What Is IRR?Internal Rate of Return Explained
The discount rate that makes an investment's NPV exactly zero — and the single most-cited number in capital budgeting decisions.
Table of Contents
What Is IRR?
Every investment involves an outlay today in exchange for cash flows in the future. The question is whether those future flows — properly discounted for the time value of money — are worth more than what you paid. The Internal Rate of Return answers a more specific version of that question: at what discount rate does the investment exactly break even on a present-value basis?
The word "internal" matters here. It signals that the rate is derived entirely from the investment's own cash flows — not from external market conditions, benchmark rates, or the investor's preferences. Two analysts looking at the same project will compute the same IRR, even if their required returns differ sharply.
Think of IRR as the implicit annual return promise embedded in an investment. When you buy a machine that costs ₹24,50,000 today and generates operating savings over five years, the IRR is the annual return rate those savings are delivering on the capital deployed. If the IRR works out to 16.4%, the investment is effectively paying you 16.4% per year — compounded annually — before returning your original capital at the end of the project's life.
IRR doesn't tell you whether an investment is good. It tells you the return rate it offers. Whether that rate clears your hurdle — that's the actual decision.
You then compare that rate to your cost of capital — the minimum return you require to compensate everyone who funded the investment. If IRR exceeds your hurdle rate, the project creates value. If it falls short, it destroys value. That comparison is the entire basis of the IRR decision rule, and we'll walk through it with specific numbers shortly.
The Math: How IRR Connects to NPV
To understand IRR precisely, start from Net Present Value. NPV is the sum of all an investment's cash flows, each discounted back to today at a chosen rate r:
CF₀ is the initial investment (negative), CF₁ through CFₙ are future cash flows, r is the discount rate, and n is the number of periods.
When r is low, future cash flows lose little value in discounting, so NPV stays high. Raise r, and those same flows shrink in present-value terms. At some specific value of r, the NPV hits exactly zero. That value is the IRR.
Solve for IRR — the discount rate that drives NPV to zero. For projects longer than two periods, there is no algebraic shortcut; IRR must be found by iteration.
There is no formula you can rearrange to isolate IRR directly for multi-period projects. The equation is a polynomial of degree n, which means it must be solved numerically — through trial and error, linear interpolation, or a spreadsheet function. This isn't a flaw in the concept; it's simply the nature of polynomial equations. In practice, Excel resolves it in milliseconds.
NPV asks: "Given a required return, how much value does this investment create?" IRR asks: "What return rate does this investment actually offer?" They use identical math — just with a different variable held fixed. When you calculate NPV, you choose r. When you calculate IRR, you solve for r.
Worked Example: Project Meridian
Northgate Manufacturing is evaluating a new precision cutting machine. The machine costs ₹24,50,000 upfront and is expected to generate operating cost savings over five years. The finance team has projected the following cash flow schedule:
| Initial Outflow | |
| Year 0 — machine purchase | −₹24,50,000 |
| Operating Cash Inflows | |
| Year 1 | ₹6,80,000 |
| Year 2 | ₹8,10,000 |
| Year 3 | ₹8,90,000 |
| Year 4 | ₹7,60,000 |
| Year 5 | ₹6,20,000 |
| Total undiscounted inflows | ₹37,60,000 |
| Net undiscounted gain | ₹13,10,000 |
The undiscounted net gain looks solid, but a rupee received in Year 5 is worth less than a rupee received today. We need to find the rate at which the present value of all inflows exactly equals the ₹24,50,000 outlay.
Step 1 — Trial at 16%. Discount each year's cash flow back to Year 0 using a 16% rate:
| Present Value of Each Cash Flow | ||
| Year 1: ₹6,80,000 ÷ 1.16¹ | ₹5,86,207 | |
| Year 2: ₹8,10,000 ÷ 1.16² | ₹6,01,980 | |
| Year 3: ₹8,90,000 ÷ 1.16³ | ₹5,70,313 | |
| Year 4: ₹7,60,000 ÷ 1.16⁴ | ₹4,19,757 | |
| Year 5: ₹6,20,000 ÷ 1.16⁵ | ₹2,95,193 | |
| Sum of PV of inflows | ₹24,73,450 | |
| NPV at 16% | +₹23,450 | |
NPV is positive (+₹23,450) at 16%. That means the investment delivers more than a 16% return — so the IRR must be higher than 16%. We need to raise the discount rate.
