IRR is the most commonly cited return metric in private real estate offerings and the most easily gamed. This is not a claim about fraud. The techniques described below are widely used, generally disclosed in PPM footnotes, and structurally designed to present the highest plausible number without technically misrepresenting anything. The LP's job is to see through them.
A projected 17 percent IRR and a projected 11 percent IRR on the same underlying asset and capital stack can both be mathematically accurate depending on three input choices: when the preferred return clock starts, what exit cap rate is assumed, and how equity multiple is calculated. Understanding each technique lets you normalize across offerings and compare deals on an equivalent basis.
IRR is highly sensitive to the timing and size of cash flows, particularly the terminal distribution at exit. A deal that returns 1.0x equity over 5 years with a large exit distribution will show a much higher IRR than one that pays regular quarterly distributions and exits at the same total multiple. The math benefits sponsors who backload returns into the sale event, which is also the event with the most valuation uncertainty.
In most syndications, capital is called in one or more tranches after the offering closes. Some deals call 100 percent of equity at close. Others call it over 6 to 18 months as renovation or construction milestones are reached.
The manipulation point is where the preferred return clock starts. Starting the pref accrual from the date each tranche is drawn (actual deployment) versus the date of commitment affects the total preferred return owed to LPs and, more importantly for IRR calculation, affects when the LP cash flow model shows the initial investment occurring.
If a sponsor uses commitment date as the starting point for LP cash flows in their IRR model but draws capital over 12 months, the LP's modeled cash flows begin earlier. This spreads the time-weighted return over a longer period, which compresses the denominator of the IRR calculation and produces a higher result. Conversely, sponsors who model cash flows starting at actual capital call will show a lower IRR on an identical deal, because the investment period is shorter.
"Does the IRR projection model cash flows from the date of commitment or the date each capital call is actually drawn? Can you provide me the underlying cash flow schedule in Excel so I can verify the timing assumptions?"
IRR in a real estate deal is almost entirely driven by the terminal value at exit. In a 5-year hold, a typical deal might generate 2 to 4 percent of its total LP returns through interim distributions, with 80 to 90 percent of value captured at the sale event. The cap rate assumed at exit therefore drives the majority of the IRR figure in the base case.
The technique is to select an exit cap rate that is either: (a) equal to the going-in cap rate, which assumes no cap rate expansion despite a materially different rate environment at exit, or (b) lower than the going-in cap rate, which assumes cap rate compression that requires either rate cuts or increased institutional demand not yet visible in current market data.
Both assumptions may be directionally reasonable. The problem is when the base case is presented as the most likely scenario rather than the optimistic scenario, and when no sensitivity analysis is provided. A deal that projects 16 percent IRR at a 5.0 percent exit cap and 9 percent IRR at a 5.75 percent exit cap has a wildly different risk profile than the base case number suggests, and that information belongs in the offering summary, not buried in an appendix.
"What is the current cap rate for comparable sales in this submarket, and can you show me the IRR sensitivity at exit cap rates of plus 50 and plus 100 basis points above your base case assumption?"
Equity multiple is often presented alongside IRR as a cross-check on return quality. A deal might show "projected 1.85x equity multiple and 16 percent IRR." What the headline figure typically does not specify: whether the multiple is gross (before sponsor fees and promote) or net (after all fees and promote distributions to GP).
On a deal with a 2 percent acquisition fee, 1.5 percent annual asset management fee, 1 percent disposition fee, and a 30 percent GP promote above the preferred return, the difference between gross and net LP equity multiple can be 0.15 to 0.25x. A gross 1.85x may be a net 1.62x to 1.70x depending on hold period and exit size.
This is not fraud. It is a disclosure gap that most LPs do not catch because they do not distinguish between gross and net in the offering summary. The underlying PPM will show the fee structure separately, but the headline numbers are often gross.
"Is the projected IRR and equity multiple gross or net of all fees and the GP promote? Can you provide the waterfall model showing the line items for each fee and the GP carried interest at your base case exit?"
What to Model Yourself
You do not need to build a full DCF to stress-test these techniques. Three targeted calculations take less than 15 minutes in a spreadsheet:
- Exit cap sensitivity. Take the sponsor's projected Year 5 NOI. Divide it by going-in cap, exit cap, and exit cap plus 75 bps. Calculate the change in total LP proceeds at each exit value. This shows you how much cap rate risk is embedded in the projected IRR.
- Net return calculation. Take the projected gross LP return, subtract total fees over the hold period (acquisition, management, disposition), subtract the GP promote on projected above-pref profits. The result is your net equity multiple. If the sponsor will not provide a fee schedule, the PPM has it.
- Cash flow timing audit. Ask for the Excel model. Look at column A: the period headers should show actual capital call dates, not commitment date. If the first negative cash flow row is month 0 and the deal draws capital over 12 months, the IRR is overstated.
None of these tests will tell you a deal is bad. They will tell you what a deal actually offers on a normalized basis. When you apply the same three-step audit to every offering you review, the comparison becomes accurate. A deal with a 13 percent projected net IRR after rigorous sensitivity testing is often more credible and more valuable than one projecting 18 percent with opaque inputs and no downside analysis.
The sponsors who can answer these questions quickly and specifically, with supporting model access, are generally the same sponsors whose realized results align more closely with their original projections. The correlation is not coincidental.