Technical Terms Used in Project Portfolio Management (Continued)

hurdle rate

A specified rate of return for a project or other investment intended to represent the minimum return that the organization will consider. The hurdle rate is also often referred to as the required rate of return.

Typically, the hurdle rate is the discount rate to be applied to the cash flows anticipated from the project. If the net present value (NPV) of cash flows using the hurdle rate as the discount rate is negative, the project is rejected. Alternatively, the hurdle rate may refer to the minimum acceptable internal rate of return (IRR) for projects—If a project's IRR is less than the hurdle rate, the project is rejected.

According to investment theory, the hurdle rate should be set equal to the rate of return that the organization could obtain by investing in alternative investment opportunities having similar risks. If the project generates a return greater than what the organization could earn elsewhere (i.e., greater than the opportunity cost of the required investment), the project will add value. Because the opportunity cost is difficult to compute, in practice, hurdle rates for projects are often specified by adding or subtracting a risk premium to the organization's marginal cost of capital, so that a higher rate is specified for project considered more risky.

Hurdle rates are generally a poor way to account for risks when prioritizing projects. As explained in one of our papers, hurdle rates tend to produce a bias toward short-term, quick payoff projects relative to projects of similar risks whose benefits accrue over longer periods of time. Also, while hurdle rates can be used to decrease the value of projects whose benefits are more uncertain, they do not achieve the converse effect of increasing the value of projects that, if not conducted or if delayed, would tend to increase risk.

internal rate of return (IRR)

The effective interest rate that equates the present value of the income stream generated by a project to the cost of that project. In other words, the IRR is the value that satisfies the equation:

Equivalently, the IRR is the discount rate that causes the project to have a zero net present value (NPV).

The IRR is sometimes misunderstood to be the annual profitability of a project investment. However, for this to be true, the cash inflows derived from the project would have to be reinvested in opportunities that produced a return equal to the project's IRR, which is typically not the case.

Like NPV, the IRR is a commonly used criterion for project-by-project selection decisions—all projects that have IRR's greater than the cost of capital are recommended for funding.

When used as a ranking metric, IRR has an advantage over NPV, it does not depend on the size or scale of the project (therefore, its use is more consistent with the concept of ranking based on "bang-for-the buck"). However, at best, the IRR is only a heuristic or approximate logic for prioritizing projects. It creates predictable and significant biases in project rankings and cannot be used to identify project portfolios that create maximum value.

IRR analysis is popular because it compares projects based on the familiar concept of rate of return (a metric analogous to interest rates charged on capital markets). For this reason, it is for many one of the easiest project evaluation methods to understand. Given two investment alternatives of comparable costs, the investment with the higher IRR should be selected. When used as a selection rule for project-by-project decision making, IRR and NPV recommend the same projects (provided that a unique IRR exists, see below). The IRR has a perceived advantage over NPV and other methods that quantify project value in dollars in that it downplays reliance on what might appear to be overly-precise dollar values.

For several reasons, the IRR cannot be used to reliably prioritize projects. First, like other purely financial metrics, the IRR ignores the non-financial components of project value. Furthermore, IRR cannot be used to correctly prioritize based on financial return alone. Ranking projects based on IRR undervalues cash flows that occur late in a project's life (assuming that the IRR is greater than the cost of capital). It therefore creates a bias for projects with early positive returns relative to projects whose returns tend to occur later. The significance of this bias is greater the longer the duration of project cash flows and the more severely constrained the capital budget (if the budget is highly constrained then the IRR's of the projects being compared will tend to be significantly higher than the cost of capital, meaning that future cash flows will be very heavily discounted).

There are several additional disadvantages with IRR. There is no specific formula that can be used to calculate the IRR; it must be found by interpolation. Thus, computing the IRR takes more computation than does the NPV (most spreadsheets do, however, provide an IRR function). The IRR of the portfolio of projects cannot be calculated from the IRR's of the individual projects, so there is no easy way to quantify the performance of the portfolio based on the analysis of the individual projects that make up the portfolio. IRR analysis cannot be extended to account for consideration of project risks or project interdependencies. Finally, there are situations of multiple solutions (no unique IRR) when project cash flow changes direction more than once.

modern portfolio theory (MPT)

A method developed by Nobel Prize winner Harry Markowitz for finding "efficient portfolios," portfolios that have the minimum possible risk for a given expected return. Also called portfolio management theory (or, more simply, portfolio theory), MPT provides a relationship between the market price of an investment, the investment's expected return, and the risk of that investment relative to the market as a whole. The relationship demonstrates that diversification—including within the portfolio different types of investments— often reduces risk. Although questions have been raised about some assumptions underlying MPT, the theory is often used by financial investment managers to help make investment allocation decisions.

From the standpoint of project prioritization, MPT has a major limitation—The theory is designed to be used for optimizing portfolios of financial securities, such as stocks and bonds, not projects. There are important differences between financial and project investments. With a stock portfolio, for example, an investor can choose any level of investment in each security. Project portfolios, on the other hand, typically require the organization to choose to do, or not to do, each project. Tool providers that claim their tools are based on MPT are likely being disingenuous—Any techniques provided for optimizing project portfolios probably have nothing to do with MPT. However, there are some situations where MPT can appropriately be applied to value projects and optimize project portfolios (see below).

In MPT, the uncertain return from an investment is represented as a random variable characterized by its mean (average or expected value) and standard deviation (a measure of variability about the mean). If the investment is a financial security (e.g., a stock) these statistics can be estimated from historical data. The standard deviation is interpreted as a measure of the risk associated with the investment. The theory assumes that, among those portfolios with a given expected return, the most attractive is the one having the least risk.

MPT shows that a key determinant of the risk of a portfolio is the degree of correlation among the individual investments; that is, the extent to which their prices (or returns) tend to move together. For example, an up tick in oil prices is often good for oil companies, but bad for airlines. Thus, oil stocks and airline stocks tend to be negatively correlated. Diversifying by including negatively correlated (or even uncorrelated) investments in a portfolio tends to decrease portfolio risk.

The risk that cannot be avoided, no matter how much you diversify, is referred to as systematic or market risk. Market risk stems from correlations among securities that arise because there are economy-wide perils that impact all businesses. MPT shows that the contribution of an investment to the risk of a well-diversified portfolio is determined not by its riskiness in isolation, but rather by its market risk (measured using a coefficient called beta). A related result is that, on average, the market provides higher expected returns for investments with higher market risks. Note that the way that the market increases the return on riskier assets is for the asset to trade at a lower price than does a similar but lower risk asset.

One situation in which MPT could be applied to projects is the following. Suppose a project would create a factory for producing some commodity traded in the marketplace, such as corn or gasoline. Commodities are similar in some respects to financial instruments; their prices are determined by large numbers of buyers and sellers, and historical data is available for estimating the statistics needed for MPT. The value of the proposed factory will largely be determined by the prices for the commodity that prevail during the lifetime of the factory. MPT could be used to understand how market risks impact the value and risk of the project and project portfolios that include the project. As another potential application, MPT can sometimes be used to compute a risk-adjusted discount rate for net present value (NPV) analysis or other techniques, such as economic value added (EVA), that account for the cost of capital.

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.