Fulton County has a split county-level case: Zillow’s 2026-06 county reading shows a $421,432 median home value, down 2.37%, while measured median asking rent is $1,907 per month, up 2.75%, and the reported gross yield is 5.43%. That price/rent divergence merits investigation by operators who can verify unit-level rents and expenses; underwriting that depends on appreciation or very thin margins deserves caution. This is county evidence, not a property-level conclusion.
FHFA’s 2025 annual repeat-transaction HPI rose 1.24%, directionally challenging Zillow’s decline. It is an appreciation index, not a home value, and its vintage and method differ from Zillow’s; the rates must not be combined. HUD’s two-bedroom FMR is a payment standard rather than an asking-rent estimate, so it cannot replace the published market-rent measure. The 0.88% effective property-tax rate is a carrying-cost input. Published market rent permits the stated gross-yield measure, but this record cannot establish net yield because operating costs, insurance, maintenance, financing and vacancy are not published.
Realtor.com MLS evidence adds a mixed marketing read: 5,971 active listings, asking prices rising, inventory and median days on market falling, and 21.72% of listings reduced. These are listing-market measures—asking prices, visible supply, marketing time and seller concessions—not closed-sale prices or proof of buyer demand; the pending ratio does not resolve that limit. Tax-return migration was net negative by 392 households, and out-movers’ average income exceeded in-movers’ by $7,732, a composition flag rather than renter-demand evidence. Non-occupant purchase mortgages were 9.86% of 12,795 purchases, identifying countywide investor participation but not bidding behavior for a target asset. Annual QCEW records covered workplace employment, not resident employment or unemployment; professional and business services is the largest disclosed private supersector, while QCEW does not describe the whole economy.
Inland flood is the dominant hazard, and the modeled expected annual building-value loss ratio is 0.09%. That is modeled loss exposure, not damage history or a parcel-specific insurance quote. Next checks are flood zone, elevation, claims and coverage terms; property-level tax assessments; comparable rents by unit type; vacancy and turnover; condition; and repair, insurance and financing costs. Without those items, county data cannot determine asset-level net cash flow, coverage cost or flood resilience.