Skip to content
thecalculatorstudio

IRR Calculator

IRR is the rate at which a project breaks even. Enter the investment and what comes back to see it, alongside the NPV, which is the figure that should rank your projects.

Last reviewed

$
Cash in each period
$
$
$
$

Enter money coming in as positive and any further spending as negative.

%

Internal rate of return

18.03%

against a required 10%

Clears the hurdle, NPV is $1,839 at 10%
NPV
$1,839
Total returned
$15,000
Net profit
$5,000
Payback
2.60 periods
Discounted at 10%
PeriodCash flowPresent valueCumulative
Now-$10,000-$10,000-$10,000
1$3,000$2,727-$7,273
2$4,000$3,306-$3,967
3$5,000$3,757-$210
4$3,000$2,049$1,839
NPV crossing zero
Discount rateNPV
0%$5,000
5%$3,273
10%$1,839
15%$636
18.03%-$0
20%-$382
25%-$1,251
30%-$1,999

IRR is a percentage, so it is blind to size. A 50% return on $1,000 beats a 20% return on $1,000,000 on this measure and loses badly in cash, which is why NPV is shown beside it. Rank projects by NPV.

The rate at which a project breaks even

IRR is the discount rate that makes net present value zero.

Invest $10,000 and receive $3,000, $4,000, $5,000 and $3,000 over four years. What rate of return is that? Watch NPV fall as the discount rate rises:

Discount rate NPV
0% $5,000.00
5% $3,272.56
10% $1,838.67
15% $636.11
18.028163% $0.00
20% −$381.94

That crossing point is the IRR: 18.03%. Discount the project at anything below it and it looks worthwhile; anything above and it does not.

So the decision rule is simple. If IRR beats your cost of capital, the project adds value. Set your required return in the calculator and it judges against that.

There is no formula. It has to be searched for

You cannot rearrange the NPV equation to give IRR. The rate appears inside a power in every term, and for more than two or three periods there is no algebraic solution at all.

So it is found numerically. This page uses bisection: pick a range known to contain the answer, test the midpoint, keep the half where the sign changes, repeat three hundred times. It is slower than Newton’s method and, unlike Newton’s method, it cannot wander off on an awkward series.

The check that it worked is the definition itself: NPV at the reported rate must be zero. The tests assert exactly that for every series.

IRR cannot rank projects

This is the mistake the measure invites, and it is worth being blunt about.

IRR is a percentage. Percentages are blind to size:

Project Investment Returns IRR NPV at 10%
Small $1,000 $1,500 50.00% $363.64
Large $1,000,000 $1,200,000 20.00% $90,909

The small project has an IRR two and a half times better and makes you $363 instead of $90,909. If you have the capital for the large one, the IRR ranking is actively misleading.

They can also disagree on timing. Same $10,000, same four years:

Year 1 Year 2 Year 3 Year 4 IRR NPV at 10%
Front-loaded $6,000 $4,000 $2,000 $1,000 15.86% $945.97
Back-loaded $1,000 $2,000 $4,000 $8,000 13.46% $1,031.35

IRR prefers the first; NPV prefers the second. They disagree because IRR implicitly assumes you can reinvest early cash at the IRR itself, which is usually optimistic.

Use IRR to decide whether. Use NPV to decide which.

When a project has more than one IRR

If the cash flows change sign more than once, NPV can cross zero more than once, and every crossing is a valid IRR.

Here is a clean example. Invest $1,000, receive $5,000, then pay $6,000 for clean-up or a refit:

Discount rate NPV
0% −$2,000.00
50% −$333.33
100% $0.00
150% $40.00
200% $0.00
250% −$61.22

Two exact roots. Both are mathematically correct and neither means anything on its own, “the IRR of this project is 100%, or possibly 200%” is not a useful sentence.

The calculator flags any series that changes sign more than once and tells you to judge it on NPV, which has exactly one value at your actual cost of capital.

This shape is common in practice: mining and energy projects with restoration costs, equipment with a major mid-life overhaul, property with a planned refurbishment.

What IRR quietly assumes

Three assumptions worth knowing, because none of them is stated on the result:

  • Cash is reinvested at the IRR. A project returning 30% assumes you have somewhere else returning 30% to put the proceeds. If you do not, your realised return is lower. This is why high-IRR short projects flatter themselves.
  • The timing is exact. IRR is extremely sensitive to when money arrives. A payment slipping two quarters can move the rate several points.
  • The forecast is right. Everything here is arithmetic on numbers you supplied. The rate is only as good as the cash flow estimates, and those are the part nobody can verify.

The usual fix: MIRR

The reinvestment assumption has a standard remedy, modified internal rate of return, which discounts the outflows at your cost of capital and compounds the inflows forward at a rate you actually expect to earn, then finds the rate connecting the two.

MIRR is almost always lower than IRR, and almost always more believable. It also has exactly one value even when the cash flows change sign repeatedly, which solves the multiple-root problem above. If a project’s IRR looks too good to be true, computing MIRR is the quickest way to find out.

Reading the rest of the output

  • NPV at your hurdle rate, the amount of value created, in money.
  • Payback: how long before the cumulative cash turns positive. Useful for risk, useless for profitability, and covered on the payback period calculator.
  • The sensitivity table: NPV at a spread of rates, so you can see the break-even point rather than trusting one number.

Before you rely on the number

  • Match the period to the flows. If your cash flows are quarterly, the IRR is a quarterly rate and must be annualised before comparing it to an annual cost of capital.
  • Include every cash movement, including working capital, tax and the residual value of anything you can sell at the end.
  • A high IRR on a tiny base is noise. Check the NPV before getting excited.
  • This is a model, not a measurement. Run it again with pessimistic cash flows and treat the gap as the real answer.

Frequently asked questions

What is IRR?

The discount rate at which a project’s net present value is zero, its break-even rate of return. Investing $10,000 and receiving $3,000, $4,000, $5,000 and $3,000 over four years gives an IRR of 18.03%.

How is IRR calculated?

By searching for it. There is no formula that rearranges to give IRR directly, so it is found numerically. This page uses bisection, narrowing the range until NPV is zero. The check that it worked is the definition itself: NPV at the reported rate must be zero.

What is the difference between IRR and NPV?

NPV is an amount of money; IRR is a percentage. They always agree about whether a single project is worth doing, NPV is positive exactly when IRR beats your required return, but they can disagree about which of two projects is better, because IRR cannot see scale.

Should I rank projects by IRR?

No. Use NPV. A 50% return on $1,000 has a far better IRR than a 20% return on $1,000,000, and makes you $500 instead of $200,000. IRR tells you whether a project clears your hurdle; NPV tells you how much it is worth.

Can a project have more than one IRR?

Yes, if the cash flows change sign more than once, money out, money in, then money out again for a refit or clean-up. Each sign change can produce another root, and no single IRR is then meaningful. The calculator flags this and tells you to judge on NPV instead.

What is a good IRR?

Higher than your cost of capital, and that is the only universal answer. The comparison is against what the money would otherwise earn at similar risk, not against a fixed benchmark. Set your required return above and the calculator judges against that.

Related calculators

Sources

This tool is for information only and is not financial advice. Figures are estimates, confirm anything you act on with your lender or a qualified adviser.