Technical Terms Used in Project Portfolio Management (Continued)

earned value management (EVM)

A method for measuring progress on projects and indicating variances in planned accomplishments, schedule, and cost expenditures. EVM, also called earned value analysis (EVA, not to be confused with economic value added) is primarily used as a way of reporting project progress to stakeholders, and government regulations often require that contractors providing services to federal and other government agencies comply with standards for using EVM. In the context of project portfolio management, EVM provides a method for reporting progress on individual projects and for demonstrating compliance with government requirements for EVM.

The basic concept with EVM is that project work be planned, budgeted, and scheduled in time-phased, "planned value" increments. Typically, these work increments are defined in a hierarchical fashion as a work breakdown structure, but for a smaller project the work elements might simply be individual project tasks. The work elements define a schedule and cost/value baseline for the project. As project work is conducted, project value is "earned." Various indices are computed that summarize project status based on comparing earned value with planed and actual costs.

The value that is assigned to each work element is termed its planned value (PV). The PV is meant to be a weighting factor that indicates the portion of the project value that, according to the plan, will be contributed by that work element at a specified time. Alternatively, a work element's PV could be defined as the number of labor hours required, or even as a subjectively assigned number of "points."

The value of the work element is earned as the work is completed. For example, the earning rule might be that 25% of the value is earned when the task is started, and the remaining 75% is earned upon completion.

Progress against the plan is reported on a regular basis (e.g., weekly or monthly) by accumulating earned value (EV) based on the earning rules. By subtracting the value of the work performed (EV) from the value of the work that was planned (PV), a schedule variance (SV) can be computed at any point of time:

Similarly, a schedule performance index (SPI) may be computed by dividing the EV by the PV:

If the SV is greater than zero (SPI is greater than 1), the work is ahead of schedule. If the SV is less than zero (SPI is less than 1), the work is behind schedule. Schedule variances can be rolled up to any level in the work breakdown schedule to provide higher-level indicators of schedule compliance.

EVM may seem confusing because of the many acronyms that are used. Since a work element's PV is traditionally chosen to be the scheduled cost of the work, the traditional term for a work element's planned value is the budgeted cost for work scheduled (BCWS). The traditional term for earned value is the budgeted cost for work performed (BCWP). The actual cost of conducting each work element is termed the actual cost of work performed (ACWP). In this context, where value and cost are both measured in dollars, a cost variance (CV) can be computed by subtracting the actual cost of work performed (ACWP) from the budgeted cost of work performed (BCWP):

EVM defines many additional indicators of technical, schedule, and cost performance that can also be calculated, and guidance is available for interpreting and addressing the various discrepancies that the indicators may reveal.

Although EVM is a well-established and effective means for managing the completion of complex projects, it's major limitation from the standpoint of project portfolio management is that it does not provide indicators for tracking or updating the anticipated ability of the project to deliver benefits to the organization. EVM might, for example, indicate that a project is under budget, ahead of schedule, and within scope, but that project could nevertheless be in trouble with regard to achieving the benefits that motivated the decision to fund it.

economic value added (EVA®)

A financial project valuation metric and related management framework developed by consulting company Stern Steward founders Joel Stern and G. Bennett Steward III (EVA® is a registered trademark of Stern Stewart). The EVA® of a project is calculated by taking net operating profit and subtracting a charge for the capital or assets deployed. The deducted amount is the "cost of capital"—what shareholders and lenders could obtain by investing in securities of comparable risk.

EVA®, also sometimes termed earned value added, provides a useful input for prioritizing projects because it quantifies the direct financial component of project value. However, other techniques are needed to account for the indirect or non-financial components of project value. Also, depending on the characteristics of projects, it may be more convenient to account for the cost of capital using the more traditional calculation of net present value (NPV).

While several other financial metrics likewise account for the cost of capital, the appeal of EVA® is that it does so in a conceptually simple and intuitive way that is easy for non-financial managers to understand. Since EVA® starts with familiar operating profits and then deducts a charge for the capital employed, it can be interpreted simply as "net profit minus the rent."

EVA® has become popular because it highlights the importance of the cost of capital when financially evaluating projects. EVA® may show, for example, that despite increasing earnings, a project is destroying shareholder value because the cost of capital associated with the required investment is too high. By assessing a charge for using capital, EVA® forces managers to think about managing assets as well as income.

As indicated above, a major weakness of EVA® is that it fails to account for non-financial project impacts (such as improved employee knowledge) that are difficult to express in terms of incremental cash flows. Also, accounting for opportunity costs by subtracting a capital charge is conceptually simple only if project start times, durations, and spending rates aren't very important (if they are, then the NPV approach of discounting cashflows using hurdle rates is computationally and conceptually simpler). Like classic NPV, EVA® does not explicitly address cash flow uncertainties, and it can be very difficult to determine the appropriate charge for the capital used by a project.

efficient frontier

In the context of modern portfolio theory, the efficient frontier is the bounding curve obtained when portfolios of possible investments are plotted based on risk and expected return. The efficient frontier shows the investment combinations that produce the highest return for the lowest possible risk. A portfolio that is not on the efficient frontier is said to be "inefficient" because another portfolio exists that has lower risk for the same return.

