The internal rate of return (IRR) is the annualized rate of return at which an investment’s net present value equals zero: the single discount rate that makes the present value of all cash inflows equal the cash outflows. In private equity and venture capital, IRR is the time-weighted rate a fund or deal earns on the capital still at work.
This guide covers what IRR means, the NPV-equals-zero formula and how spreadsheets solve it, a worked cash-flow example with the resulting IRR, why IRR and MOIC can disagree on the same deal, the gross-versus-net distinction, what counts as a good IRR in venture and private equity, and where IRR breaks down.
The short version:
- What it is: the annual rate that makes an investment’s net present value (NPV) equal zero.
- The formula:
0 = Σ [ CFt / (1 + IRR)^t ], solved iteratively rather than with a closed formula. - What a 20% IRR means: the investment is projected to earn about 20% per year on the capital still deployed, compounding over the hold.
- IRR vs MOIC in one line: IRR is the speed (time-weighted rate); MOIC is the size (absolute multiple). The same 3.0x can be a 24.6% IRR or a 42.2% IRR depending on when cash comes back.
- What is “good”: for venture and private equity, a high-teens to 20%+ net IRR is generally considered strong; lower-risk strategies clear at far lower rates.
What IRR means
IRR is a project’s own (or “internal”) rate of return. It is called internal because the calculation depends only on the investment’s own cash flows and excludes external factors such as the risk-free rate, inflation, or the investor’s cost of capital. The rate falls directly out of the timing and size of the money going in and coming out, and nothing else.
Intuitively, a 20% IRR means the investment is projected to earn about 20% per year on the capital that remains at work, compounding over the holding period. It is not a flat 20% of the original amount each year: IRR is time-weighted and assumes interim returns compound at that rate. The higher the IRR, the higher the projected annualized return.
Because IRR is expressed as a percentage per year, it answers a specific question, what annual rate did this earn?, rather than how many times did I multiply my money? That second question belongs to a different metric, MOIC, and the gap between the two is the heart of this page.
The IRR formula
IRR is defined as the discount rate that sets net present value to zero. Written out across every period of the cash-flow stream:
0 = Σ [ CFt / (1 + IRR)^t ] for t = 0 to n
Where:
- CF0 is the cash flow at time zero: the initial investment, entered as a negative number (an outflow).
- CFt is the cash flow in period t: distributions and the eventual exit, entered as positives (inflows).
- t is the period index, from 0 (today) through n (the final period).
- IRR is the rate being solved for.
The formula is the same net-present-value equation used to discount future cash flows, set equal to zero and rearranged to solve for the rate instead of the value. For most cash-flow streams IRR has no closed-form solution; it cannot be isolated algebraically and is instead found iteratively, by trial and error or a root-finding method such as Newton’s method, testing rates until NPV lands on zero. That iterative search is why IRR is computed in a spreadsheet rather than by hand.
Calculating IRR in Excel or Sheets
Spreadsheets ship two functions for the job. For cash flows spaced one period apart, =IRR(values) takes the ordered range of flows (the initial outflow as a negative, distributions as positives) and returns the rate. For cash flows that fall on irregular dates, =XIRR(values, dates) pairs each amount with its actual date and annualizes correctly. XIRR is the honest choice whenever flows are not evenly annual, because real distributions rarely arrive on neat anniversaries. Both implement the same NPV-equals-zero search described in Microsoft’s IRR and XIRR function references.
A worked IRR example
Consider an investor who commits $10,000,000.00 at Year 0, the only outflow. The investment returns nothing in Years 1 through 4, then pays a single exit distribution of $30,000,000.00 at the end of Year 5.
IRR is the rate r that sets the NPV to zero:
-10,000,000 + 30,000,000 / (1 + r)^5 = 0
Rearranged, this is (1 + r)^5 = 3.0, so r = 3.0^(1/5) − 1 ≈ 0.2457, an IRR of about 24.6%. This is the clean case: a single inflow against a single outflow has a closed-form answer (the n-th root of the multiple), so no iteration is needed. Most real cash-flow streams have several inflows and require the iterative search above.
