The Big Problem with IRR That Could Wreck Your Investment Decisions
You're reviewing two projects for your company. Even so, project B shows 18%. Project A shows an IRR of 25%. Easy choice, right? Throw some money at Project A and move on with your day Simple as that..
Except here's the thing — that 25% number might be lying to you. Not because someone fudged the numbers, but because IRR has a fundamental flaw that makes it unreliable in many real-world situations. And if you're basing major capital decisions on it without understanding this flaw, you could be leaving serious money on the table Easy to understand, harder to ignore. Practical, not theoretical..
Let me explain what that flaw is, why it matters, and how to protect yourself from it.
What Is IRR (Internal Rate of Return)?
IRR is one of the most popular methods for evaluating capital investments. At its core, it's the discount rate that makes the net present value (NPV) of all cash flows from a project equal to zero.
In plain English: IRR tells you the effective annual return a project is expected to generate. If your company's cost of capital is 10% and a project's IRR is 15%, that project looks attractive — it's returning more than it costs to finance Turns out it matters..
The formula itself is straightforward in concept:
Initial Investment = Sum of (Cash Flow / (1 + IRR)^t)
You solve for IRR, and whatever number you get is your expected return. Most financial calculators and spreadsheet software (Excel's IRR function) will calculate it for you instantly.
That's the theory. Now here's where it falls apart.
Why IRR's Biggest Disadvantage Matters
The multiple IRR problem — sometimes called the "sign change problem" — is arguably the most serious flaw with IRR as a capital budgeting tool. And it shows up more often than most people realize.
Here's what happens: when a project's cash flows change direction (from negative to positive and back to negative again), the IRR equation can produce more than one solution. Three IRRs. Two IRRs. Sometimes more.
Which one do you use for your decision? There's no clear answer, and that's the problem.
This isn't some theoretical edge case that only shows up in textbook examples, either. It happens in real projects all the time — think of mining operations that require big remediation costs at the end, or manufacturing equipment that needs a major overhaul mid-life, or any project with a lease termination buyout built in.
When Cash Flows Flip Signs
Let me give you a concrete example. Say you're evaluating a project with these cash flows:
- Year 0: -$1,000,000 (initial investment)
- Year 1: +$2,500,000 (strong first-year returns)
- Year 2: -$1,600,000 (major expense or outflow)
The cash flows go from negative to positive (good) and then back to negative (not so good). Consider this: that's two sign changes. And in this scenario, you'd get two IRRs — one around 7% and another around 133% It's one of those things that adds up. Surprisingly effective..
So which IRR is the "real" return? Neither? Both?
If you only looked at the 133% number, you'd think this was an incredible investment. But the actual economics of the project are far more complicated than either number suggests.
Why This Misleads Decision-Makers
The problem is that IRR was designed to work with a simple assumption: you invest money upfront, then receive positive cash flows over time. That's a clean, single sign change (from negative to positive), and it produces a single, useful IRR.
But when cash flows flip direction even once more — negative, positive, negative, positive — the math breaks down. The IRR equation becomes a polynomial, and polynomials with multiple sign changes can have multiple roots.
Most business people never learn this. They punch the numbers into Excel, get a single IRR (because Excel's IRR function just returns one number, usually the first one it finds), and make a million-dollar decision based on a number that might not even be the relevant return.
Honestly, this part trips people up more than it should.
So yes, the multiple IRR problem deserves the attention it gets. Even so, it's not a minor technicality. It can directly lead you to approve projects that look great on paper but deliver mediocre (or worse) actual returns.
Other Significant Disadvantages of IRR
While the multiple IRR problem is the big one, it's not the only weakness worth knowing about Easy to understand, harder to ignore..
The Reinvestment Rate Assumption
IRR assumes that every cash flow generated by a project gets reinvested at the same rate as the IRR itself. That's rarely realistic That's the whole idea..
Say a project has an IRR of 20%. The theory says your company will reinvest that Year 1 cash flow at 20% to generate the project's overall return. But in reality, your company might only be able to reinvest that money at 8% or 10% — your actual cost of capital or the return you'd get on alternative investments That's the part that actually makes a difference..
This assumption inflates the effective return of projects with early cash flows and penalizes projects that generate most of their returns later. It's a hidden bias that IRR doesn't disclose.
Ignoring Project Scale
IRR is a percentage, not a dollar amount. This means it doesn't tell you anything about how much money you're actually making — just what percentage return you're earning.