In the context of project portfolio management, the efficient frontier typically refers to the bounding curve that is obtained when project portfolios (or sometimes individual projects) are plotted based on cost and some quantity that is intended to represent portfolio (or project) attractiveness (ideally, the y-axis should represent the value or worth of the portfolio to the organization). In this context, a portfolio that is not on the efficient frontier is inefficient because another portfolio exists with greater value for the same cost. Click here for a precise definition and detailed discussion.

expected commercial value (ECV)

A probability-weighted value for a project with uncertain outcomes similar to expected net present value (ENPV). As with ENPV, scenarios are defined to represent different project outcomes, and each scenario is assigned a probability. A project value is computed for each scenario. The expected commercial value is obtained by multiplying each scenario's value by the scenario probability and adding the results. Estimated commercial value is another term for ECV.

Depending on the techniques used to estimate the value of the project under each scenario (and on the techniques used to estimate the probabilities of the scenarios), ECV can be a useful way to address project uncertainties. However, as indicated below, the technique often involves simplifications that may or may not be appropriate.

Typically, ECV denotes a simplified version of ENPV often appropriate for projects that generate new products. The project is broken into stages which are represented in a decision tree. The first stage is the product development stage, where the probability of technical success is P_ts. The second stage is the product launch, where the probability of commercial success is P_cs. If D is the development cost, C is the cost of commercially launching the project, and PV is the present value of future earnings assuming a commercially successful project, then:

In reality, of course, technical and commercial success are not yes/no outcomes. There are varying degrees of technical success and, assuming the product is launched, commercial sales could be anywhere within a range of possibilities. Still, depending on the application, the simple formula may provide a sufficient approximation. More generally, because ECV is a simplified version of ENPV, it has the limitations of the more general approach (including omission of non-financial sources of project value and potential for inadequate treatment of risk).

expected internal rate of return (EIRR)

A minor modification of the internal rate of return (IRR) sometimes used to prioritize projects (such as new product development projects) whose costs and future cash flows are highly uncertain. In the formula for computing IRR, project costs are replaced by the expected value of initial-year project costs, and project cash flows are replaced by the year-by-year expected value of project cash flows. Thus, The EIRR is the solution to the equation:

In other words, to use the EIRR, alternative project cost and future cash-flow scenarios are defined. For example, the various stages and associated cash flows for the project (such as development, testing, and commercialization) may be represented in a decision tree. Probabilities are assigned to each scenario. The expected value of project costs and expected value of each year's net cash flow are computed by multiplying probabilities by cash flows and adding. The EIRR is then computed as the discount rate that equates the discounted value of expected future cash flows with the expected project cost.

When applied to multi-stage, high-risk projects, the EIRR behaves in an intuitive way. For early stage projects with a low probabilities of ultimate success, expected cash flows tend to be low so the EIRR tends to be low. However, if such a project is funded, its EIRR tends to grow (assuming initial project outcomes are successful) as project costs are sunk and early-stage failure scenarios are avoided. A late stage project (one that has successfully avoided early and middle stage risks) tends to have a very high EIRR. Because of the strong influence of project stage on the EIRR, typical advice is that project-by-project comparisons using EIRR be conducted ony for projects at the same stage of development and that separate budgets be established for funding projects within the different stages.

As a project prioritization metric, the EIRR has the advantages and disadvantages described for the IRR, plus the advantages and difficulties associated with assigning probabilities to alternative scenarios.

expected net present value (ENPV)

An enhancement of the net present value approach that explicitly addresses uncertainty. Depending on how it is applied, ENPV can produce estimates of uncertainty in the value of the overall project portfolio and adjust project value to account for risk. It can also be coupled with methods for quantifying the non-financial or indirect components of project value. It is, therefore, a useful tool for computing project and project portfolio value. However, the computations necessary to compute ENPV can be difficult. The method is best reserved for very large and risky projects.

With ENPV, rather than calculate a single time-stream of project cashflows and other project impacts, alternative scenarios are defined representing the range of possibilities. Simulation techniques are often used to generate the alternative scenarios, which may be represented in as a decision tree (a graphic structure wherein alternative sequences of choices and outcomes are displayed as branches in the tree and the various paths through the tree represent the alternative scenarios) or event tree (similar to a decision tree, but without nodes and branches representing alternative choices). Probabilities are associated to each scenario in the tree. A project NPV is computed for each scenario, and the ENPV is the probability-weighted sum of the values.

As described under net present value, selecting discount rates is often problematic. If risk is important, risk-adjusted discount rates are often recommended, with different risk-adjusted rates being appropriate for different scenarios. Alternatively, techniques based on risk tolerance can be used to account for risk (these techniques generally involve using a risk-free discount rate for computing EPNV).

In addition to the difficulties mentioned above related to selecting the discount rate, another limitation of ENPV is that historical data is generally unavailable for estimating probabilities. Thus, probabilities must typically be assigned subjectively.

expected value

Term used to represent the result of a mathematical computation performed using probabilities. Suppose there is an uncertain (random) variable X that may produce various "payoffs" (values). Suppose the possible payoffs are denoted x₁, x₂,..., x_N, and suppose that these alternative payoffs occur with probabilities p₁, p₂,...p_N, respectively. The expected value of the variable is sum of each possible payoff multiplied by its probability:

If the uncertain variable can take on a continuum of possible values (e.g., any value between 0 and 1), then its expected value is computed by weighting the possible values using the variable's probability density function and using integral calculus.

The expected value is the average return one would expect over many "trials" or opportunities for the uncertainty to occur. See expected commercial value and expected net present value for examples of measures based on expected value.