NPV vs IRR: Two Ways to Value an Investment

Net present value (NPV) and internal rate of return (IRR) are the two standard tools for deciding whether a stream of future cash flows is worth an upfront investment today. NPV answers “how much value does this create, in today’s dollars, at a discount rate I choose?” IRR flips the question around: “at what discount rate would this investment exactly break even?” They usually agree on which investment is better — but not always, and knowing when they diverge matters more than knowing the formulas.

How NPV works

NPV takes every cash flow an investment is expected to produce, discounts each one back to today’s value using a chosen discount rate, and sums them up, then subtracts the initial cost:

NPV = Σ [CFₜ / (1 + r)ᵗ] − Initial investment

Here r is the discount rate — usually a required rate of return or the cost of capital — and t is the time period each cash flow arrives in. A dollar received five years from now is worth less than a dollar today, both because of inflation and because that dollar could have been earning a return elsewhere in the meantime; the discount rate captures both effects in one number. If NPV comes out positive, the investment is expected to create more value than the chosen discount rate demands; negative NPV means it doesn’t clear that bar.

The discount rate is the whole ballgame here, and it’s a judgment call, not something the formula hands you — it reflects how compound interest works in reverse, and it should reflect the return you could get on an alternative investment of similar risk.

How IRR works

IRR is the discount rate that makes NPV exactly zero for a given set of cash flows — the break-even rate of return. Rather than choosing a discount rate up front, you solve for it: what rate would make this project’s discounted cash flows exactly cover its cost?

There’s no clean algebraic formula for IRR in the general case — it’s typically found by iteration (trial and error, or software solving it numerically). Once you have it, the decision rule is simple: if IRR exceeds your required rate of return (your cost of capital), the investment clears the bar; if IRR falls short, it doesn’t.

NPV vs IRR

	NPV	IRR
Answers	How much value in today’s dollars?	What’s the break-even rate?
Output	A dollar amount	A percentage
Requires choosing	A discount rate up front	Nothing — it solves for the rate
Best for comparing	Projects of different sizes	Communicating return as a single number
Reinvestment assumption	Reinvested at the discount rate	Reinvested at the IRR itself
Multiple solutions possible	No	Yes, if cash flows change sign more than once

Where they disagree: mutually exclusive projects, scale

The two methods can rank projects differently in a few specific situations. The most common is scale: a project requiring $10,000 might have a lower IRR than a project requiring $1,000, but a far higher NPV, because IRR is a percentage that ignores the absolute size of the investment while NPV captures dollar value directly. If you can only choose one project, NPV tells you which one actually adds more value; IRR alone can point you toward the smaller, “more efficient” project even when the larger one makes you more money overall.

The two methods also embed different reinvestment assumptions. NPV implicitly assumes intermediate cash flows are reinvested at the discount rate you chose — typically a conservative, realistic figure like your cost of capital. IRR implicitly assumes they’re reinvested at the IRR itself, which for a high-IRR project can be an unrealistically optimistic assumption, overstating how attractive the project really is.

Finally, cash flows that change sign more than once — an investment with a large cost midway through its life, for instance — can produce multiple mathematically valid IRRs, or none at all, making the metric ambiguous in exactly the cases where a single clean answer would be most useful. NPV doesn’t have this problem; it always produces one number for a given discount rate.

Which one to trust

For comparing mutually exclusive projects of different sizes — where you can only pick one — NPV is the more reliable metric because it measures value in actual dollars rather than a rate that ignores scale, similar to how a DCF valuation discounts an entire business’s projected cash flows into one comparable number. IRR remains genuinely useful as a communication tool — “this project returns 18%” is intuitive in a way “$340,000 in NPV” isn’t — and as a quick screening threshold against a hurdle rate. The practical approach most finance teams take is to compute both: use IRR to screen and communicate, then defer to NPV when the two disagree on ranking, the same way a bond investor weighs coupon rate against yield to maturity rather than trusting either number in isolation.

The takeaway

NPV and IRR both discount future cash flows to make them comparable to money today, but NPV reports the result in dollars at a rate you choose, while IRR reports it as the break-even rate itself. They agree most of the time — but scale differences, reinvestment assumptions, and sign-changing cash flows can make them disagree on which investment is better, and when that happens, NPV’s dollar-denominated answer is the one to trust for an actual go/no-go decision.