When evaluating a portfolio of candidate projects, the relationship between them affects how the comparison is set up.
Independent projects. Doing one doesn’t depend on whether you do another. You evaluate each on its own merits: any project whose expected return exceeds the MARR (or has positive PW) is accepted, regardless of what the others look like. If you have a budget constraint, you may have to ration capital — pick the best subset that fits — but absent that, independent projects are accepted or rejected one at a time.
Mutually exclusive (ME) projects. Choosing one precludes the others. Two competing designs for the same machine; three different replacements for the same aging asset; four ways to build the same bridge. You can do at most one. The evaluation reduces to “which one is best?”, not “which ones clear the threshold?”
Related but not mutually exclusive. The expected cost or benefit of one depends on whether you do another. Building a new factory might enable a second project (a logistics warehouse) at lower cost; doing both together is different than doing each in isolation. These cases are most accurately handled by treating combinations as the alternatives (do A only, do B only, do both, do neither) and evaluating each combination as a separate ME alternative.
The decision rules differ:
| Relationship | Decision rule |
|---|---|
| Independent | Accept each project whose PW > 0 (equivalently IRR > MARR) |
| Mutually exclusive | Pick the one with the highest PW (or use incremental IRR for IRR-based comparison) |
| Related but not ME | Enumerate combinations; treat as ME among combinations |
A common mistake with ME projects: picking the one with the highest IRR. IRR is not a maximisation target across ME alternatives — the project with the highest IRR may have a smaller total benefit than another with lower IRR but larger scale. The correct ME-IRR method is incremental analysis: starting from the lowest-cost project, ask whether moving up to the next costlier project earns at least MARR on the incremental investment. See Internal rate of return for the procedure.
For independent projects with budget constraints, the right framework is capital rationing — maximise total PW subject to total investment ≤ budget. This is a knapsack-style optimisation in general, though in practice a heuristic of “rank by PW/cost ratio, pick top until budget runs out” usually does fine.
For the comparison methods that operate on these project relationships, see Present worth method, Annual worth method, Internal rate of return, Comparison of alternatives.