Crash-Risk Composite — Methodology
A weighted composite of five publicly-published housing-stress signals, normalized to 0–100 per state, where higher = greater downside pressure. Not a forecast — a leading-edge indicator that's transparent, cited, and free.
Inputs & weights
| Factor | Weight | What it measures | Source |
|---|---|---|---|
| Price-to-income overvaluation | 30% | Median home price ÷ median household income; 5x is stretched, 7x is historically high. | U.S. Census ACS, IMF/Shiller methodology |
| YoY price weakness | 20% | Year-over-year change in median sale price; negative = active correction. | Redfin Data Center, FHFA HPI |
| Inventory surplus vs long-run | 20% | Active listings vs trailing 24-month average; surplus > 30% = soft. | Redfin monthly listings reports |
| Days-on-market spike | 15% | Current DOM vs trailing 24-month average; +20% spike + DOM > 60 days is the classic buyer-leverage signal. | NAR Realtors Confidence Index methodology |
| Permits contraction | 15% | YoY change in new private housing permits; builders pull back 6–12 months ahead of price softness. | U.S. Census Building Permits Survey (BPS) |
Normalization
Every input is mapped onto a 0–100 scale where 100 = max downside pressure. A few examples to make this concrete:
- V/I = 3.5 → normalized 0 (affordable). 5.0 → 50. 7.0+ → 100.
- YoY = -10% → 100. 0% → 50. +10% → 0.
- Inventory surplus +50% → 100. 0% → 50. -25% (deficit) → 0.
- DOM spike +40% → 100. 0% → 50. -20% → 0.
- Permits -30% YoY → 100. 0% → 50. +20% → 0.
Composite = weighted sum, also 0–100. Tier labels: 70+ = elevated downside risk, 50–69 = moderate, 30–49 = low, <30 = minimal.
What this is NOT
- It is not a forecast. A high crash-risk score in 2024 did not predict an actual crash — it's a "lean bearish" leading-edge indicator.
- It is not a national macro signal. Fed policy, recession, employment shocks are not in the input set. A state can score "low risk" and still drop 20% in a national recession.
- It is not market timing. Real-estate transactions take months and round-trip costs are 6-10% — never use a single composite to make a buy/sell decision in isolation.
- Coastal markets (CA / FL / NV / AZ) historically score higher because their volatility is wider — that doesn't mean they're guaranteed to drop more, just that their downside variance is larger.
Sources
- IMF Global Housing Watch
- FHFA House Price Index
- Redfin Data Center
- NAR Realtors Confidence Index
- U.S. Census Building Permits Survey
- Shiller, Robert. "Irrational Exuberance" methodology (Yale Press)
Open methodology. The score's source is the same per-state JSON every public consumer can access via /api/states/{slug}.json. Anyone can reproduce the composite — that's the design.