The Complete Guide to Profitability Index in the Real World

The profitability index is the ratio of the present value of inflows to the initial investment. It is NPV rescaled: PI = 1 + NPV/Investment. A zero-NPV project has PI exactly 1.0; each dollar of NPV per dollar of investment adds 0.01 to the index.

Because PI is a ratio, it is scale-free — a $5 million project and a $500 million project can be compared directly. That is precisely its power and, as with IRR, precisely its danger when projects differ greatly in size.

Standalone: accept any project with PI > 1.0 and reject below 1.0 — logically identical to the NPV rule. PI above 1 means inflows’ present value exceeds the outlay; value is created. At exactly 1.0 the project earns the required return and nothing more.

PI shines when the constraint is the budget rather than project quality. Ranking candidate projects by PI (bang per buck) and funding from the top until the budget exhausts approximates the maximum total NPV achievable under rationing — the classic knapsack logic of capital budgeting.

For single go/no-go decisions, PI and NPV always agree; the ratio is a monotone transform of the dollar measure. They diverge only in ranking. When two positive-NPV projects are mutually exclusive, the larger project may have both the higher NPV and the lower PI — and the right choice is the larger NPV, because wealth is measured in dollars.

PI also inherits NPV’s input problems — it is only as good as the cash flow forecasts and the discount rate. Its practical role is specific: a budget-allocation lens. Many companies compute all three metrics for every project — NPV for the decision, IRR for communication, PI for capital rationing — and let each answer the question it is built for.

The mechanics take three moves. First, discount every inflow at the hurdle rate to get the present value of the inflow stream. Second, divide that present value by the initial investment. Third, read the result: every 0.01 above 1.00 is one cent of value created per invested dollar. The computation reuses everything you know from NPV — only the final division changes.

Because the last step is a division, PI responds sharply when the denominator shrinks. Halving the investment while halving the inflows leaves NPV halved but PI unchanged — the ratio, not the dollar figure, is the lens. That is exactly why PI belongs where dollars of budget, not dollars of wealth, are the constraint.

Capital rationing is the normal state of most companies: divisions submit more positive-NPV projects than headquarters will fund. When projects are divisible, ranking by PI and funding top-down maximizes total NPV per budgeted dollar — the greedy algorithm of capital budgeting. Each accepted project is the best bang-per-buck available at that moment.

Real projects are lumpy, and lumpiness breaks pure ranking: the top-ranked $18M project may fit worse into a $20M budget than the second- and fourth-ranked projects combined. The practical protocol is PI-ranked shortlisting followed by integer optimization — test the best combinations that actually fit the budget, and confirm the winner by its total NPV. PI narrows the field; total NPV still picks the portfolio.

PI’s strengths are real: it is scale-free, intuitive, theoretically consistent with NPV for accept/reject choices, and uniquely useful under rationing. It communicates well to boards — “1.25 dollars of value per dollar invested” lands harder than an abstract NPV figure — and it normalizes comparisons across divisions with different capital intensities.

The watch-outs mirror IRR’s: as a ratio it is silent about scale, so it must never arbitrate mutually exclusive projects of different sizes. It also assumes the rationing is real — if the “budget” is a political fiction, optimizing PI optimizes the wrong constraint. And like every discounted metric, it inherits forecast risk: a precise PI on invented cash flows is precise nonsense.