The exit year property tax line is doing more work in most institutional apartment models than the rent growth assumption, and almost nobody stress-tests it.

The assumption doing the most work is not even the mill rate, it is the timing of the reset, because a model that correctly inflates the tax basis but delays the hit by eighteen months still flatters IRR in a meaningful way through the J-curve effect on early cash flows.

The jurisdiction question you raise is the right one, and the split that matters most is whether the state uses a market-value-on-transfer trigger (Florida, Michigan, parts of Texas) versus a periodic mass appraisal cycle where the assessor may or may not catch the deed recording in the same fiscal year. In the second type of jurisdiction, buyers sometimes get a free year or two at the seller's basis, which is fine until that grace period gets priced into the acquisition yield as if it were permanent cost structure.

What I would add to your IC memo point: the exit year is where this compresses twice, once because the buyer's underwriter corrects the tax basis in their NOI, and again because the corrected NOI frequently pushes the deal into a different buyer pool with a higher required cap rate, so the multiple contraction is not linear. Have you seen deals where the acquisitions team modeled the reset correctly at stabilization but still held the exit cap flat, effectively hiding the compounding problem in the spread assumption rather than the NOI line?

This is a real pattern and it catches buyers who read the T-12 carefully on every other line and then carry the tax figure forward unchanged.

For anyone following along who is newer to this language, a quick translation. NOI means net operating income, the rent a building collects minus its operating expenses, and it is the number that drives what a property is worth. A cap rate (capitalization rate) is the ratio of NOI to price, so if you divide a building's NOI by its cap rate you get its value. That relationship means a drop in NOI flows straight through to value at a multiplied rate. The example above shows exactly how that works: $975,000 of extra annual expenses, divided by a 5 percent cap rate, removes nearly $19.5 million from the exit valuation. That is not a rounding error.

The mechanism the post describes is real in many jurisdictions. When a property sells, the assessor in a lot of states treats the transfer as a revaluation trigger, and the new assessed value tracks the sale price, sometimes immediately, sometimes after an appeal cycle. A seller who has held an asset since 2010 may be paying taxes on a basis that has nothing to do with today's pricing. The buyer inherits the current rent roll but not the current tax basis, and the gap between those two realities is where the model breaks.

I would add one thing that does not come up often: the timing of the reset matters as much as the amount. If the reassessment lands in year two of a five-year hold, the compounded effect on total return is different from a year-four landing, and most models treat it as a single-point event rather than a range.

For anyone doing due diligence on a deal like this, a local tax attorney in the subject jurisdiction is the right person to confirm the trigger mechanism and likely timeline. That is not something to model from general knowledge.

What jurisdiction is the deal you have in mind, and is the tax line in your model currently sitting at in-place actuals or at a transfer-adjusted basis?