Step 2 — Trial at 18%. A higher rate reduces present values across the board:
| Present Value of Each Cash Flow | ||
| Year 1: ₹6,80,000 ÷ 1.18¹ | ₹5,76,271 | |
| Year 2: ₹8,10,000 ÷ 1.18² | ₹5,81,636 | |
| Year 3: ₹8,90,000 ÷ 1.18³ | ₹5,41,695 | |
| Year 4: ₹7,60,000 ÷ 1.18⁴ | ₹3,92,006 | |
| Year 5: ₹6,20,000 ÷ 1.18⁵ | ₹2,71,016 | |
| Sum of PV of inflows | ₹23,62,624 | |
| NPV at 18% | −₹87,376 | |
NPV turns negative (−₹87,376) at 18%, which means 18% is too high — we've over-discounted. The zero-NPV crossover falls between 16% and 18%.
Step 3 — Linear interpolation. With NPV = +₹23,450 at 16% and NPV = −₹87,376 at 18%, we approximate the exact zero-crossing:
IRR ≈ 16% + [23,450 / (23,450 + 87,376)] × 2% = 16% + 0.42% ≈ 16.4%
Discounting the five cash flows at 16.4% produces a PV of inflows of approximately ₹24,50,520 — within ₹520 of the ₹24,50,000 outlay. The tiny residual is rounding from the two-point interpolation. Excel's =IRR() function iterates to 16.41% with far greater precision.
Project Meridian's IRR is approximately 16.4%. Whether Northgate should proceed depends on how that compares to their cost of capital — the topic of the next section.
How to Calculate IRR in Excel (and Google Sheets)
Manual interpolation is useful for understanding the mechanics, but in practice no analyst solves multi-period IRRs by hand. Excel and Google Sheets both include built-in IRR functions that run the same iterative process thousands of times in under a second.
Set up your cash flow column
In a single column — say B2 through B7 — enter your cash flows in chronological order: Year 0 in B2, Year 1 in B3, down to Year 5 in B7. The Year 0 value must be entered as a negative number; it represents your outflow.
Apply the IRR function
In any empty cell, type =IRR(B2:B7) and press Enter. Excel returns a decimal (e.g., 0.1641). Format the cell as Percentage to display 16.41%. The function assumes all cash flows are spaced exactly one period apart.
Use XIRR for irregular dates
If cash flows occur on specific calendar dates rather than at neat year-end intervals, use =XIRR(values, dates). Enter the cash flow amounts in one column and the corresponding dates in a parallel column. XIRR handles monthly, quarterly, or completely uneven spacing correctly — the standard =IRR() function assumes equal intervals and will give a wrong answer if applied to irregular cash flows.
Provide a guess if the function errors out
If =IRR(B2:B7) returns #NUM!, add a starting guess: =IRR(B2:B7, 0.1). By default, Excel begins iterating from 10%; an unconventional cash flow pattern may cause convergence to fail without a reasonable starting point. Try guesses close to your expected answer.
If your cash flows are monthly, =IRR() returns a monthly IRR. Comparing a monthly IRR directly to an annual WACC will wildly overstate the project's attractiveness. Annualise monthly IRR with: (1 + monthly_IRR)^12 − 1. This conversion step is commonly missed.
The IRR Decision Rule: Comparing IRR to WACC
An IRR figure in isolation means nothing. It only becomes a decision tool when compared against a benchmark — and that benchmark is typically the company's Weighted Average Cost of Capital (WACC) or an explicitly set hurdle rate.
WACC represents the blended cost of all the capital a company has deployed — equity (the return shareholders require) and debt (interest expense, after tax). It is the minimum return any project must earn to pay back everyone who funded it. A project that earns exactly WACC creates zero net value. One that earns more creates value; one that earns less destroys it.
Applying this to Project Meridian: Northgate's WACC is 12%. Project Meridian's IRR is 16.4%. Since 16.4% clears the 12% hurdle by 4.4 percentage points, the project should be accepted. That 4.4pp spread is also the project's margin of safety — how much the actual cash flows could disappoint before the investment turns value-destructive.
| Project IRR | 16.4% |
| Company WACC | 12.0% |
| IRR spread over WACC | +4.4 percentage points |
| NPV discounted at 12% WACC | +₹2,71,046 |
| Decision | Accept ✓ |
Both signals — positive NPV and positive IRR spread — confirm the same decision. This is always the case for conventional projects: when IRR exceeds WACC, NPV at WACC is positive.
