What returns actually mean (projections are not promises)

A projected return answers a hypothetical question: if the stated assumptions occur, what does the model calculate?

It does not tell you what will occur. It does not promise a distribution, a sale date, or the return of principal. It is not evidence that the sponsor can control rents, expenses, interest rates, buyers, lenders, or time.

This is general education, not legal, tax, securities, or investment advice. Private investments can be illiquid, projected outcomes may not occur, and an investor can lose some or all invested capital. Evaluate the actual documents and your circumstances with qualified professionals.

The metrics answer different questions

Return terms are not interchangeable labels for “good.” Each one measures a different part of a hypothetical outcome:

• Cash-on-cash return compares a period’s cash distributions with the invested capital used in the calculation.
• Internal rate of return (IRR) is sensitive to both the amount and timing of modeled cash flows.
• Equity multiple compares total modeled distributions with modeled invested equity, without expressing timing by itself.
• Preferred return describes a priority in the distribution waterfall. It is not a guaranteed payment or guaranteed rate of return.
• Average annual return can compress an uneven, multi-year modeled path into one average that does not show when cash is assumed to arrive.

When one metric is used to answer a different question, stop. An equity multiple does not tell you when cash arrives. An IRR does not tell you that its timing assumptions will occur. A preferred-return provision does not manufacture cash for distribution.

A hypothetical pattern is not an expected result

Consider two entirely hypothetical cash-flow patterns for an assumed $100,000 investment. These figures are examples only; they are not an offering, forecast, target, or representation of any likely outcome.

In hypothetical Pattern A, the model assumes distributions of $5,000 after each of the first three years and $120,000 of sale proceeds at the end of year three. In hypothetical Pattern B, the model assumes no interim distributions and $135,000 at the end of year three.

Both patterns total a hypothetical $135,000 of distributions, including the assumed final proceeds. They do not provide the same timing or liquidity. Neither pattern says that the property will generate the cash, that a buyer will pay the modeled price, or that investors will receive any stated amount.

That is why a headline metric is never the full receipt. You need the dated cash flows, the assumptions that produced them, the fees deducted along the way, and the waterfall that allocates whatever cash actually exists.

Every projection has conditions attached

Modeled outcomes may depend on rent growth, occupancy, collection rates, operating expenses, renovation cost and timing, interest rates, debt terms, reserves, fees, sale costs, hold period, exit value, and the distribution waterfall.

Change one assumption and the output changes. Change several at once—as real properties tend to do—and the headline can become irrelevant.

The model is useful as a map of assumptions. It becomes dangerous when its output is repeated without the word “projected,” when the assumptions are hidden, or when a reader treats a calculated scenario as a promise.

Trace the number backward

Review the investor presentation and model, if provided, alongside the PPM risk factors, operating agreement waterfall, debt terms, historical T-12, current rent roll, capital budget, and sale assumptions. Keep historical facts separate from sponsor estimates and model inputs.

Ask questions that expose dependency:

• Which assumption changes the hypothetical investor outcome the most?
• Which figures are historical, which are contractual, and which are projections?
• What happens in the model if rent growth is lower, expenses are higher, or renovations take longer?
• What debt maturity, extension, or refinancing assumptions affect the timing?
• What sale price or exit cap rate is assumed, and what happens if that assumption worsens?
• Which fees and waterfall tiers are applied before cash reaches investors?
• Is return of principal included in a figure that a reader might mistake for profit?

An answer made of confidence is not a sensitivity analysis. You need the changed inputs and the changed outputs.

Build the case that refuses to cooperate

Copy the model’s stated assumptions onto one page. Label every number as historical, contractual, or projected. Then create a separate hypothetical downside case with slower leasing, higher expenses, delayed timing, and less favorable sale pricing. Do not call that case a prediction either.

The purpose is not to guess the future correctly. It is to see what must go right, what the sponsor cannot control, how quickly modeled distributions can disappear, and whether the legal documents match the language used in the pitch.

If you remember one sentence, use this one: a projection is a question with assumptions, never an answer about your future money.