San Juan County presents a valuation-versus-liquidity tension: Zillow’s county median home value was $558,474 in 2026-06, up 2.87% year over year, while FHFA’s repeat-transaction HPI fell 3.16% in annual 2025 data. These are different methods and vintages and cannot be combined. The conflict warrants investigation by buyers relying on resale support and caution from underwriters seeking dependable income; the county record does not establish a closed-sale price trend.
No county market asking rent is published, so gross yield cannot be computed. HUD FMR of $1,227 is a payment standard, not evidence of asking rent, and cannot fill that gap. At a 0.32% effective property-tax rate, taxes are a carrying-cost input alongside the stated home-value benchmark, but insurance, maintenance, financing, and assessed-value evidence are absent. Rent comparables and tax bills are needed before reaching an income or all-in carrying-cost conclusion.
Realtor.com’s 2026-06 MLS market shows visible inventory up 73.17%, a median 81 days on market, and a 9.86% pending-to-active ratio. These are active-listing supply and marketing measures, not closed-sale prices or proof of demand. QCEW annual 2025 covered employment at county workplaces declined 5.27%; Leisure and hospitality is the largest disclosed private supersector, not the whole economy. The migration record shows 20 inbound tax-return households averaging $46,450 AGI, but no outflow, so net migration cannot be assessed. Investor share was 7.69% of 13 purchase mortgages, a limited measure of non-occupant competition.
Inland flood is the dominant hazard, and modeled expected annual building loss equals 0.26% of building value. That county-level metric does not quantify a specific parcel’s flood exposure, insurance availability, deductible, or mitigation needs. Next checks are property-level flood maps and insurance quotes, rent comps, assessment and tax records, listing histories, and closed transactions. Without those inputs, neither durable cash flow nor exit pricing can be underwritten.