DLRadar · Methodology

Auditable Distress Intelligence

Every score on DLRadar is computed from public-record sources by deterministic formulas. No black-box ML. No fabricated data. Every claim traces to a SQL view name you can audit.

Deterministic scoring13 scoring engines33,775 ZIPs · 3,194 counties21+ public sources

Live coverage

Counts pulled live from production views — refreshed when this page is re-rendered (every 60 minutes).

33,775

ZIPs scored

3,194

Counties tracked

4,327

Banks scored

33,058

Layer 4 clusters

161,202

Listings tracked

Qualified ≥30%/≥10%

States

27.4/100

Avg cluster score

The acquisition funnel

Every engine below feeds one of four stages. Stages compose into a single workflow.

01 · SIGNAL

Detect distress + market stress + bank pressure across 9,611 ZIPs

ZIP Stress · Bank Stress · Insurance Distress · 12 Distress Categories

02 · UNDERSTAND

Cluster + classify markets by phase, capital readiness, exit timing

Layer 4 Clusters · Liquidity+Exit · Market Phase · Heatmap

03 · ACCESS

Surface qualified listings + tracked opportunities per asset class

DLSpecial · Asset Opportunity · Alpha Rankings · DLLenderRadar

04 · EXECUTE

Today's actionable list + capital allocator + alerts

Daily Deals · Real-Time Deal Feed · Fund Mode allocator

13 scoring engines

Each card shows what the engine scores, the inputs it consumes, the deterministic formula, the SQL view it lives in, and how often it refreshes. Every claim on DLRadar can be audited by reading the view definition.

🏦Layer 4 — Capital Readiness Clusters

Surface: /bank-distress

Five-tier ZIP cluster ranking — Deploy Now → Buy Now → Watch → Monitor → Avoid — combining bank pressure, ZIP stress, vacancy, listing density, and exit timing.

Inputs

→ Bank movement score (FDIC QoQ asset/equity delta)
→ ZIP stress composite (3-axis)
→ DLSpecial listing density per ZIP
→ Vacancy/demand signal (USPS + ACS)
→ Liquidity + Exit predictor band

Formula

cluster_score = capital_readiness_score · 0.35 + zip_stress_score · 0.25 + bank_movement_score · 0.20 + (avg_discount_score) · 0.15 + (population_score · job_growth_score) · 0.05

View / Source

analytics.layer4_clusters_safe_v1 → analytics.layer4_clusters_with_bank_v1

Refresh

Hourly via dlradar-layer4 timer

Output: 5-tier strategy label · cluster_score 0–100 · capital_required estimate

⏱Liquidity + Exit Predictor

Surface: /bank-distress (per cluster + per property)

Time-to-exit band per cluster — how long until a buyer can resell or refinance. Used by Fund Mode to allocate capital under a budget.

Inputs

→ DOM trajectory (median + 90th percentile)
→ Discount-cut velocity
→ ZIP stress + vacancy
→ Bank movement direction

Formula

exit_band ∈ { FAST_0_45, NORMAL_45_90, SLOW_90_180, ILLIQUID_180+ } · derived from threshold cuts on DOM_p50, discount_velocity_30d, and zip_stress_axis_2

View / Source

analytics.layer4_clusters_liquidity_exit_safe_v1

Refresh

Hourly with Layer 4

Output: exit_band · capital_strategy · time_to_exit_days_p50

📊ZIP Stress Composite (3-Axis)

Surface: /zip-stress/[zip]

Three independent axes per ZIP — Market Stress (pricing pressure), Structural Risk (long-horizon fragility), Distress Activity (events firing now).

Inputs

→ FHFA HPI YoY + drawdown depth (market)
→ FEMA disaster history + NOAA climate (structural)
→ NFIP flood claims + insurance pressure (structural)
→ Foreclosure filings + auction events (distress)
→ CFPB DQ 30–89 / 90+ (distress)
→ Census ACS rent burden + vacancy (structural)

Formula

axis_market = w1·hpi_drawdown + w2·hpi_yoy_decel + w3·price_stress · axis_structural = w4·climate + w5·insurance + w6·institutional + w7·construction · axis_distress = w8·foreclosure + w9·mortgage_dq + w10·auction · composite = (axis_market + axis_structural + axis_distress)/3

View / Source

analytics.v_zip_stress_rollup_v1 (read by /zip-stress + /heatmap)

Refresh

Every 8 hours via ZIP stress timer

Output: Composite 0–100 + 3 axes · 33,643 ZCTAs covered

💸Bank Stress + Momentum

Surface: /radar/banks · /banks/[cert]

FDIC-backed credit pressure scored per institution. Quarterly stress level + 7-day momentum delta surfaces banks moving fastest into trouble.

