NPV vs. IRR: What's the Difference and Which Is Better?
By Markets & Investing DeskUpdated Nov 10, 2026
Key takeaways
- NPV is a dollar amount. IRR is a percentage: the discount rate at which a project's NPV equals exactly zero.
- For conventional cash flows (money out first, money in afterward), a positive NPV and an IRR above the required rate of return lead to the same accept-or-reject decision.
- IRR can mislead when cash flows change sign more than once or when projects differ in size or timing, and it has no risk adjustment until compared with a hurdle rate. NPV handles all of these.
NPV and IRR come from the same discounted cash flow math, but they answer different questions. Net present value (NPV) tells you how many dollars of value a project adds today after discounting its future cash flows at your required rate of return. Internal rate of return (IRR) tells you the percentage return the project’s cash flows generate on their own, without reference to any outside rate. When the two disagree, NPV is generally the better decision rule, because it measures the thing a business is trying to maximize: added value in dollars. IRR is useful for quick comparisons, and it is most reliable as a supporting figure alongside NPV.
How each measure works
Net present value
To calculate NPV, you estimate a project’s cash flows period by period, discount each one back to today using a rate that reflects the project’s cost of capital and risk, and add them up. The upfront investment enters as a negative cash flow at time zero.
NPV = Σ [Rₜ / (1 + i)ᵗ] − initial investment
- Rₜ = net cash flow (inflows minus outflows) in period t
- i = discount rate, the return you could earn on an alternative investment of similar risk
- t = the period number
A positive NPV means the project’s discounted inflows exceed its cost, so it earns more than the required return and adds value. An NPV of zero means it earns exactly the required return. A negative NPV means the project returns less than the discount rate; on its own projections, it destroys value and should be rejected.
Internal rate of return
IRR uses the same equation but turns it around. Instead of picking a discount rate and solving for NPV, you set NPV to zero and solve for the rate:
0 = Σ [Rₜ / (1 + IRR)ᵗ] − initial investment
IRR is usually found by trial and error or with a spreadsheet function. The result is the break-even discount rate: the cost of capital at which the project neither adds nor destroys value.
The relationship between NPV, IRR and the required return
The required rate of return (RRR), often the company’s weighted average cost of capital (WACC) adjusted for project risk, is the hurdle. For a conventional project:
| If… | Then NPV at the required rate is… | Decision |
|---|---|---|
| IRR > required rate | Positive | Accept |
| IRR = required rate | Zero | Indifferent |
| IRR < required rate | Negative | Reject |
A project whose IRR is below the WACC therefore has a negative NPV, not a positive one. The rule holds for conventional cash flow patterns; the exceptions are covered below.
Worked examples
Example 1: Calculating NPV and IRR for one project
A hypothetical company is considering a $100,000 equipment purchase that it expects to generate the following cash flows. Its required return is 10%, and cash flows are assumed to arrive at the end of each year.
| Year | Cash flow | Discount factor at 10% | Present value |
|---|---|---|---|
| 0 | −$100,000 | 1.0000 | −$100,000.00 |
| 1 | $30,000 | 0.9091 | $27,272.73 |
| 2 | $40,000 | 0.8264 | $33,057.85 |
| 3 | $50,000 | 0.7513 | $37,565.73 |
| 4 | $40,000 | 0.6830 | $27,320.54 |
The present value of the inflows is $125,216.85, so the NPV is about $25,217. The project adds roughly $25,000 of value beyond what a 10% return would deliver.
To find the IRR, try higher rates until NPV reaches zero. At 20%, NPV is about +$1,003. At 21%, it is about −$1,002. The IRR therefore sits at roughly 20.5%, comfortably above the 10% hurdle. Both measures say accept.
A related figure, the profitability index, divides the present value of inflows by the initial investment: $125,217 / $100,000 ≈ 1.25, meaning about $1.25 of present value per dollar invested.
Example 2: When projects conflict on timing
Now suppose a company can choose only one of two projects (they are mutually exclusive). Each costs $10,000, and the required return is 10%.
- Project A returns $12,000 after one year.
- Project B returns $16,000 after three years.
| Project A | Project B | |
|---|---|---|
| IRR | 20.0% | About 17.0% |
| NPV at 10% | About $909 | About $2,021 |
IRR ranks A first, NPV ranks B first. The reason is timing. Project A returns cash quickly at a high rate, while Project B earns a lower rate but on the full $10,000 for three years.
