A REAL OPTIONS PERSPECTIVE TO ENTERPRISE ARCHITECTURE AS AN INVESTMENT ACTIVITY
Institute of Systems Science
National University of Singapore
The ever-increasing expenditure on information technology (IT) is accompanied by an increasing demand to measure the business value of the investment. This has prompted enterprises to take an architectural view of their information systems (IS) and supporting technologies. However, many crucial enterprise architecture frameworks and guidelines are characterized by lack of adequate theoretical or conceptual foundations. Important but inadequately formulated concepts include architectural assessments, governance and architecture maturity models. These, though central to the enterprise architecture development process, remain in their current formulations largely wisdom driven rather than engineering based approaches. Absence of adequate scientific or mathematical foundations for enterprise design and engineering significantly impede enterprise architecture initiatives. The current body of knowledge is limited to reference architectures where the implementation challenges are left to the enterprises themselves.
This paper views enterprise architecture development as largely a
process of decision making under uncertainty and incomplete knowledge. Taking value
maximization as the primary objective of the enterprise architecture decision-making
process, the paper attempts to develop guidelines for value enhancement. The paper assumes
that portion of the value of enterprise architecture initiative is in the form of embedded
options (real options), which provide architects with valuable flexibility to change
plans, as uncertainties are resolved over time. Plausibility of using such an approach to
develop a better account of critical enterprise architecture practice is focused on three
Enterprise engineering and integration are crucial components in architecting enterprises . Large numbers of enterprise engineering initiatives are (reference) architecture driven. A reference architecture shows the anatomy of the life cycle of an enterprise (logical structure of activities). The common theme in all current reference architectures is the existence of the Enterprise Life Cycle concept [Bernus et. al., 2003]. This allows an enterprise to be conceptualized, conceived, designed, developed, operated, maintained and possibly retired (or renewed). Most reference architecture lifecycle representations revolve around these phases / activities.
A study of all existing reference architectures reveals that many important enterprise architecture (EA) guidelines and best practices lack adequate theoretical or conceptual foundations. This is partly explainable by the fact that most of the developments of reference architectures have traditionally been driven by the industry where quick implementation is preferred over theoretical / conceptual rigor. In order to address such issues there have been a few significant efforts to concretize recommendations and best practices into generalized reference architectural approaches mainly by the International Federation of Information Processing (IFIP) and the International Federation of Automatic Control (IFAC), which proposed the Generalized Enterprise Reference Architecture and Methodology (GERAM) [Bernus and Nemes, 1996]. This proposal was then developed by the IFIP-IFAC Task Force and became the basis of the International Standard ISO 15704: 2000. GERAM defines a toolbox of concepts for designing and maintaining enterprises through their entire lifecycles [Bernus et. al., 2003]. It represents the common (baseline) set of requirements that other above mentioned reference architectures must fulfill in order to be GERAM compliant. All of the existing reference architectures mentioned earlier have been mapped to GERAM requirements [Noran, 2003].
Organizations embarking on an EA journey usually prefer to use one of
the available reference architectures to speed up implementation and take advantage of
collated best practices [Perks and Beveridge, 2003]. Concepts and guidelines in reference
architectures, which are central to the architecture development process, remain in their
current forms more heuristical than scientific. In practice, the lack of scientific or
mathematical foundations for enterprise architecture significantly impede enterprise
engineering for three primary reasons:
Using Enterprise Life Cycle concept as the base and GERAM requirement of an EA initiative to have a project based and methodology driven approach [IFIP-IFAC Task Force, 1999], this paper takes an integrated view of EA as largely a process of decision making under uncertainty and incomplete knowledge. EA literature stresses the need for EA initiatives to take value enhancement as the primary objective rather than technical perfection [Bernus et. al., 2003]. The traditional return on investment (ROI) expects costs of investment to be returned within the scope of the initiative at hand making it too tactical. Assessing the value of investments in EA, whose impact may not be apparent immediately, requires another measure. Gartner  proposes the use of return on assets (ROA) as an alternative, which focuses more on value enhancement through increases in productivity of their capital assets.
With the above background, the first part of the paper is a brief discussion on the importance of economics in enterprise engineering with the view that enterprise architecture activity is one of investing valuable resources under uncertainty [Perks and Beveridge, 2003] with the goal of maximizing value added , rather than lowering total cost of ownership (TCO). While it is possible to adopt a complex view of value, the paper takes a more limited view of market value added to the enterprise. Economic value of an enterprise is greatly influenced by structure. The reason being that structure dictates behavior (including flexibility) that is displayed by an enterprise in the face of changing and uncertain business environment. Under such conditions flexibility in the architecture development process can provide great value by potential to avoid risks and take benefits of new opportunities as they come by. The paper discusses some of the uncertainties in architecting an enterprise and the traditional approaches to address EA related economics.
