Technical Terms Used in Project Portfolio Management (Continued)

multi-attribute utility analysis (MUA)

Also called multi-attribute utility theory, multi-objective decision analysis and multi-criteria decision analysis, MUA is a decision analysis method for quantifying the value of something (e.g., a project) based on its characteristics, impacts, and other relevant "attributes." MUA is useful for project portfolio management because it provides a relatively simple and defensible way to quantify project value, including non-financial (or "intangible") components of value.

The term MUA derives as follows. In economics, "utility" is a measure of the value or satisfaction derived from something. Sometimes, utility depends on a single attribute. For example, the utility to me of the cash in my wallet depends on one attribute; namely, the number of dollars. The more dollars I have, the greater the utility of those dollars. Mostly, though, the utility of something depends on more than one attribute. For example, if I travel to Europe, the utility of the cash in my wallet would depend on two attributes—the number of dollars and whether those dollars are US or Euros. "Multi-attribute utility" is a measure of value that depends on or is determined by more than one attribute of the thing being valued.

More precisely, MUA is an approach for deriving a "utility function" (a decision model) that, according to decision theory, quantifies a decision maker's preferences over the available alternatives to a decision. The utility function, U, is such that the best alternative is the one that maximizes U. Thus, if we could determine the function U, we could calculate which of the alternatives to a decision is the most desirable. MUA is a step-by-step process for determining U, a process that is efficient in the common case where multiple criteria (attributes) determine the desirability of alternatives.

Utility functions for evaluating projects are typically multi-attribute because the desirability of a project depends not just on the dollars the project returns (one attribute), but also on various additional criteria that capture other types of benefits that may accrue from the project (e.g., improved safety, increased knowledge, etc.). The concept of MUA is that the correct multi-attribute function U will be much easier to find if it can be written as some simple combination of single-attribute functions (e.g., as a sum of a utility function describing the relative desirability of various levels of financial performance and another utility function describing the desirability of various levels of safety). If the utility function separates in this way, then, deriving the function is easier because each single-attribute function can be assessed without reference to the other attributes.

MUA provides a step-by-step process for deriving a multi-attribute utility function (developing a decision model) as a combination of single-attribute functions. The first step is to structure the objectives of the decision into a hierarchy. The next step is to define measures for the achievement of each objective and to develop a single-attribute utility function for each measure. Tests are performed on the measures to determine how the single-attribute utility functions should be combined (e.g., added, multiplied, etc.). Since the objectives are not necessarily equally important, weights must be assigned, and techniques are provided to correctly set the weights based on the decision maker's answers to questions indicating willingness to trade off various levels of performance on the different measures.

MUA can be easily integrated with more traditional financial measures of project value. For example, within MUA, net present value (NPV) can be used as a metric for accounting for the project's direct financial value. Likewise, MUA can be used with expected net present value (ENPV) to include an explicit accounting of project uncertainties.

MUA has been applied to many different types of decision problems and there is a vast literature on the subject. The major benefit of the approach is that it produces a single number, expressible in equivalent dollars, that measures the overall value (utility) of a project. Since this number is derived through a step-by-step process beginning with the specification of objectives, the logic is open and explicit, can be reviewed, and may be changed if any assumption made at any step is judged to be inappropriate. Scores and weights are also explicit and are developed according to established techniques and can often be cross-referenced to other sources of information on relative values. Because MUA provides a detailed "audit trail" of assumptions for reviewers, it has been used by government agencies to help make controversial public policy decisions.

A key characteristic of MUA is its explicit and extensive reliance on judgment. The necessary judgments include value judgments provided by policy makers (e.g., Which do we want more: a project that would generate $2 million or a project that would reduce the average number of annual worker injuries by 25%?) and technical judgments provided by specialists (e.g., How do we estimate the impact of a project on the annual number of worker injuries?). This reliance on subjective judgment is sometimes interpreted as a weakness, as applications may appear overly subjective. Judgments, however, are required for virtually all important decisions. The fact that MUA makes those judgments explicit is an advantage. Since the judgments and assumptions are represented as inputs to a decision model, interested parties can explore via sensitivity analysis whether changes would alter conclusions.