In private equity and infrastructure finance, hurdle rates are often set explicitly above WACC to reflect deal-specific risk or return targets. A PE fund might require a minimum 20% IRR before calculating carried interest — any project below that threshold doesn't cross the performance threshold regardless of its absolute dollar return. The principle is the same; the benchmark simply differs from a corporate WACC.
IRR vs NPV: When the Two Metrics Disagree
For a single, standalone project with conventional cash flows — one outflow at the start, inflows only after — IRR and NPV always deliver the same accept/reject signal. But when you are ranking mutually exclusive projects, they can give you different orderings. And one of those orderings will be wrong.
The root problem is scale blindness. IRR is a percentage rate — it has no memory of how much capital is in play. A project that converts ₹1,000 into ₹1,900 in one year has a 90% IRR. A project that converts ₹1,00,00,000 into ₹1,14,00,000 in one year has a 14% IRR. IRR says the first project is more than six times better. NPV at 10% says the second project created ₹31,48 more absolute value for shareholders. For a firm whose objective is to maximise shareholder wealth — not return rates — NPV wins that argument every time.
| Dimension | IRR | NPV |
|---|---|---|
| What it measures | Rate of return (%) | Absolute value created (₹) |
| Scale sensitivity | Scale-blind — ignores investment size | Scale-aware — larger project, larger NPV |
| Reinvestment assumption | Assumes interim flows reinvested at IRR — often unrealistic for high-IRR projects | Assumes reinvestment at WACC — more realistic |
| Multiple solutions possible? | Yes — if cash flows change sign more than once | Never — always one unique NPV |
| Mutually exclusive projects | Can give incorrect ranking due to scale and timing effects | Always gives correct value-maximising ranking |
| Communication | Highly intuitive — "we achieved a 22% return" | Less intuitive — "NPV is ₹4.2 crore" |
| Recommended for ranking? | No — unreliable for mutually exclusive choices | Yes — NPV is the theoretically correct ranking metric |
Timing is the second source of disagreement. Consider two projects with the same NPV at your current WACC, but one pays most of its returns early while the other pays more, later. At a low discount rate, the back-loaded project might have a higher NPV (the time-value penalty is mild). At a higher rate, the front-loaded project takes over. The discount rate at which their NPVs intersect is called the crossover rate (or Fisher Rate). Below the crossover, NPV and IRR can rank the two projects differently.
NPV measures how much richer shareholders become in real currency — which is the actual objective of value-maximising firms. IRR measures return as a percentage, which only has meaning relative to the scale of capital. When they disagree on mutually exclusive projects, NPV gives the correct ranking every time.
Three Times IRR Will Mislead You
IRR's popularity comes from its intuitive, percentage-based framing. That same intuition also makes it easy to misapply. These three failure modes are what separate analysts who use IRR well from those who use it blindly.
The Scale Problem
Project A (IRR 28%, requires ₹50 lakh) is better than Project B (IRR 19%, requires ₹5 crore).
Project B could generate many times more absolute value — even at a lower IRR — simply because far more capital is compounding. A 19% return on ₹5 crore creates ₹95 lakh of gain; a 28% return on ₹50 lakh creates ₹14 lakh. When choosing between mutually exclusive projects, rank by NPV. IRR ranking is only reliable when projects are identical in scale and duration.
The Reinvestment Rate Assumption
A 30% IRR means the investor earns 30% per year on their capital over the life of the project.
IRR implicitly assumes that every intermediate cash flow — every rupee received in Year 1, Year 2, and so on — is reinvested at the IRR itself (30% in this case). For most businesses, that's unrealistic: their next-best opportunity earns something closer to their WACC. The true compounded return on capital is lower than the stated IRR whenever the reinvestment rate is lower than the IRR. MIRR (covered in the next section) corrects this by making the reinvestment rate explicit.
The Multiple IRR Problem
Every project has exactly one IRR, and Excel's =IRR() function will always find it.
