For a long time in my career, if a project showed a strong IRR, especially north of 22–24 percent, it felt almost self-evident that the deal was good. The spreadsheets looked clean, the assumptions felt reasonable, and the number itself carried a quiet authority.
That belief started cracking not because of the theory, but because of the execution. In a recent project that looked excellent on paper, IRR was comfortably above our internal hurdle. Sales were assumed to start early, collections were phased neatly, and the exit looked well timed.
But as construction progressed, the lived experience of the project told a very different story. Costs were front-loaded, approvals took longer than expected, and sales, while eventually strong, did not convert into cash at the pace the model had promised. Month after month, the team was not debating returns but managing liquidity. Payments had to be sequenced carefully, buffers were constantly discussed, and the stress had nothing to do with profitability and everything to do with timing. But we survived and delivered a good product, we maintained our reputation, and we learned lessons that made every subsequent project better.
That was the moment I realised that IRR had not lied, but it had spoken only one part of the truth.
IRR is fundamentally a measure of speed. It tells you how quickly capital is expected to come back, assuming the timing works broadly as planned ( which it seldom does!). What it does not tell you how painful the journey is before you get there.
Here’s the simplified example that made it obvious to me:
Two projects, two different stories:
- One project might return 130 on an investment of 100 in under two years. The IRR will look fantastic, often close to 30%, but the wealth created is still limited.
- Another project might return 220 on the same 100, but over six or seven years. The IRR will look modest in comparison, yet the value created is far higher.
More importantly, IRR is blind to the most dangerous phase of any real estate project, which is the period before sales and collections stabilise. It does not show you how deep the cash deficit becomes at the worst point, or how sensitive the project is to a six-month delay in approvals or demand. A project can have a strong IRR and still be financially uncomfortable for most of its life, especially if the cash inflows are back-ended while construction costs and interest run ahead of them.
Over time, I stopped asking whether an IRR was good or bad and started asking what sat beside it. I look closely at:
- The equity multiple to understand whether the project actually creates meaningful wealth, not just fast turnover
- Peak negative cash to understand how much stress the project will place on the balance sheet before it stabilises
- The cashflow curve, which tells me whether the story in the spreadsheet has any chance of surviving contact with site reality
My perspective today is simple. IRR is useful, but incomplete. It describes velocity, not comfort, and not resilience. In projects , those distinctions matter more than most models admit. The projects that succeed over cycles are not the ones with the sharpest IRR graphs, but the ones that balance reasonable speed with real cash discipline and enough margin for error when reality inevitably deviates from plan.
Whenever someone presents a project with a confident IRR number, I no longer challenge it. I just ask a quieter question. What does the cash position look like in the worst month of this project? The answer to that question usually tells you far more than the IRR ever will.