Inputs

→ Equity / assets ratio
→ ROA
→ Asset trend QoQ
→ Net charge-off ratio
→ Deposit composition (volatile vs core)
→ Branch concentration
→ FDIC enforcement actions

Formula

stress_score = clamp(0,100, 50 - 4·equity_ratio_pct - 8·roa_pct + 3·nco_ratio_pct + 2·asset_decline_qoq_pct + enforcement_penalty) · momentum = stress_score_today − stress_score_7d_ago

View / Source

analytics.bank_stress_momentum_v1 + analytics.v_zip_bank_stress_exposure_v1

Refresh

Daily via dlradar-fdic + quarterly call report ingest

Output: stress_score 0–100 · momentum delta · 4,431 banks tracked

🏛DLLenderRadar — Lending Capacity

Surface: /banking/lenders · property + ZIP panels

Inverse of bank stress — banks with public financial signals suggesting capacity to deploy real-estate loans. Not a loan-approval recommendation.

Inputs

→ Tier 1 capital ratio
→ Liquidity ratio (cash + securities / assets)
→ Loan-to-deposit ratio
→ Credit quality (noncurrent loans / total loans)
→ Loan-growth pressure (deposit growth − loan growth)
→ Branch market fit (proximity + ZIP context)

Formula

lending_capacity_score = 0.30·capital_strength + 0.20·liquidity + 0.20·credit_quality + 0.15·loan_growth_pressure + 0.10·local_presence + 0.05·borrower_fit

View / Source

banking.lender_capacity_scores_v1 + banking.zip_lender_capacity_v1

Refresh

Quarterly with FFIEC Call Reports + FDIC SoD

Output: Per-bank score 0–100 · per-ZIP top lender list · 4,327 banks scored

🗺Heatmap Composite (6-Component)

Surface: /heatmap

Single rank per state/ZIP combining six independent stress dimensions for top-down market scanning.

Inputs

→ FHFA HPI drawdown
→ Climate severity (NOAA storm events × frequency)
→ Insurance pressure (NFIP rate trajectory)
→ Institutional exposure (HUD + FHA concentration)
→ Foreclosure activity (filings per 1k housing units)
→ Auction event density

Formula

composite = (drawdown + climate + insurance + institutional + foreclosure + auction) / 6 · banded into PEAK / EXPANSION / NEUTRAL / CONTRACTION / RECOVERY

View / Source

analytics.v_geo_heatmap_all_v1

Refresh

Every 8 hours

Output: Composite 0–100 across 9,611 ZIPs · 3,150 counties

📈Market Phase — HPI Cycle Radar

Surface: /market-phase

Classifies every US county and ZCTA into one of five cycle phases using FHFA HPI trajectory + drawdown direction + county headwind.

Inputs

→ FHFA HPI YoY %
→ HPI 5-year acceleration
→ Drawdown from local peak
→ County headwind score

Formula

phase = case · YoY > 0 ∧ accel > 0 ∧ drawdown < 5% → EXPANSION · YoY > 0 ∧ accel ≤ 0 → PEAK · YoY ≤ 0 ∧ drawdown < 10% → NEUTRAL · YoY ≤ 0 ∧ drawdown ≥ 10% ∧ accel < 0 → CONTRACTION · YoY ≥ 0 after drawdown → RECOVERY

View / Source

analytics.us_county_headwind_snapshot_v1 + FHFA HPI ingest

Refresh

Monthly with FHFA HPI release

Output: 5-phase classification · 33,643 ZCTAs · 3,150 counties

🎯Asset Market Prediction

Surface: /opportunities (per asset class)

Direction label per (asset_class × county) — surfaces markets moving from EARLY to ACTIVE before inventory appears.

Inputs

→ Macro pressure score (rates + labor + HPI)
→ Bank stress in county
→ Listing inventory delta (28-day)
→ Permit activity (Census BPS 12mo)

Formula

label ∈ { EARLY, HEATING_UP, ACTIVE, COOLING, AVOID, NO_DATA } · derived from threshold cuts on macro_pressure_score, candidate_flow_28d, and inventory_velocity

View / Source

analytics.macro_prediction_labels_v1

Refresh

Daily

Output: Direction label · per asset class × county

⭐Alpha Rankings (TOP_100 / TOP_200)

Surface: /opportunities · landing top alpha

Composite alpha score per EARLY county — surfaces highest-conviction emerging markets before the herd.

