A dollar that doubles, or a million that grows a fifth
Here is a choice. Project one: put in $1 today and get $2 back in a year. Project two: put in $1 million and get $1.2 million back in a year. You can do only one.
The first has an internal rate of return of 100 percent. The second, 20 percent. Ranked by IRR, the dollar wins by a mile. Ranked by net present value at a 10 percent cost of money, the dollar adds about 82 cents of value and the million adds about $90,900. Nobody in their right mind takes the dollar.
Put that starkly, the answer is obvious. It stops being obvious inside a company, where the two projects arrive in different decks, from different teams, months apart, and the one with the higher percentage gets the better headline. That is the whole NPV-vs-IRR debate in one picture: a rate of return is a speed, and a company does not spend speed. It spends money.
IRR tells you how fast the money grows. NPV tells you how much of it you end up with.
NPV vs. IRR: two projects, two rankings
Two projects, each costing money today and paying back in one lump in a year you choose. The defaults are an illustration: both cost $1 million; one returns $1.5 million in two years, the other $2.6 million in six. Enter your own cost of capital and projects, and watch which measure prefers which.
The return the money could earn elsewhere at similar risk.
10%Extra value B adds over A, today: $227,963
A thinking aid from Stratrix. The arithmetic is exact for one cost today and one payment in a chosen year; real projects pay over many years, so use the Excel steps on this page for those. Not investment advice.
Results
| Value | |
|---|---|
| Project A: internal rate of return | 22.5% |
| Project B: internal rate of return | 17.3% |
| IRR of B less IRR of A (points) | -5.21% |
| Project A: net present value | $239,669 |
| Project B: net present value | $467,632 |
| Extra value B adds over A, today | $227,963 |
Your inputs
| Value | |
|---|---|
| Your cost of capital (discount rate) | 10% |
| Project A: cost today | $1,000,000 |
| Project A: payment back | $1,500,000 |
| Project A: year the payment arrives | 2 |
| Project B: cost today | $1,000,000 |
| Project B: payment back | $2,600,000 |
| Project B: year the payment arrives | 6 |
About the method
- IRR here is the yearly rate at which the cost grows into the payment: (payment ÷ cost) to the power 1 ÷ years, less 1.
- NPV is the payment discounted at your cost of capital for the number of years until it arrives, less the cost today.
- When the IRR gap and the value gap point in opposite directions, the measures disagree; NPV is the one measuring value added.
Get the full report: PDF report, Word (editable), Excel
Two measures built from the same cash
Both start from the same list: what a project costs, and the cash it returns in each future year. Net present value discounts each future amount back to today at the company's cost of capital (the return investors could earn elsewhere at similar risk) and subtracts the cost. A positive NPV means the project adds that much value today. Internal rate of return asks a different question: at what discount rate would the NPV be exactly zero? It is the project's own rate of growth, compared afterward against a hurdle rate.
For a single project, judged yes or no, the two usually give the same answer: if the IRR beats the cost of capital, the NPV at that cost of capital is positive. The trouble starts when you have to choose between projects, or when the cash flows change sign more than once.
Three ways the two measures part company
Scale. A percentage has no size. A 100 percent return on a dollar beats a 20 percent return on a million by IRR and loses by every measure a shareholder cares about. Whenever two options differ in size, the IRR comparison is close to meaningless on its own.
Timing. In the calculator's default pair, the quick project returns 22.5 percent a year and the slow one 17.3 percent. IRR picks the quick one. But at a 10 percent cost of capital, the quick project adds about $240,000 of value and the slow one about $468,000. The slow project compounds a larger sum for longer, and IRR does not count how long the money stays at work. The ranking flips at a discount rate of about 14.7 percent: below it the slow project is worth more, above it the quick one is. Which one is right depends on the company's real cost of capital, which is exactly the number IRR leaves out.
Cash flows that change sign. A project that costs money up front, earns for some years and then costs money again at the end (a mine that must be restored, a plant that must be decommissioned) can have two IRRs, or none at all. NPV still gives one answer at any discount rate.
| Situation | IRR says | NPV says | Trust |
|---|---|---|---|
| One project, yes or no, normal cash flows | Accept if IRR beats the hurdle | Accept if NPV is positive | Either; they agree |
| Two projects of very different size | Prefers the higher percentage | Prefers the larger value added | NPV |
| Same size, different timing | Prefers the faster payback | Depends on the cost of capital | NPV, at the right rate |
| Cash flows change sign more than once | May give two answers, or none | One answer at any rate | NPV |
| Capital is limited and projects compete for it | Not designed for it | Rank by NPV per dollar invested | NPV per dollar |
Why IRR stays popular anyway
If NPV is the better guide, why does IRR survive in every investment committee? Because a percentage travels. It needs no explanation, it compares across projects of any size without anyone checking the size, and it lets a team avoid arguing about the cost of capital, which is the one number every project's NPV depends on. When John Graham and Campbell Harvey surveyed 392 chief financial officers, 75.7 percent said they always or almost always use IRR, a hair more than the 74.9 percent who said the same of NPV 1.
