What You're Actually Trying to Calculate
The internal rate of return is the discount rate that makes the net present value of all cash flows equal zero. Most people encounter it when evaluating whether a project or investment is worth pursuing. It gets presented as this definitive metric, but in practice it is just one lens among several for assessing returns. Understanding how it works and where it breaks down matters more than memorizing the definition. If you are looking for a straightforward walkthrough, here it is. You take each cash flow in a series, assign a time period to it, and then solve for the rate r that satisfies this equation: sum of (cash flow divided by one plus r, all raised to the power of the period) equals zero. The key detail most tutorials skip is that the equation cannot be rearranged algebraically for anything beyond a single period. You solve it iteratively using a spreadsheet solver, a financial calculator, or successive approximation. In Excel or Google Sheets you enter the cash flows in consecutive cells and use the IRR function. The function assumes equal time intervals between each value by default. If your intervals are irregular, you switch to XIRR and provide the corresponding dates alongside each cash flow. That distinction alone prevents a lot of basic errors.
I remember working on a commercial real estate deal a few years back where the sponsor's pro forma had construction draws spread across thirteen months with uneven amounts. Someone ran the IRR function on those raw values and reported a 22 percent return. The problem was that the months did not represent equal risk periods because the draw schedule included a six-month dry spell in the middle when no capital was deployed. The IRR output was technically correct for the input series, but it was misleading as a performance measure because it treated idle months the same as active months. The fix was straightforward: I converted the draw schedule into monthly equity commitments mapped to actual calendar dates and recalculated using XIRR. The return dropped to 14.3 percent. That number aligned much better with what the deal actually delivered. There are nuances that do not show up in introductory material. One is the sign change issue. The IRR calculation can produce multiple valid rates when cash flows alternate between positive and negative more than once. A classic example is a project that requires an initial outlay, generates returns, then requires a major renovation cost further down the timeline. The mathematical result may give you two IRR values, and neither is necessarily wrong. The practical problem is that you cannot rely on IRR alone to rank these projects. In those cases you should fall back on NPV calculated at your actual hurdle rate, or use the modified internal rate of return, which compounds interim cash flows at a reinvestment rate you specify rather than assuming they earn the IRR itself. Another counter-intuitive point is that IRR favors shorter projects over longer ones even when the longer project creates more absolute value. This happens because IRR is a percentage measure, and percentage returns compress when you stretch cash flows over many periods. A project returning 30 percent over two years looks impressive next to a project returning 18 percent over ten years, but the latter may contribute significantly more to portfolio growth depending on your capital base. If you are comparing mutually exclusive projects, always cross-check with NPV. The project with the higher IRR is not automatically the better choice.
The reinvestment rate assumption deserves its own attention. Standard IRR implicitly assumes that every intermediate cash inflow is reinvested at the same rate as the IRR itself. That is rarely realistic. If your IRR comes out to 25 percent, you are not going to find enough opportunities at 25 percent to compound those interim payments. The MIRR approach lets you set a separate reinvestment rate, typically your weighted average cost of capital or a short-term treasury yield, which produces a result closer to what actually happens in practice. Pitfalls I see repeatedly in beginner work include treating IRR as a standalone decision rule, ignoring the scale of the investment, and failing to account for taxes and fees. An IRR calculation based on gross cash flows will overstate the return you actually keep. Running it without subtracting management fees, transaction costs, or tax drag creates a gap between reported and realized performance that compounds over time. Another common mistake is comparing IRRs across projects with different capital structures. Debt financing changes the timing and size of equity cash flows, which shifts the IRR independently of operational performance. That is why equity IRR and unlevered IRR can tell you very different stories about the same asset. For most people working through their first IRR analysis, the practical takeaway is to build a model that tracks actual calendar dates, uses XIRR for irregular timing, flags projects with multiple sign changes, and runs a parallel NPV calculation at a stated discount rate. That process takes about twenty minutes to set up on a simple model, and it saves you from making decisions based on a number that looks impressive but does not reflect the economics of the investment.
IRR remains useful because it translates a stream of cash flows into a single percentage that is easy to communicate. It is not useful when you need precision, when cash flows are unconventional, or when you are choosing between projects of different sizes. In those situations you rely on NPV and sensitivity analysis instead. Knowing which tool applies to which problem is what separates people who use IRR from people who get fooled by it.