The second part of the paper elaborates some of the available
approaches to value flexibility and provides a brief overview of options pricing theory
which is the basis for real options approach used in this paper. This approach is then
discussed in the context of three critical areas: (1) options in timing of important
architectural investment decisions; (2) compound options in architecture development
process; and (3) options interpretation to architectural maturity levels. Five concrete
real options are suggested as an approach to address some of the identified EA initiative
Traditional enterprise engineering focuses more on structure and technical perfection leading to lower TCO than value added (or asset productivity) [Gartner, 2002]. This paper takes the view that enterprise architecture development initiative is one of investing valuable resources under uncertainty with the aim of maximizing value added to the organization. This view is consistent with one of the documented business benefits of The Open Group Architecture and Framework (TOGAF) which mentions, better return on existing information and reduced risk for future investments [Perks and Beveridge, 2003]. While it is possible to adopt a complex view of value, this paper takes a narrow view: that value is measured in terms of asset productivity improved for the enterprise. This view is supported by the fact that according to Gartner, by 2007, IT asset productivity will drive market capitalization.
With respect to timing of architecture related investments, most reference architectures rely on rules of thumb. TOGAF for instance provides a Standards Information Base (SIB) for enterprises to choose from a multitude of products and standards to realize the enterprise architecture [The Open Group, 2003]. However, the need is to have a clear basis for reasoning about the timing of such investment decisions to have a better-founded approach to managing complex enterprise architecture initiatives. This need is driven by the fact that enterprises find it difficult to enhance their IT Investment Management Maturity [U.S. General Accounting Office, 2004] unless IT decisions are based on sound architectural factors.
UNCERTAINTIES IN ENTERPRISE ARCHITECTURE
|On one dimension TOGAF ADM imposes a phased structure on an architecture development initiative, where the aim during each phase is to discover / uncover new information and reduce key uncertainties, with decisions about whether to invest in subsequent phases. On the temporality dimension, TOGAF Architecture and Solutions Continuum stresses the development of alternatives, creating an option to choose the most appropriate one.|
|Having architected for flexibility, the key decision for the enterprise
is knowing when to take action in order to maximize enterprise value.
TRADITIONAL ENTERPRISE ENGINEERING ECONOMICS
Analyzing capital investments involves justification and assessment of these investments and it is frequently used as a criterion for investment decision-making or capital budgeting. Value is the single time-value discounted figure that represents all future net profitability [Mun, 2002]. Over time, value of an asset may or may not be identical to its market price. Traditionally, the three prevalent approaches to valuation of strategic assets include: the market approach, the income approach and the cost approach. The market approach views value of comparable assets in the market, assumed to be at an equilibrium level due to market forces, as a way to value assets. Usually market approach looks at several criteria to make such assessments [Mun, 2002]. The income approach on the other hand looks at future cash flows and future potential profits, attempting to quantify such numbers to a present value. Some of the measurements include internal rate of return (IRR) and NPV. The cost an organization would incur should it decide to replace or reproduce the assets future profitability potential, including cost of its strategic intangibles, is used in the cost approach. Despite availability of several approaches [Irani et. al., 1997; Wen et. al., 1998], assessment of investments in IT in general is still a fledgling practice [Keen, 1991; Banker et. al., 1993]. According to a survey conducted by Ballantine and Stray in 1999, the most utilized financial criteria to evaluate IT projects found that more than 25% of organizations use NPV method. The point to note is that all investment evaluation approaches look at IT in general and currently there are no specific approaches to assess and analyze investments in enterprise architecture initiatives. The most fundamental drawback is that static NPV fails to capture the future value of flexibility under uncertainty [Mun, 2002; Brach, 2003]. The associated investment decision rule is that investment should be made if the NPV is positive, otherwise not.
NPV approach is wrought with several problems that include undervaluing an asset due to lower cash in the shorter time frame, variability in the weighted average cost of capital discount rate and forecasting errors in assessing future cash flows [Coleman and Jamieson, 1994; Remenyi et. al., 1996; Mun, 2002]. The bottom line is that static NPV is generally not a good measure for valuing an enterprise because it tends to overlook one of the key sources of value: the flexibility of the enterprise to adapt to newly discovered and dynamic information, markets and environment.
