for a
finitely lived, perfectly reversible option on a flow. For this real option
that allows frequent and costless switching between the maximum of two asset
flows, we first examine the perpetual and then the finite cases in terms of
switching thresholds and values. The finite option value is inferred from the
perpetual using an annuity argument. Applications include energy and
commodity consumption costs where switching between flows can occur
frequently and costlessly.
This paper introduces the first passage
time approach to study optimal option exercise rule for geometric Brownian
motion process to a boundary. I have derived analytical results on the first
passage time probability, density and its expectation. The results on the
first moment of the first passage time clarify some recent controversies on
the sign of uncertainty on investment. The first passage time provides an alternative
characterisation of optimal exercise rule. In
addition, we establish a new framework for testing real option models. The
approach is applicable to other stochastic modelling
in finance and economics.
We propose a log-transformed binomial
lattice approach for pricing options whose payoff depends on several state variable following a joint diffusion process. Our method
extends the log-transformed approach proposed by Trigeorgis (1991) to several
state variables and improves other known lattice algorithms (Boyle, Evnine and Gibbs (1989) and Kamrad and Ritchken (1991)).
The method we propose is consistent, stable and efficient. We present some
applications of our method both to financial and real option pricing
problems.
Panel Discussion: Challenges and Future Prospects II
(Academic Perspectives)
Moderator: Alex Triantis (U. Maryland)
Panelists include:
Bhagwan Chowdhry
(UCLA)
Blake Johnson (Stanford U.)
Robert McDonald (Northwestern U.)
Eduardo Schwartz (UCLA)
Gordon Sick (U. Calgary)
Lenos
Trigeorgis
(U. Cyprus, U. Chicago and ROG)
|