A $10,000 investment returning $2,000 (20% IRR) looks better, percentage-wise, than a $1 million investment returning $100,000 (10% IRR). But the second project puts $100,000 in your pocket, while the first puts in $2,000.
If you only look at IRR, you might keep choosing small, high-return projects while passing up the large investments that actually move the needle on your bottom line.
Not Measuring Absolute Value
Related to the scale problem: IRR doesn't tell you anything about NPV. A project can have a great IRR and still destroy value if its NPV is negative. You'd never know it from IRR alone.
This is why most finance professionals recommend using NPV as the primary decision metric and treating IRR as a supplementary measure. Practically speaking, nPV tells you exactly how much wealth the project creates in today's dollars. IRR just tells you the rate.
Common Mistakes People Make with IRR
The biggest mistake is treating IRR as a standalone decision tool. Now, it's not. It's one input among several — NPV, payback period, profitability index — that should inform your capital allocation decision.
Another common error: using IRR to compare projects of different lengths. A 3-year project with 15% IRR isn't necessarily better than a 10-year project with 12% IRR. The shorter project might require you to reinvest your money at unknown rates for seven additional years, which changes the actual comparison significantly.
People also tend to forget that IRR is just a number that comes out of a formula. It doesn't account for risk, strategic value, or operational constraints. A project with a lower IRR might be worth pursuing because it opens a new market, builds institutional knowledge, or reduces dependence on a single supplier.
Practical Tips for Using IRR Correctly
Here's what actually works:
Always calculate NPV alongside IRR. If the NPV is negative, the project destroys value — regardless of how attractive the IRR looks. Use NPV as your primary decision metric.
Check your cash flow pattern. If cash flows change signs more than once, don't rely on IRR alone. The multiple IRR problem means you can't trust a single percentage to represent the project's return Worth keeping that in mind..
Use the Modified IRR (MIRR) if you need a rate-based metric. MIRR lets you specify a reinvestment rate (usually your cost of capital) rather than assuming reinvestment at the IRR itself. It also produces a single, unambiguous result But it adds up..
Look at the actual dollar returns. Calculate how much cash the project will put in your pocket, in absolute terms. A project with lower IRR but higher total NPV might be the better choice.
Compare projects on equal footing. If you're evaluating mutually exclusive options, make sure they're comparable in terms of scale, risk, and lifespan. IRR can mislead you when comparing apples to oranges.
Frequently Asked Questions
Can IRR be negative?
Yes. This leads to if a project's cash flows (discounted appropriately) never recover the initial investment, the IRR calculation will produce a negative number. This indicates the project loses money.
What is a good IRR?
There's no universal answer. Worth adding: a "good" IRR depends on your cost of capital, the risk of the project, and what returns are available on alternative investments. Think about it: generally, IRR should exceed your cost of capital by a meaningful margin to account for risk. Many companies use a hurdle rate — a minimum acceptable return — to screen projects.
Why does Excel sometimes return an error for IRR?
Excel's IRR function requires at least one negative value (the initial investment) and one positive value (a subsequent cash flow). In real terms, if your cash flow series doesn't meet these requirements, or if the function can't find a solution, it returns an error. This often happens with unconventional cash flow patterns Worth keeping that in mind. Simple as that..
Counterintuitive, but true.
How is MIRR different from IRR?
MIRR (Modified Internal Rate of Return) addresses two weaknesses of standard IRR: the reinvestment rate assumption and the multiple IRR problem. MIRR assumes cash flows are reinvested at your cost of capital (or a specified rate) and produces a single, unambiguous result Easy to understand, harder to ignore..
Should I always use NPV instead of IRR?
Most finance professionals recommend NPV as the primary metric because it directly measures value creation in dollar terms. IRR is useful as a supplementary measure — it gives you a sense of the project's return — but it shouldn't be the sole basis for capital budgeting decisions.
The Bottom Line
IRR isn't broken. And it's a useful tool when applied correctly. The problem is that it's often applied without understanding its limitations, and those limitations can lead to bad decisions Simple as that..
The multiple IRR problem is the most serious flaw — when cash flows flip direction, a single IRR number can be meaningless or actively misleading. Combined with the reinvestment assumption and IRR's inability to capture scale or absolute value, you have a metric that needs serious caution.
Here's what I'd do: always run NPV alongside IRR. And if your cash flow pattern produces multiple IRRs, don't try to force a single number to do the work. Treat IRR as a useful signal, not the final answer. Use MIRR, or go back to NPV as your decision guide.
Your capital budget decisions are too important to trust to a single percentage.