The three main approaches to help make the value of flexibility tangible include: (1) dynamic discounted cash flow analysis; (2) decision analysis (using utility theory), and; (3) real options. Each approach is appropriate under certain conditions and has its strengths and weaknesses. This paper uses the real options approach due to its suitability and wider applicability within the context of EA investments. The appropriateness of the real options approach will further be apparent as the paper progresses.
Given the high level of investments required to take advantage and
optimize benefits of Enterprise Architecture, incorporating flexibility into EA investment
decisions that accounts for active management and adaptation to changing market
conditions, creates a more efficient and realistic investment framework. Also as EAMMF is
a five level system (see Figure 4), each level representing higher degree of EA Management
maturity, the possibility of options to move from lower to higher maturity levels do
exist. These options can be called EAMM Options. However, application of OPM to capital
investments in EA initiatives must address the following issues [Taudes, 1998].
In statistical decision theory risk is defined as the expected value of a loss
function. Risk is commonly measured in terms of consequences and likelihood, where
likelihood is a qualitative description of probability or frequency. In the context to
risk, it is critical to note that risk arises from choice [Emblemsvag, 2004]. Three
crucial facts associated with risk are:
This brings us to concept of uncertainty, which is defined as the lack of information and knowledge or the state of being uncertain, in doubt or hesitant. Uncertainty exists is all situations that are unknown, unpredictable, open ended or complex.
Uncertainty and the value of flexibility in the face of uncertainty are at the core of both enterprise architecture development and finance. Reducing risk by introducing uncertainty is what this paper proposes and it uses Monte Carlo methods to achieve this. Monte Carlo methods usually rely on introducing uncertainty into the models since no uncertainty exists, allowing assessment of the impact of uncertainties on the outcomes. The exact use of Monte Carlo methods in the context of EA decision-making is elaborated in a later section.
Enterprise architecture development is based today largely on experience and
heuristics. Reference architectures like ARIS, PERA, CIMOSA, GRAI / GIM, TOGAF, FEAF
typically provide architecture principles as an aid to the architecture development
process. Architecture principles are general rules and guidelines, intended to be enduring
and rarely modified, which support the deployment of all IT resources and assets across
the enterprise, reflecting a level of consensus among the various elements of the
enterprise [The Open Group, 2003]. Developed from generic IT principles, architecture
principles are usually of two types [The Open Group, 2003]:
A study of TOGAF Version 8.1 and Federal Enterprise Architecture Framework (FEAF) Version 1.1 which specify 20 and 8 architecture principles, respectively [CIO Council, 1999; The Open Group, 2003], reveals that principles are: (1) Qualitative and not based on documented scientific or mathematical theories; (2) Focus of aspects like decision making, prioritizing, linkage to business objectives, and prudent management of investments among others. In general, the need to delay decisions for as long as possible can be inferred (directly and indirectly) from several architecture principles like: maximizing benefit to the enterprise, aiming technology independence, requirements based change, displaying responsive change management, coordinating technology investments and controlling technical diversity. Under these conditions, use of options pricing theory presents potential to understand why architecture principles seem to work, to explore conditions in which they are / are not valid and to develop prescriptive models for architecture decision making. This is made possible as real options thinking make key variables linked to value of decision flexibility under uncertainty, explicit. Understanding the underlying mathematics in architecture principles allows organizations to reason on when, why and how to apply a certain principle and what results are to be expected.
V. AN OPTIONS INTERPRETATION OF ENTERPRISE MODULES
Enterprise modules are implemented building blocks, which are utilized as common
resources in enterprise architecture development [IFIP-IFAC Task Force, 1999]. A building
block is a package of functionality specified to meet business requirements [The Open
Group, 2003]. TOGAF for instance identifies to types of building blocks, Architecture
Building Blocks (ABB) representing bundles of functionality and Solutions Building Blocks
(SBB) that represent real products and custom developments. Building Blocks / Enterprise
Modules are characterized by:
As is evident from the characteristics above, Enterprise Modules /
Building Blocks actually facilitate implementation of several architecture principles.
Having discussed earlier the application of real options to decide on the optimal time of
architecture investment (value of timing), this section focuses on value of modularity.
