Has anyone else noticed that the suburban workforce deal is now where everyone thinks the yield is hiding

The assumption doing the most work in those decks is usually rent growth, but the one that actually breaks the model first is expense ratio drift, and insurance is only part of it.

What's the debt structure on the deals you're still seeing pitched?

The assumption doing the most work in most of those decks is the expense ratio. Sponsors are still modeling T12 actuals from a period when insurance was a third of what it is now, tax assessments hadn't caught up to the 2021-era sale price, and deferred maintenance was being funded by stimulus-era cashflow. Stack a current insurance quote, a reassessed tax line, and a realistic capex reserve onto those deals and the cap rate that looked like a yield refuge becomes a breakeven exercise or worse.

Your Georgia example is the right way to think about it. $180k on a 96-unit deal is roughly $1,875 per unit annually just in insurance before you get to taxes, management, or payroll. If the pro forma was built at $600-800 per unit, that delta alone is enough to blow through most of the cash-on-cash margin that made the deal look attractive. The sponsors who closed on bridge debt in 2021-2022 are now facing that math at refi time with rates that haven't returned to their underwriting assumptions, and the gap between stabilized NOI and debt service isn't closing fast enough.

The risk I'd add to what you're describing: the workforce tenant base is also thinner on the collections side than the T12 shows, for exactly the reason you named. Moratorium-era data is embedded in operating histories that underwriters are still treating as representative. A deal that showed 96-97% collections through 2021 and 2022 may have been running on a structural subsidy that had nothing to do with tenant quality or management competence.

The product type probably works in markets where supply is actually contracting and where you can buy at a basis that reflects the real operating cost structure, not the 2021 version of it. The question is whether the current pricing on those decks you're seeing reflects that or whether the sponsor is still backing into the purchase price from a pro forma that needs the old expense environment to pencil.

What's the cap rate spread between what they're marketing and what the deal actually looks like when you recast expenses to current actuals?