Now hold the totals fixed but change the timing. Same $10,000,000.00 invested at Year 0, and the same $30,000,000.00 returned in total, but now $15,000,000.00 comes back at the end of Year 2 and the remaining $15,000,000.00 at the end of Year 5. Solving -10,000,000 + 15,000,000 / (1 + r)^2 + 15,000,000 / (1 + r)^5 = 0 iteratively gives an IRR of about 42.2%.
The investor put in the same dollars and got back the same dollars. The only thing that changed was when half the money returned, and IRR jumped from 24.6% to 42.2%. The takeaway is plain: IRR rewards getting money back sooner. Capital returned earlier can be put back to work earlier, and the metric prices that in.
Both streams returned exactly 3.0x the capital invested, the same MOIC, yet their IRRs differ by more than 17 percentage points. So which figure is the “real” return?
IRR vs MOIC: time-weighted rate vs absolute multiple
The two streams above expose the core distinction. MOIC (multiple on invested capital) is total value returned divided by capital invested: a 3.0x means the investment tripled the money. It measures magnitude and is blind to timing: both streams are 3.0x because both returned $30,000,000.00 on $10,000,000.00. IRR measures speed (the annualized, time-weighted rate) and is acutely sensitive to when cash arrives. The earlier-cash stream carries the higher IRR despite the identical multiple.
This is also where IRR can mislead. Because earlier distributions raise IRR, a fund can flatter its headline IRR by accelerating early cash: a subscription line that defers capital calls, or a quick partial exit, lifts IRR without changing the eventual multiple. IRR can also be inflated by unrealized (paper) marks that have not been returned in cash, and a fund’s gross IRR and net IRR can diverge sharply once management fees and carried interest are deducted. None of that touches MOIC, which reports the same multiple regardless of timing or marks. That is why funds report IRR and MOIC together.
| Attribute | IRR (Internal Rate of Return) | MOIC (Multiple on Invested Capital) |
|---|---|---|
| What it measures | Annualized rate of return | Absolute multiple of capital returned |
| Units | Percentage per year (e.g., 24.6%) | Multiple (e.g., 3.0x) |
| Time value of money | Yes, time-weighted | No, timing-blind |
| Sensitive to cash-flow timing | Yes, earlier cash raises IRR | No, same dollars, same multiple |
| Can be gamed by | Accelerating early distributions; paper marks | Hard to game; reflects total value |
| Best answer to | ”What annual rate did this earn?" | "How many times did I multiply my money?” |
Neither number is sufficient alone. MOIC says nothing about how long the capital was tied up; IRR says nothing about how much money was actually made. Reporting both, alongside the holding period, keeps either figure from telling half the story.
Gross vs net IRR
As with MOIC, IRR is reported in two forms, and the label matters. Gross IRR is calculated on the underlying deal cash flows, before fund-level costs; it measures the performance of the investments themselves. Net IRR is calculated after management fees, fund expenses, and carried interest are deducted; it measures what limited partners actually earn. Gross IRR is always equal to or higher than net IRR, and the gap can run several percentage points, widening with higher fees and carry.
A benchmark quoted as “a 25% IRR” is ambiguous until it specifies gross or net. Limited partners care about net IRR, because that is the return on the capital they paid in; general partners often cite gross IRR to isolate deal selection from fee drag. When comparing funds or deals, confirm that both figures are stated on the same basis.
What is a good IRR?
What counts as a good IRR depends on the risk and asset class, and on whether the figure is gross or net. As a rough, non-prescriptive guide: lower-risk, income-oriented strategies may be acceptable in the high single digits; moderate-risk investments target roughly 10%-15%; and venture and private equity generally aim for the high teens to 20%+ net, compensating investors for illiquidity and the risk of loss. These ranges are consistent with the MOIC benchmarks reported for the same strategies.
The load-bearing caveat is that a high IRR over a short hold can still be a small absolute gain. Doubling $100,000 in three months is a spectacular IRR and a $100,000 profit; doubling $100,000,000 over seven years is a modest IRR and a far larger one. This is precisely why IRR must be read alongside MOIC and the holding period rather than ranked on its own.