The rate at which the two projects have equal NPV is called the crossover rate. You find it by setting the two NPV equations equal: 12,000 / (1 + r) = 16,000 / (1 + r)³, which gives (1 + r)² = 1.333 and r ≈ 15.5%. Below 15.5%, Project B has the higher NPV; above it, Project A does. At the company’s actual 10% cost of capital, B adds more value, and NPV identifies that correctly.
Example 3: When cash flows change sign
Consider a hypothetical project with these cash flows:
- Year 0: −$50,000 initial outlay
- Year 1: +$115,000 return
- Year 2: −$66,000 in new costs to revamp the project
Setting NPV to zero produces two valid solutions: 10% and 20%. Both are mathematically correct IRRs, so neither one alone tells you whether to proceed.
NPV gives a clear answer once you plug in the required return. The project has a slightly positive NPV only when the discount rate is between 10% and 20% (at 15%, NPV is under $100). At any rate below 10% or above 20%, NPV is negative. Even inside that window the value added is tiny relative to the money at stake, which a single IRR figure would hide.
Multiple IRRs can arise whenever cash flows switch from negative to positive more than once, as with projects that need major reinvestment midway or carry large closing costs such as decommissioning or cleanup.
Why IRR can mislead
It ignores scale
IRR measures the rate, not the amount. A hypothetical project that turns $1,000 into $1,500 in a year has a 50% IRR and, at a 10% discount rate, an NPV of about $364. A project that turns $100,000 into $120,000 has a 20% IRR but an NPV of about $9,091. If capital is available for either, the lower-IRR project creates 25 times as much value.
It assumes reinvestment at the IRR
The IRR calculation implicitly assumes interim cash flows can be reinvested at the IRR itself. For a project with a 40% IRR, that assumption may be unrealistic. NPV assumes reinvestment at the discount rate, which is usually closer to what a company can actually earn. The modified internal rate of return (MIRR) addresses this by compounding inflows at a specified reinvestment rate and discounting outflows at a financing rate, which also produces a single answer when cash flows change sign.
It has no built-in risk adjustment
NPV builds risk into the answer through the discount rate: a riskier project gets a higher rate and a lower NPV. IRR is calculated from the cash flows alone. A 15% IRR could be excellent for a low-risk expansion and inadequate for a speculative venture, so it needs a risk-appropriate hurdle rate before it means anything.
It struggles with changing discount rates
NPV can apply a different rate to each period if the cost of capital is expected to change over a project’s life. IRR produces one rate to compare against one hurdle, which makes it awkward when the hurdle itself varies.
Which is better: NPV or IRR?
For choosing between competing projects, NPV is the more reliable rule. It is stated in dollars, accounts for project size and timing, reflects risk through the discount rate, and gives one answer regardless of how cash flows are patterned. When NPV and IRR rank mutually exclusive projects differently, the NPV ranking is the one aligned with maximizing value.
IRR still earns its place in several situations:
- Communicating returns. A percentage is easy to compare with a borrowing cost, a hurdle rate or the return on other assets.
- Uncertain discount rates. If the appropriate discount rate is hard to pin down, IRR shows the break-even cost of capital. A project with a 25% IRR stays worthwhile across a wide range of plausible hurdle rates.
- Screening independent projects. When projects do not compete for the same capital and have conventional cash flows, IRR and NPV give the same accept-or-reject answer, and IRR is a quick first filter.
Many analysts report both: NPV to make the decision, IRR to describe the return. The payback period (how long until the initial investment is recovered) ignores the time value of money and any cash flows after the payback date, so it works as a liquidity check rather than a value measure.
Both measures are only as good as the cash flow forecasts behind them, which typically start from operating cash flow estimates like those on a cash flow statement.
FAQs
How do you calculate NPV and IRR in a spreadsheet?
Use NPV(rate, cash flows from year 1 onward) and then add the year 0 cash flow separately, since the function treats its first value as arriving one period from now. IRR(cash flows from year 0 onward) returns the IRR for evenly spaced periods; XIRR handles cash flows on specific dates. Both IRR functions can return only one root, so check the cash flow signs before relying on the result.
The Bottom Line
Managers and analysts deciding where to commit capital should use NPV as the main decision rule, especially when choosing between mutually exclusive projects of different size or timing. IRR remains a useful companion figure for communicating returns, screening independent projects with conventional cash flows, and showing the break-even cost of capital when the discount rate is uncertain. Either measure depends on the quality of the cash flow forecasts behind it.
