Net present value (NPV) is generally considered the stronger capital budgeting tool because it directly measures how much wealth a project adds in dollar terms, while internal rate of return (IRR) expresses results as a percentage that can mislead you when projects differ in size, timing, or cash flow patterns. Both methods discount future cash flows to evaluate investments, but they answer fundamentally different questions. NPV tells you how much richer a project makes the firm. IRR tells you the rate of return a project earns. When the two methods disagree, NPV gives the more reliable answer.
NPV Directly Measures Wealth Creation
The core advantage of NPV is conceptual simplicity: a positive NPV means the project creates value for shareholders, and the number itself tells you exactly how much. If a project has an NPV of $267,946, that figure represents the extra wealth it generates for the company’s owners in today’s dollars. A negative NPV means the project destroys value. There’s no ambiguity in the decision rule.
This direct link to shareholder wealth is why finance theory treats NPV as the gold standard. The goal of any firm is to maximize the value of its owners’ stake, and NPV is the only metric that quantifies that goal in absolute terms. IRR, by contrast, gives you a percentage. A 30% return sounds impressive, but it doesn’t tell you whether the project adds $500 or $5 million to the firm’s value. That distinction matters enormously when you’re choosing between competing investments.
The Reinvestment Rate Problem
One of the most commonly cited differences between the two methods involves what happens to cash flows a project generates along the way. The traditional explanation goes like this: IRR assumes that intermediate cash flows get reinvested at the project’s own rate of return, while NPV assumes reinvestment at the discount rate (typically the company’s cost of capital).
If a project has an IRR of 30%, the IRR calculation implicitly assumes every dollar of cash flow generated mid-project can be reinvested at 30%. That’s a hard assumption to justify. Most companies can’t reliably find new investments that earn the same outsized return. NPV’s assumption, that interim cash flows earn the firm’s cost of capital, is far more conservative and realistic. It reflects what a company could reasonably expect to earn by paying down debt, buying back shares, or funding other typical operations.
It’s worth noting that some academics argue neither method technically “requires” a reinvestment assumption baked into the formula itself. The IRR is simply the return generated by each dollar that remains invested in the project, regardless of where interim cash flows go. But in practice, when you’re comparing projects with different timing patterns, the implicit reinvestment assumption is exactly where the two methods diverge and where IRR gets into trouble.
IRR Can Produce Multiple Answers
IRR has a mathematical weakness that NPV simply doesn’t have. When a project’s cash flows switch between positive and negative more than once (called non-conventional cash flows), the IRR equation can spit out multiple solutions. Consider a project with these cash flows: you invest $1,600 upfront, receive $10,000 in year one, then pay out $10,000 in year two. That project has two IRRs: 25% and 400%. Both are mathematically valid. Neither is useful for making a decision.
This happens because the IRR formula is a polynomial equation, and polynomials with multiple sign changes can have multiple roots. The more times cash flows flip between inflows and outflows, the more potential IRRs you get. NPV never has this problem. For any given discount rate, a project has exactly one NPV. You always get a single, clear answer.
Non-conventional cash flows aren’t rare edge cases, either. They show up in mining projects (which require reclamation costs at the end), real estate developments with phased investments, and any business with significant mid-project capital outlays.
IRR Misleads on Project Size
IRR’s percentage format creates a blind spot known as the scale problem. Imagine two mutually exclusive projects: Project A requires a $10,000 investment and has an IRR of 50%. Project B requires $500,000 and has an IRR of 25%. IRR ranks Project A first. But Project B, despite its lower percentage return, could generate far more total value for the company. NPV captures this because it measures absolute dollars. A 25% return on half a million dollars typically creates more wealth than a 50% return on ten thousand.
This size disparity problem is especially dangerous when companies use IRR as a screening tool. Small projects with high percentage returns get prioritized over larger projects that would contribute much more to firm value. IRR essentially favors efficiency over scale, which is the wrong priority when your goal is to maximize total wealth rather than percentage returns.
Timing Differences and the Crossover Rate
Even when two projects require the same initial investment, IRR and NPV can disagree if the projects generate cash at different times. One project might pay back most of its value early, while another delivers larger cash flows later. IRR tends to favor projects with faster paybacks because early cash flows, when reinvested at the (often inflated) IRR, compound more. NPV evaluates both projects at the same discount rate and focuses on which one is worth more today.
The discount rate at which NPV ranks two projects equally is called the crossover rate. Below this rate, one project has the higher NPV. Above it, the other does. The crossover rate reveals exactly where IRR rankings and NPV rankings diverge. If your company’s actual cost of capital falls on one side of that crossover point, IRR might point you toward the wrong project. NPV, because it uses your actual cost of capital as the discount rate, always identifies the project that adds the most value at the rate that matters to your firm.
When IRR Still Has Value
None of this means IRR is useless. It communicates something NPV doesn’t: margin of safety. If a project’s IRR is 22% and your cost of capital is 10%, you know there’s a 12-percentage-point cushion before the project stops creating value. That buffer is intuitive and easy to communicate to non-financial stakeholders. Executives and board members often find a percentage return easier to grasp than a dollar-denominated present value.
IRR also works perfectly well for independent, conventional projects (one upfront cost followed by a series of positive cash flows). When you’re evaluating a single investment on a pass/fail basis, IRR and NPV will always agree: if the IRR exceeds the cost of capital, the NPV is positive, and the project should be accepted.
The problems surface when you’re ranking multiple projects against each other, dealing with unusual cash flow patterns, or comparing investments of different sizes. In those situations, relying on IRR alone can lead you to pick projects that earn a high percentage on a small base while passing on projects that would create substantially more value.
MIRR: A Partial Fix
The Modified Internal Rate of Return (MIRR) was developed specifically to address IRR’s reinvestment rate flaw. Instead of assuming cash flows are reinvested at the project’s own return, MIRR lets you specify a reinvestment rate, usually the company’s cost of capital. It also eliminates the multiple-solution problem by converting non-conventional cash flows into a conventional pattern before calculating a single rate of return.
MIRR is more accurate than standard IRR and better suited to real-world conditions. It separates the cost of financing from the return on reinvested cash, which makes the result more realistic. However, it still expresses results as a percentage, so it inherits the scale problem. A small project can still show a higher MIRR than a large project that creates more total value. For that reason, MIRR is best used alongside NPV rather than as a replacement for it.
The Bottom Line on Choosing a Method
When NPV and IRR conflict, the standard recommendation in finance is to follow NPV. It avoids the multiple-solution trap, accounts for project size, handles any cash flow pattern, and directly measures the thing companies are trying to maximize: total value created. IRR and MIRR are useful supplements that express returns in a format people find intuitive, but they shouldn’t override an NPV analysis when the two disagree. The strongest capital budgeting process uses NPV as the primary decision tool and IRR as a secondary lens for communicating risk and return.