NPV tells you how much money — in today's dollars — your business generates above the minimum you require. IRR tells you the rate at which that investment yields. Both measure the same thing from opposite angles. And sometimes they fight.
Two indicators, one project
When you evaluate an investment project, two numbers always show up: NPV and IRR. Both measure whether the investment is worth it, but they answer different questions.
Before going further, you need a well-defined MARR — both indicators rely on it, and a poorly calibrated one breaks both. If you already have it, load your project and look at both numbers before reading on.
NPV says how much value your project creates, in dollars. IRR says how fast it's creating that value.
What each one measures
NPV — Net Present Value
NPV answers: how much value does this investment create? It's an absolute number in dollars. An NPV of $50,000 means the project generates $50,000 more than investing that same money at the MARR.
NPV = −I₀ + Σ [ CFₜ / (1 + MARR)ᵗ ]
Criterio de decisión
- NPV > 0
- Accept — the project creates value
- NPV = 0
- Indifferent — exactly matches the MARR
- NPV < 0
- Reject — the project destroys value
IRR — Internal Rate of Return
IRR answers: at what interest rate does this investment yield? It's a percentage. If your IRR is 35%, every dollar invested returns 35% per year. To accept the project, IRR must beat MARR.
0 = −I₀ + Σ [ CFₜ / (1 + IRR)ᵗ ]
Donde
- IRR
- Internal Rate of Return — the r that makes NPV = 0
Criterio de decisión
- IRR > MARR
- Accept — the investment beats the minimum required
- IRR < MARR
- Reject — the investment doesn't cover opportunity cost
When they disagree
When you evaluate one project alone, NPV and IRR agree: if one approves, the other does too. Conflict shows up when you compare two alternative projects. Project A may have higher NPV but lower IRR than Project B.
| Project | Investment | NPV | IRR |
|---|---|---|---|
| A — Large location | $1,000,000 | $450,000 | 28% |
| B — Small location | $200,000 | $180,000 | 55% |
A creates more absolute value: $450,000 vs. $180,000. B yields at a higher rate: 55% vs. 28%. Which do you choose?
- If you have the capital and can invest in A, choose A: you maximize value created.
- If capital is scarce and you can reinvest B's surpluses at similar rates, B may be better.
The golden rule: NPV wins
When IRR and NPV conflict for mutually exclusive projects, financial theory says NPV wins.
IRR implicitly assumes flows get reinvested at the same rate as the IRR. In practice, that almost never happens — no one consistently earns 55% on a parallel investment. NPV uses MARR as the reinvestment rate, which is realistic.
IRR is easier to communicate to non-financial partners. "This business yields 45% per year" sounds better than "NPV is $380,000." In your pitch, use both. In your decision, win with NPV.
Payback: the third vertex
NPV and IRR cover the "how much" and the "at what rate." The "when" is missing. That question is answered by Payback Period: how many years until you recover the initial investment.
A project can have positive NPV and high IRR but an 8-year Payback. If your liquidity can't sustain that timeframe, no number justifies the decision. The three metrics coexist — each answers what the others don't.
Cases where IRR fails
- Flows with multiple sign changes (partial financing, large reinvestments in intermediate years). There can be multiple valid IRRs, or none. NPV always gives a unique answer.
- Projects of very different scale. A 60% IRR on $10,000 is less useful than a 25% IRR on $1,000,000.
- Very different horizons. Comparing a 2-year project to a 10-year one by IRR is confusing — annualized rates aren't equivalent.
Practical summary
| Situation | Use NPV | Use IRR |
|---|---|---|
| Evaluating a single project | ✓ | ✓ |
| Comparing projects of different scale | ✓ (prefer) | ⚠️ careful |
| Projects with irregular flows | ✓ | ⚠️ may fail |
| Presenting to investors/partners | ✓ as backup | ✓ more intuitive |
| Flows with multiple sign changes | ✓ | ✗ may be invalid |
Before choosing a project, also run a sensitivity analysis — what happens to your NPV and IRR if price drops 10%, costs rise 15%, demand falls short. Without that step, the numbers are a flat snapshot.
Calculate both in five minutes
The good news: you don't have to choose which to calculate. Factibilidad.io gives you NPV, IRR, Payback, and sensitivity in a single run, alongside the full scenario for your project.
Apply what you just read — load your project and compare them side by side.
¿Te fue útil? Compartilo