The rehab budget and the rent estimate are not independent variables, but most BRRRR models treat them that way.

This is a clean articulation of a risk that most beginner BRRRR resources skip entirely, so I want to make sure the underlying mechanics are readable for anyone newer to the thread.

A few terms first. A cap rate (capitalization rate) is a way appraisers measure a property's value based on its income. Simplified: annual rent income divided by the cap rate gives you the appraised value. So if two units each rent for $1,700 a month, annual income is $40,800. At a 7.5% cap rate, the appraised value is roughly $544,000. Drop the rent to $1,700 from $1,850 and that income falls by $3,600 a year, which at the same cap rate reduces the appraised value by about $48,000. The math in the original post follows directly from that.

LTV means loan-to-value. A 75% LTV cash-out refinance means the lender will lend you up to 75 cents for every dollar of appraised value. So $48,000 less in appraised value means roughly $36,000 less cash returned. That is real capital, not a rounding error.

The point about rent being a market number that keeps moving while construction runs is the part I'd highlight for anyone building their first BRRRR model. Four to six months is a long time in a local rental market. Underwriting rent on the day you write the scope and then not revisiting it at month four is optimistic by assumption, and the deal only reveals that at the refi table.

The stress-testing approach described here, running the model at 8% below projected rent and at a slightly higher cap rate, is the right habit to build early. The Analyst archetype tends toward this kind of sensitivity analysis naturally, but it is worth naming for anyone who is not yet there.

What market type are you seeing this pattern in most, somewhere with flat rent growth, or somewhere actively softening?

The 48k appraisal hit in that example is doing something worth separating out: most of it came from the cap rate applied to the rent shortfall, not from the rent shortfall itself. A 150-per-unit monthly miss sounds manageable until you divide it by a 7.5 cap, and then it becomes a number that changes whether the deal works at all. That math is why the refi appraisal is the most sensitive output in a BRRRR model, and why rent assumptions need a timeline attached, not just a figure.

There are two ways to build the model with this in mind. You can underwrite to rent at the time of stabilization by pulling comps on recently leased units and then discounting that number by whatever the local vacancy trend suggests for the following quarter. Or you can underwrite to current rent and apply a haircut to the projected appraised value directly, which is slower to explain to a lender but puts the stress where it belongs. The first method wins when you have enough comparable lease data to project forward with confidence. The second wins when the submarket is genuinely hard to read and you want the conservatism baked into the number a lender sees.

What is the typical gap between your scope-finalization date and your projected first lease date on the deals you are modeling?