Descartes' Rule of Signs states that a polynomial has at most as many positive roots as it has sign changes. The IRR equation is a polynomial — if your cash flow series changes sign twice (outflow at Year 0, inflows in Years 1–4, then a large cleanup cost in Year 5), the equation can have two mathematically valid solutions. Both are technically "correct," but neither can be used as a reliable decision criterion. When you have multiple sign changes, switch to NPV — evaluated across a range of discount rates — or use MIRR.
The multiple IRR problem is not theoretical. Mining projects with large decommissioning costs, nuclear plants with end-of-life decommissioning, and real estate developments with phased construction — all involve significant cash outflows partway through the investment's life, creating the sign changes that can produce two or more valid IRRs.
MIRR: The Fix for IRR's Biggest Flaw
The Modified Internal Rate of Return (MIRR) addresses the reinvestment rate problem directly. Where IRR conflates the rate at which you fund the initial investment with the rate at which you reinvest intermediate cash flows, MIRR separates them — letting you assign each independently.
The MIRR calculation works like this: take all positive cash flows (the inflows) and compound them forward to the end of the project at a specified reinvestment rate, usually WACC. Take all negative cash flows and discount them to Year 0 at a specified financing rate, again usually WACC. Then compute the IRR on just those two terminal values — the future value of inflows at the end, and the present value of outflows at the start.
In Excel: =MIRR(values, finance_rate, reinvest_rate). A common starting point is to use WACC for both rates.
Returning to Project Meridian: its IRR was 16.4%, and Northgate's WACC is 12%. When you apply MIRR using 12% as both the financing and reinvestment rate, the result is roughly 14.1% — meaningfully lower than 16.4%, because the assumption that interim flows can be reinvested at 16.4% has been replaced with the more conservative 12%. Both still clear the 12% WACC hurdle, so the accept decision holds — but MIRR gives a more honest picture of the effective return.
MIRR is most valuable in two specific situations: when a project has a very high IRR (making the reinvestment assumption implausibly optimistic), and when cash flows change sign more than once (eliminating the multiple-IRR problem, since MIRR always produces a unique answer). For routine capital budgeting with conventional cash flows and IRRs in the 12–25% range, the IRR–MIRR difference is typically small enough that both point to the same accept/reject decision.
One practical advantage of MIRR over IRR is that it always produces exactly one result, regardless of how many sign changes the cash flow series has. If a project has environmental remediation costs at the end and Excel returns two IRRs (or a #NUM! error), use =MIRR(values, WACC, WACC) for an unambiguous, defensible answer.
Key Takeaways
- IRR is the discount rate that makes an investment's NPV equal to zero — it is the annual return rate an investment implicitly offers, derived entirely from its own cash flows.
- The decision rule: IRR > WACC = Accept. When an investment's return rate clears the cost of capital, it creates shareholder value. When it falls short, it destroys it.
- IRR and NPV agree on standalone projects but can conflict on mutually exclusive ones. When they give different rankings, follow NPV — it measures absolute value creation in currency, which is what shareholders care about.
- IRR has three known failure modes: scale blindness, an optimistic reinvestment rate assumption, and multiple valid solutions from cash flows with more than one sign change. Know all three before trusting an IRR figure alone.
- MIRR corrects the reinvestment flaw by letting you set a realistic reinvestment rate separately from the financing rate. Use it for high-IRR projects or any cash flow series with multiple sign changes.
- In Excel: =IRR() for equal-period conventional cash flows; =XIRR() for irregular dates; =MIRR() when you need explicit control over the reinvestment rate. All three are standard tools in financial modelling.
Quick Quiz
Four questions to check your understanding. Click an answer to reveal the explanation.
1. What is the Internal Rate of Return by definition?
2. A company's WACC is 13.5% and a proposed project has an IRR of 11.2%. What should the company do?
3. A mining project has the following cash flow pattern: large outflow at Year 0, positive inflows from Years 1–6, then a large negative outflow in Year 7 (site remediation). How many valid IRRs might this project have?
4. Project X has IRR 31% and requires ₹8 lakh, creating NPV of ₹1.2 lakh at 12% WACC. Project Y has IRR 22% and requires ₹80 lakh, creating NPV of ₹9.6 lakh at 12% WACC. The two are mutually exclusive. Which should the firm choose?