Although real options have not been used in enterprise architecture activities, it is
interesting to see the existence of embedded options when designing enterprises in a
modular manner. As seen earlier, the primary idea to facilitate enterprise module
development is to enhance the ability to assemble enterprise entities from a set of
available building blocks. Thus enterprise modules (i.e. ABB) create options to select the
best solution from a set of solutions (SBB). Solution variants are usually developed by
experimentation, and the challenge is to create real options for enterprises (architects)
to modify internal details without impacting the rest of the system (as long the
interfaces of the SBBs do not change). This makes investment in developing building blocks
provide benefits that can be expressed as the value of such an option. The actual
specification and development of building blocks takes place at two levels, shown in
VI. METHODOLOGY FOR ENTERPRISE ARCHITECTURE INVESTMENT MANAGEMENT
Typically EA initiatives consist of steps that include: needs identification, concept development, requirements analysis, preliminary and detailed design, implementation, operation and decommission. In addition to life cycle phases, EA development involves understanding the needs of different (and relevant) stakeholders and incorporating their perspectives. These stakeholders perspectives are called views. The detailed representation of the EA development lifecycle and various views are shown in Figure 6.
Any methodology proposed for investment management of EA development activities must synchronize with the overall EA initiative, as it is imperative to understand broadly the EA initiative in order to develop a reliable and acceptable investment management methodology for it. The proposed methodology, adapted from IT investment methodology put forth by Kulatilaka, Balasubramanian and Storck  consists of eight steps: (1) assessment of current capabilities, (2) identification and specification of desired capabilities, (3) design of a contingent investment program to attain desired capabilities, (4) estimation and valuation of costs and benefits of realized capabilities in terms of cash flows, (5) modeling uncertainties using event trees, (6) identification and incorporation of managerial flexibilities using decision trees, (7) performing real options analysis (ROA) and an optional (8) optimization of portfolio and resources. As the primary benefit of embedding real options within the investment process is to allow managerial flexibility, addressing investment risks is a critical factor [Wen et. al., 1998; Benaroch, 2002]. Hence, the proposed investment management methodology based on real options approach, elaborated in the following sections, incorporates management of investment risks through introduction of uncertainties. The following subsections elaborate each of the eight steps.
While there are several activities involved in each of the above phases in Figure 7, the following sections will only elaborate those that are relevant in the context of an EA initiative. Real options based investment analyses elaborated in subsequent sections would largely be used to select options and provide basis for control and evaluation.
Application of option pricing theory to generic IT investments have largely focused on evaluating and analyzing such investments that embed a single, a priori known option (such as deferral option, prototype option) [Benaroch, 2002]. Because real options are not inherent in any IT investment, they must be deliberately planned and intentionally embedded in a target IT investment in order to incorporate managerial flexibility and risk control mechanism. As an IT investment (including EA investment) involves multiple risks, there are several potential ways to configure investment paths using series of compound options [Benaroch, 2002].
In order incorporate managerial flexibility in EA initiative
investments, it is important to understand the types of risks that an Enterprise
Architecture initiative faces. While there is no documented list of specific risks faced
during EA development process, analyzing risks in generic IT / IS provide us some
insights. IS research has categorized IT risks into those related to IT development [Alter
and Ginzberg, 1978; Davis, 1982; McFarlan, 1982; Boehm, 1989] and those arising outside
the scope of IT development [Clemons and Weber, 1990; Clemons, 1991; Keen, 1991; Kemerer
and Sosa, 1991; Deise et. al., 2000]. Synthesizing literature on IT risks, they can be
broadly placed into three categories [Benaroch, 2002]. They are:
MAPPING ENTERPRISE ARCHITECTURE OPTIONS TO RISKS
The use of real options to support enterprise architecture development decision-making is an attractive possibility, but it is crucial to understand some of the critical issues and challenges in adopting this approach, specially for organizations who have no or limited prior experience with real options based investment analysis and decision making and elementary investment management practices. These issues, elaborated in subsequent sections, are categorized into:
Knowledge of Current Asset Value: Assessing the value of EA
initiatives is imprecise and still evolving. This difficulty stems from the fact that
benefits of EA are largely intangible and extended over time, besides its success be
qualitative in nature [Bernus et. al., 2003]. While there is evidence that EA maturity is
positively related to IT / IS performance [Perks and Beveridge, 2003], difficulties in
estimating EA initiatives will remain a serious impediment to accurate valuation of some
|Some major issues and challenges that organizations are likely to face
IT Investment Management Maturity: Adopting a real options based approach to investment analysis necessitates a paradigm shift where corporate valuation may no longer depend on traditional fundamentals but rather on future expectations. Real options are a new way of thinking rather than mere application of advanced analytical methods. As seen in Figure 8, an organization moves up the investment maturity framework in a phased and gradual manner. Evidently, organizations already at the upper reaches of the maturity stages are more likely to embrace the proposed new approach. Hence it might be beneficial for managers to assess their investment maturity before exploring the adoption of real options based framework in the context of EA initiatives.