Inputs

→ Macro/banking context (40%)
→ Bank stress signals (40%)
→ Candidate listing flow (20%)

Formula

alpha_score = 0.4·context_score + 0.4·bank_score + 0.2·candidate_flow_score · ranked descending within EARLY-labeled counties · top 100 + top 200 maintained

View / Source

analytics.deal_priority_v1 → analytics.deal_conviction_v1

Refresh

Daily with prediction labels

Output: Ranked county list with alpha_score 0–100

🔍Asset Opportunity Tracker

Surface: /opportunities (5 pages)

Stage classifier per (asset_class × county) — moves opportunities through CANDIDATE → ACTIVE → WATCHING → CONTEXT_READY pipeline.

Inputs

→ Listing density per asset class
→ Discount % distribution
→ Bank presence + stress
→ Days since first signal

Formula

stage = case · candidate_count ≥ threshold_active ∧ avg_discount ≥ 30% → ACTIVE_INVENTORY · candidate_count ≥ threshold_candidate → CANDIDATE_DEALS · context_data_complete → CONTEXT_READY · else → WATCHING

View / Source

analytics.asset_opportunities_v1

Refresh

Hourly with Layer 4

Output: Stage label · 5 asset class views · per-county opportunity count

🔔Real-Time Deal Feed

Surface: /new-deals · alerts

Event detection layer — flags state changes across the funnel (cluster moves to BUY_NOW, exit becomes FAST, new top deal in your ZIP).

Inputs

→ Layer 4 cluster transitions
→ Liquidity+Exit band changes
→ DLSpecial new qualified rows
→ Bank stress band changes

Formula

event ∈ { NEW_TOP_DEAL, CLUSTER_BECAME_BUY_NOW, EXIT_BECAME_FAST, BANK_STRESS_CRITICAL, NEW_DLSPECIAL_HIT } · fired when state_now ≠ state_prev with severity threshold

View / Source

analytics.deal_events_v1 + alerts.user_alert_rules_v1

Refresh

Every 5 minutes

Output: Time-stamped event stream → bell badge + email queue

💎DLSpecial Qualification

Surface: /listings

Strict deal filter — every qualified row meets a hard discount threshold against verifiable comps. No estimates.

Inputs

→ List price (live scraper feed)
→ Market value (appraiser just_value or ACS-blended median)
→ Taxable value (county tax roll)

Formula

qualified ⟺ (list_price ≤ market_value · 0.70) OR (list_price ≤ taxable_value · 0.90) · enforced at MV write-time + UI read-time

View / Source

analytics.listings_inventory_ui_safe_v1 + ui.v_discount_listings_passthrough

Refresh

Every 8 hours via DLSpecial dispatcher

Output: qualified flag + discount % + rule label per listing · 14,966 tracked / 7,409 qualified

🌊Insurance Distress (NFIP / FEMA)

Surface: ZIP stress report (insurance section)

Monthly ZCTA insurance pressure — surfaces ZIPs where insurance is becoming uneconomic (a leading abandonment signal).

Inputs

→ NFIP claim trends (frequency × loss severity)
→ Premium pressure (NFIP rate revisions)
→ Policy lapse rate
→ FEMA disaster declaration density

Formula

insurance_pressure = 0.40·claim_severity_zscore + 0.30·premium_trajectory + 0.20·lapse_rate + 0.10·disaster_density · banded 0–100

View / Source

analytics.zcta_insurance_distress_monthly_v2

Refresh

Monthly with NFIP + FEMA releases

Output: Per-ZCTA insurance score 0–100

12 distress categories per property

Inside Platinum + Supreme, every property is scored 0–100 across these independent categories. The stack countNumber of categories scoring above the distress threshold (typically ≥25). Stacked categories are high-conviction targets. tallies how many fire above threshold. Hover for details.

🏦

Mortgage Stress

Score 0–100

LTV ratio + ARM reset exposure + payment-to-income burden + HMDA denial rates per census tract.

📋

Tax Lien

Score 0–100

Delinquent property tax exposure from county recorder filings + lien sale schedules + redemption status.

⚖️

Judgment / Lis Pendens

Score 0–100

Civil judgments, mechanics liens, lis pendens — pending litigation or forced-sale risk.

