New Madrid County presents a split underwriting signal. Zillow’s county observation labeled 2026-06 puts the median home value at $104,768, down 0.12% year over year, while FHFA’s annual observation labeled 2025 shows a repeat-transaction index up 5.63% and 47.36% over five years. These are different vintages and methods, so they should not be averaged into one growth rate. The tension favors investigation by an investor who can verify property-level economics; anyone relying on appreciation alone should be cautious.
Measured market rent is not published, so gross yield cannot be computed. HUD’s two-bedroom FMR is $890 per month, but it is a payment standard, not an asking-rent estimate. The effective property-tax rate is 0.73%, with a median annual tax of $814. Those costs must be tested against actual rent, insurance, repairs, vacancy and financing; the record supplies none of those property-level operating inputs. Because no Realtor.com listing price or closed-sale evidence is supplied, price-to-rent support and negotiation depth are also untested.
Demand evidence is mixed, not strong enough to establish tenant or buyer depth. Net migration was -45, and the AGI gap was -$741, meaning average income of moving households was lower inbound than outbound. QCEW covered employment located in the county averaged 7,372 jobs, down 1.76%, while average weekly wage was $888, up 3.14%. Trade, transportation, and utilities was the largest disclosed private supersector; QCEW is workplace covered employment, not resident employment or unemployment. Investor purchases were 20 of 115 total purchases, or 17.39%: competition is visible but not dominant, and this does not prove leasing demand.
Earthquake is the dominant hazard. The modeled climate-loss ratio is 0.37% of building value per year, but it is not an insurance quote or a property-specific loss estimate. Next, obtain market rent, property condition, vacancy, insurance terms and earthquake deductibles or retrofit needs, then test taxes and operating costs at the asset level. Evidence count is seven of eight, but without rent the underwriter cannot determine gross yield, rent coverage or whether price is income-supported. County figures also do not establish a metro conclusion.