Alignment to Investment Management Lifecycle: It is critical for organizations to recognize the fact that when analyzing EA investments, alignment of the Architecture Development Methodology to the investment lifecycle is imperative to create and build viable EA options that can be analyzed and embedded within IT investments. The investment lifecycle typically consists of the following stages [Benaroch, 2002]:
o Inception: Most likely facilitated by earlier investments, investment opportunity exists.
o Recognition: Investment is perceived as a viable opportunity to be exploited.
o Building: Investment decision is made and the opportunity is realized.
o Operation: Investment is in operation producing direct and measurable payoffs, which can be compared against planned payoffs.
o Retirement: Investment, while still in operation, continues to produce in direct payoffs, including new investment opportunities and a possibility of initiating new investment cycles.
o Obsolescence: Investment has served its purpose of providing both direct and indirect benefits. Assets and capabilities yielded by the investment become obsolete.
ISSUES RELATED TO ENTERPRISE ARCHITECTURE DEVELOPMENT
The primary purpose of an EA is to inform, guide and constrain the decisions for the enterprise, especially those related to IT investments [CIO Council, 2001]. Essential reasons for developing an EA include [CIO Council, 2001]:
Alignment: Ensuring that the enterprise operates in a way that is aligned with the managements intent.
Integration: Connectivity and interoperability is managed across the value chain.
Change: Facilitating and managing change to any aspect of enterprise, supported by full visibility and transparency.
Time-to-market: Reducing systems development and application generation timeframes.
Convergence: Striving towards a standard IT product portfolio as in the Technical Reference Model (TRM).
As is evident from the paragraph above, the
critical driver for an EA initiative is to control and manage IT investments. Value
enhancement is seldom mentioned explicitly as one of the primary goals of an EA. The shift
from controlling IT investments to enhancement of organizational value can be a large one
for many organizations, often requiring a change in the mindset. The shift is made more
challenging by the following factors, that ultimate influences an organizations
ability to adopt the real options based approach.
SUGGESTIONS FOR FURTHER RESEARCH
Effective use of enterprise architectures is a recognized strength of successful
enterprises. Metaphorically, an EA is to an organizations operations and systems as
a set of blueprints is to a building. EA provides a clear and comprehensive view of the
structure and operations of the organization. This paper presented a real options based
approach to view investments in EA. This approach stresses that EA initiatives have
several associated uncertainties impacting the benefits derived, i.e. substantial portion
of the benefits are derived from implementation opportunities. In contrast to traditional
DCF based methods, where uncertainty of possible future implementation opportunities are
penalized, the proposed approach actually acknowledges the existence of risks and
uncertainties and allows investments to be configured accordingly. Incorporating
managerial flexibility involves understanding and acknowledging the existence of temporal
aspects in the investment cycle. This allows managers to build investment configurations
that suit their organizations and implementation scenario. However adopting real options
based approach has several implications on organizational business and investment
management practices. They include:
The importance of EA Options is driven by the fact that, typically, the rationale for such initiatives is the eroding status quo, that is, within the current business environment options are not economically justified, but when conditions change, for instance, a regulation (or a business practice) requiring all IS / IT entities to have minimum architecture maturity level, they can bring several benefits, including being an important differentiator. As the value of EA options becomes obvious to competitors inducing them to exercise their options, the value of such options can witness large erosion. Hence in such a scenario it is important for organizations to act early, thus avoiding loss of valuable opportunities if the value of options from active management is undervalued or neglected.
Alter, S. and M. Ginzberg (1978) Managing uncertainty in MIS
implementation, Sloan Management Review, 20 (1), pp. 23 31.
C4ISR Architecture Working Group (AWG) (1997) C4ISR Architecture Framework Version 2.0,
U.S. Department of Defense.
ABB Architecture Building Block
ABOUT THE AUTHOR
Pallab Saha is a member of the faculty with the Institute of Systems Science, National University of Singapore. He has a Ph.D. (Engineering) in Information Systems. His current research interests are in areas of enterprise engineering, business process management, business dynamics and utility of information technology in implementing business strategies. Prior to academics, he has held consulting positions in several large IT organizations and worked on projects for Fortune 100 companies.