The reinvestment rate fallacy
IRR’s most important flaw is that it implicitly assumes all positive cash flows generated by a project are reinvested at the IRR rate until the end of the project horizon. For a project with a 40% IRR, this means every pound of early cash flow is assumed to grow at 40% per year until the project ends — which is rarely achievable in practice. This biases IRR upward for projects with early, large cash flows, and creates a fictitious comparison between projects of different duration. Mathematically: IRR is the solution r to: ∑ CFₜ / (1+r)ᵗ = 0. The implicit assumption is that each cash flow CFₜ is reinvested at rate r from time t to the end of the horizon. For a typical firm with a cost of capital of 10%, reinvesting at 40% (the IRR) is simply not possible — the IRR overstates the realistic compound return of the investment.
How private equity manipulates IRR
Private equity uses IRR as the primary return metric, creating specific incentives to game it. Common techniques: front-loading dividend recapitalisations (early cash distributions boost IRR even with no improvement in underlying business value); use of subscription lines of credit (borrowing for the first 6–12 months delays calling investor capital, shortening the holding period over which the investment return is measured, boosting IRR); and holding duration management (selling winners quickly — achieving 3x on a 2-year hold shows a higher IRR than 3x on a 4-year hold). The LP community has increasingly adopted TVPI (Total Value to Paid-In, a multiple on invested capital) alongside IRR to avoid these gaming techniques. A high IRR on a short hold must be evaluated alongside the absolute multiple of money returned to assess whether value was genuinely created.
“IRR is the single most abused metric in all of finance. It can be engineered, gamed, and misinterpreted. Treat it as a preliminary screen, not a final verdict.”
What this means for you
Always consider IRR alongside NPV and investment multiples. Two projects with identical IRRs can have dramatically different NPVs if they differ in scale (a 30% IRR on £1m is less valuable than a 20% IRR on £100m). When evaluating private equity or private credit investments, ask for TVPI and DPI (Distributed to Paid-In — cash actually returned) alongside IRR — the combination is much harder to game than IRR alone. For corporate capital budgeting, NPV should be the primary criterion; IRR is a useful check, but when NPV and IRR give conflicting rankings for mutually exclusive projects (which happens when projects have different scale or duration), always follow the NPV ranking.