There is nothing wrong with showing the IRR. The mistake is letting it choose between projects. On that, the textbooks do not hedge: when projects compete, "the rule is to take the project with the highest NPV" 4. Show both, decide on NPV, and when the two disagree, say so out loud and explain why: it is almost always scale or timing, and saying which one is the most useful sentence in the business case.
In Excel: three functions and one trap
- Lay out the cash flows by yearYear 0 (today) in the first column, then one column per year. The year 0 amount is negative: it is what you spend.
- NPV, with the trap avoidedExcel's NPV function treats its first value as arriving at the end of the first period 2. So leave year 0 out of the function and add it separately: =NPV(rate, year1:yearN) + year0. Putting year 0 inside the function discounts today's spending by a year and overstates the result.
- IRR=IRR(year0:yearN) returns the rate at which the NPV of all the flows, year 0 included, is zero 3. If the flows change sign more than once, try different guesses (the function's optional second argument) and see whether you get more than one answer.
- XNPV and XIRR for real datesWhen cash does not arrive once a year on the same day, use =XNPV(rate, values, dates) and =XIRR(values, dates), which discount each amount by its actual date.
What to put in the business case
- Lead with NPV, at a cost of capital someone has signed off on. State the rate and where it came from.
- Show IRR beside it, labeled as the rate at which the project breaks even, not as the reason to do it.
- If you are comparing options, show the NPV of each and the rate at which their ranking flips. A board can judge whether its cost of capital is near that rate.
- Say which assumption the NPV turns on. A sensitivity analysis ranks them; the top one is usually worth more attention than the choice of measure.
- If capital is scarce, rank by NPV per dollar invested, then fund down the list until the money runs out.
Questions
- What is the difference between NPV and IRR?
- NPV is the value a project adds today: its future cash discounted at the cost of capital, minus its cost. IRR is the discount rate at which that NPV is zero, the project's own rate of return. NPV is an amount of money; IRR is a percentage.
- Which is better, NPV or IRR?
- For choosing between projects, NPV, because it measures the value added in money. IRR can favor a small project with a high percentage over a large one that adds far more value, and a quick payback over a slower project that is worth more at the company's cost of capital.
- Why do NPV and IRR give different rankings?
- Usually because the projects differ in size or in timing. IRR ignores size, and it does not account for how long money stays invested at the company's cost of capital. Cash flows that change sign more than once can also give several IRRs.
- Can a project have two IRRs?
- Yes. If the cash flows change from negative to positive and back to negative, for example a project with a large cleanup cost at the end, there can be more than one discount rate at which the NPV is zero. NPV still gives a single answer at any given rate.
- How do you calculate NPV in Excel?
- Use =NPV(rate, year1:yearN) + year0. Excel's NPV function assumes the first value arrives at the end of the first period, so the money spent today goes outside the function.
Sources
- Graham, J. R. and Harvey, C. R. (2001). The theory and practice of corporate finance: evidence from the field. Journal of Financial Economics 60, 187–243.
- Microsoft (2024). NPV function. Microsoft Support.
- Microsoft (2024). IRR function. Microsoft Support.
- Dahlquist, J. and Knight, R. (2022). Principles of Finance, section 16.5: Choosing between projects. OpenStax.
For information and strategic analysis only, not investment advice. Stratrix does not recommend buying, selling or holding any security.
Tools to try
- Ansoff move finderFramework: Ansoff Matrix
- Business model canvas builderFramework: Business Model Canvas
- Change readiness checkFramework: Kotter's 8-Step Change Model
- Five Forces scorerFramework: Porter's Five Forces
- SMART goal builderFramework: SMART Goals
- SWOT builderFramework: SWOT Analysis
More from Stratrix
“What did the decision you did not take actually cost?”
One decision per pack: the framework, a working model, the sourced cases and the board material.
See the pack — $499 →Test a decision, free
Set a live decision against the companies that made it before you.
AcquisitionsThe Distance TestDivestituresThe Divestiture TestDiversificationThe Diversification TestMake or buyThe Make-or-Buy TestStratrix Team
Every Casebook, three Strategy Packs a seat each year and Ask Stratrix, for teams that face these decisions all year.
$1,999 a year · $499 a seatWhat is included →Weekly email
Stratonomics, every Tuesday
One data story every Tuesday: what happened, what it means, and the figures you can quote. Free.