Limitations of IRR
IRR is a standard return metric, but four limitations recur:
1. Timing sensitivity and gameability. Because earlier distributions raise IRR, the metric can be flattered by accelerating cash: deferring capital calls or pulling forward a partial exit lifts IRR without improving the eventual multiple.
2. It ignores absolute size. IRR is a rate, so it says nothing about how much money was made. A high IRR on a small, short deal can be a trivial dollar gain, the gap that MOIC fills.
3. The reinvestment assumption. Standard IRR implicitly assumes interim cash is reinvested at the IRR itself, which is often unrealistic for a high IRR. The modified internal rate of return (MIRR) adjusts for this by assuming reinvestment at a separate, specified rate.
4. Multiple or no solution. A cash-flow stream that changes sign more than once, for example outflow then inflow then a later outflow (a follow-on or capital call), can mathematically produce more than one IRR, or none at all, because the NPV equation can cross zero multiple times.
Where IRR shows up: funds, exits, and waterfalls
IRR is the rate limited partners use to judge a fund and the rate general partners report alongside MOIC in every performance update. It also gates economics directly: an IRR-style hurdle, or preferred return, is the threshold a fund must clear before the general partner takes carried interest, and whether that hurdle is measured deal-by-deal or across the whole fund is the dividing line between the European and American waterfall models.
Both IRR and MOIC are computed off the exit cash flows a distribution waterfall produces, modeling how proceeds split across each share class is what generates the per-holder cash flows these metrics then summarize. Tools like Waterfalls model those exit cash flows across the cap table; they do not report a fund’s official IRR, which should come from the fund’s books and administrator.
Frequently asked questions
What is IRR in simple terms?
IRR (internal rate of return) is the annual rate of return at which an investment breaks even in present-value terms: the single discount rate that makes the net present value of all its cash flows equal zero. A higher IRR means a higher projected annual return.
What does a 20% IRR mean?
A 20% IRR means the investment is projected to earn about 20% per year on the capital still at work, compounding over the holding period. It does not mean a flat 20% of the original amount each year; IRR is time-weighted and assumes returns compound.
Is a 20% IRR good?
For higher-risk private investments such as venture or private equity, a high-teens to 20%+ IRR is generally considered strong; for lower-risk investments a far lower IRR can be acceptable. IRR should always be read alongside MOIC, because a high IRR over a short hold can still be a small absolute gain.
What is the difference between IRR and MOIC?
IRR measures the annualized, time-weighted rate of return and is sensitive to when cash comes back; MOIC measures the absolute multiple of capital returned and ignores timing. The same investment can show 3.0x MOIC with very different IRRs depending on how early the cash is distributed.
How do you calculate IRR?
IRR is the rate that sets net present value to zero, solved iteratively rather than with a closed formula. In practice you use a spreadsheet: =IRR(values) for evenly spaced periods, or =XIRR(values, dates) when cash flows fall on irregular dates.
What are the limitations of IRR?
IRR can be inflated by early distributions, ignores the absolute size of the gain, assumes interim cash is reinvested at the IRR itself, and can return multiple values, or none, when cash flows change sign more than once. It is read alongside MOIC for that reason.
The bottom line
IRR is the annualized, time-weighted rate of return that sets an investment’s net present value to zero. Its strength is that it prices the timing of cash, and that is also its weakness, because earlier distributions and paper marks can flatter the headline figure while saying nothing about how much money was actually made. That is the gap MOIC fills, which is why funds report the two together, on a clearly stated gross or net basis, with the holding period.
IRR summarizes the rate; what produces it is the distribution that splits exit proceeds across the cap table. Waterfalls models option pools, dilution, and exit-waterfall distributions from a cap table, so the actual exit cash flows (and the IRR and MOIC they imply) can be computed rather than estimated. It is a modeling tool for those cash flows, not a fund-administration or returns-reporting system.
This article is for educational purposes and is not investment, legal, or tax advice; consult a qualified professional before relying on any return metric to make a decision.