🏘️

HOA / Code Violation

Score 0–100

HOA lien filings + municipal code violations + special assessment delinquencies.

🏚️

Vacancy

Score 0–100

USPS vacancy flags + utility disconnect + Census ACS vacancy rates.

🔑

Landlord Stress

Score 0–100

Eviction filing rates + rent-to-mortgage gaps + multi-property owner financial strain.

🏗️

Construction / Permit

Score 0–100

Stalled construction — expired permits + stop-work orders + CO gaps >12mo.

👴

Elderly / Probate

Score 0–100

Owner age signals + death record matches + trust/estate ownership patterns.

🏢

Commercial Distress

Score 0–100

Tenant vacancy + CMBS delinquency + NOI decline.

🏛️

IRS / Federal Lien

Score 0–100

IRS federal tax liens + SBA defaults + EIDL/PPP non-compliance.

🌊

Environmental

Score 0–100

FEMA flood zone + NOAA disaster frequency + EPA Superfund proximity + NFIP repetitive loss.

📜

Heir / Title Defect

Score 0–100

Heir property indicators + missing probate + fractional ownership + chain-of-title breaks.

Global distress composite (Platinum / Supreme dossier)

global_distress = 0.20·mortgage + 0.15·tax_lien + 0.10·judgment + 0.10·vacancy + 0.08·hoa + 0.08·landlord + 0.07·elderly + 0.06·environmental + 0.05·construction + 0.04·commercial + 0.04·irs_lien + 0.03·heir_title

View: analytics.v_property_distress_v1 · Refresh: hourly with promote

Public data sources

Every signal traces to one of these. No scraped private data, no synthetic estimates.

FHFA HPI

House Price Index, monthly, county + ZCTA

https://www.fhfa.gov/HPI

FDIC Call Reports

Quarterly bank financials per certificate

https://banks.data.fdic.gov

FDIC Summary of Deposits

Annual bank branches + deposits per ZIP

https://www.fdic.gov/sod

FFIEC Call Reports

Detailed regulatory bank data

https://www.ffiec.gov/npw

Census ACS / PUMS

5-year tenure, rent burden, vacancy, income

https://www.census.gov/acs

Census BPS

Building permits, monthly per place

https://www.census.gov/construction/bps

HUD HCV / Section 8

Voucher utilization + payment standards

https://www.huduser.gov

HUD FMR

Fair Market Rents, annual

https://www.huduser.gov/portal/datasets/fmr.html

HUD Assisted Housing

Project-based subsidized properties

https://www.huduser.gov

FEMA Disaster

Declared disaster history per county

https://www.fema.gov/openfema

NFIP Claims

Flood insurance claims, redacted

https://www.fema.gov/openfema

NOAA Storm Events

Severe weather event log

https://www.ncdc.noaa.gov/stormevents

FRED

Federal Reserve macro series (rates, labor)

https://fred.stlouisfed.org

BLS LAUS / QCEW / CES

County unemployment + payrolls

https://www.bls.gov

CFPB Mortgage Performance

DQ 30–89 + 90+ per county

https://www.consumerfinance.gov

FFIEC HMDA

Mortgage application + denial records

https://ffiec.cfpb.gov/data-browser

EPA FRS

Facility registry — industrial proximity

https://www.epa.gov/frs

USDA NASS

Land + farm values, cash rent

https://www.nass.usda.gov

RSMeans

Replacement cost benchmarks

https://www.rsmeans.com

County Recorder Filings

Liens, judgments, foreclosure events (per-county scrapers)

USAuctions Public Records

Auction calendar + bid records

Reproducibility

Every score on this platform is computed by SQL — the view names are listed on each engine card above. Inside Platinum + Supreme, every property dossier shows the underlying signal values. Inside DLBankDistress + DLLenderRadar, every cluster + bank exposes its component scores. Inside DLSpecial, every qualified row shows the exact discount calculation.

No ML

Zero black-box predictions. Every score is a deterministic formula on public inputs.

No fabricated data

When a signal is missing, the field shows '—'. We never synthesize values.

Auditable

Every claim → view name → public source. You can trace any score back to its inputs.

Now see the system in action

The bundle that fits depends on whether you're scanning markets, accessing real listings, or executing daily.

Bundle 1 · Market Intelligence

$799/mo

✓ Platinum property research
✓ Radar Markets · Heatmap + prediction
✓ Find pressure before listings appear

View access →