A limitation of MUA is that it is not easy to apply correctly. Meeting the technical requirements necessary to satisfy the assumptions of the approach requires skill, and applications generally must be guided by specialists in the field. Furthermore, like all other decision tools, the decision model produced by MUA involves simplifications that may introduce errors into recommendations.

net present value (NPV)

The traditional method for quantifying the financial attractiveness of a project. NPV, also called discounted cash flow (DCF), represents the amount in today's dollars (present value) by which all income projected from the project exceeds all costs.

Basically, NPV attempts to answer the question, "What is the equivalent, lump-sum worth of this project?" According to NPV logic, given two projects, the one with the larger NPV should be preferred. Also, any project with a positive NPV should be viewed as a good investment.

NPV computes the present value for a project by discounting estimated future incremental cash inflows and outflows. Typically, the discount rate is chosen to represent a required rate of return or target yield for the capital invested. To accurately calculate a project's NPV, it is necessary to estimate the life-cycle cash flows that would result from doing the project— not just the project costs, but also all of the financial benefits that would result from the project. For example, if a project improves productivity, the future cost savings that would result should be included in the estimated cash stream. Cost estimates reflect the total cost of ownership (TCO) perspective. Thus, costs include project investment costs, future operating costs, and any "exit" costs associated with ultimately phasing out whatever it is that the project produces. Since taxes can have a significant impact, cashflows are often calculated "after-tax," accounting for depreciation, working capital, and other considerations.

NPV cannot be used directly as a metric for ranking projects because it ignores the size of the projects being compared. Larger projects tend to have larger NPV's. Thus, projects with large NPV's tend to consume greater portions of the available budget. However, NPV can be used to translate the financial benefits expected from a project into an equivalent dollar value which, if divided by the project cost, can be a useful metric for ranking projects (assuming the projects are independent).

By requiring that a single, nominal cash flow be identified for the project, NPV ignores uncertainty. This limitation may be addressed by using NPV in conjunction with simulation. If alternative project cash flows are simulated, they may be used to generate a range or distribution of project NPV's. This distribution may be interpreted as describing the uncertainty over the actual value that the project will generate. See expected net present value (ENPV) for more discussion.

The major limitation of NPV for project prioritization is that it underestimates the true value of projects (e.g., the impact of a project on shareholder value) because it leaves out "intangible" project benefits that are difficult to express as incremental cash flows. One such omitted component of value is "option value," the value associated with options embedded in or created by the project which may allow management to better respond to future risks and take advantage of future opportunities. See real options analysis for an explanation of option value. One method for addressing this bias is to add to the estimated project cash flows additional dollar amounts that represent the equivalent dollar value of the non-financial project benefits. See multi-attribute utility analysis for a method for doing this.

Another problem with NPV is that it is not always clear what discount rate should be chosen. According to finance theory, the correct rate is the return available from investing in securities equivalent to the risk of the project being evaluated. Research on real options shows that the discount rate ought to be adjusted over time depending on how uncertainties are resolved and on the project-management strategy. Using a constant discount rate for a project implicitly assumes that uncertainty increases over time in a specific way (geometrically). If the discount rate is adjusted upwards to account for the risk of the project, there will be a bias toward short-term, quick payoff projects because project benefits that occur in the more distant future will be severely discounted.

payback period

The amount of time it takes for the cumulative cash flows from a project to equal the initial investment. An investment will have paid for itself in the year, or month, where the cumulative cash flow first becomes positive. Payback period is a popular metric for evaluating projects because of its simplicity, emphasis on liquidity, and obvious responsiveness to external financing pressures. However, because payback period does not provide a measure of project or portfolio value, it cannot be used as a metric for prioritizing projects.

Another major limitation of the payback period method is that it ignores cash flows after the payback period. For example, a small project may have a break-even point at six month from the start of the project. A larger project that costs twice as much may not break-even until 12 months after the start of the project. Based on the analysis of the payback periods, the company may choose to go with the smaller project. The analysis would ignore the fact that after two years, the larger project would have produced three times the dollar savings as the smaller project. Also, as no discounting is involved, the payback period overlooks the time value of money (cost of capital).