Data Sources — prei¶

This document catalogs all data sources used by prei for market intelligence, property valuation, and investor analytics.

Market Scoring Data¶

Used by the Markets page (/markets/) and core/services/market_scoring.py to compute market scores per ZIP code.

Data Point	Source	Type	Update Frequency
Price-to-Rent Ratio	Calculated from property data (purchase price ÷ annual rent)	Derived	Per-property entry
Unemployment Rate	Bureau of Labor Statistics (BLS) via public API	External	Monthly
Population Growth	US Census Bureau ACS 5-year estimates	External	Annual
Employment Diversity	BLS County Employment & Wages	External	Quarterly
Market Verdict	Composite score from above signals	Derived	On refresh

Market Scoring Weights¶

Signal	Weight	Data Source
Price-to-Rent Ratio	30%	Property records / Zillow public data
Unemployment Rate	25%	BLS LAUS series
Population Growth	25%	Census ACS
Employment Diversity	20%	BLS QCEW

Market Refresh¶

The MarketRefreshView (/markets/refresh/) triggers a re-score of all ZIP codes found in the current user's properties. It scopes to Property.objects.filter(user=request.user) — no cross-user access.

# core/services/market_scoring.py
def score_market_by_zip(zip_code: str) -> dict:
    """Return dict with pr_ratio, verdict, unemployment, pop_growth, overall_score."""

Property Financial Data¶

Used by the Underwriting Score engine (core/services/scoring.py).

Data Point	Source	Type
Purchase Price	User input via Property Form	User-provided
Monthly Rent (Gross)	User input via Property Form	User-provided
Operating Expenses	User input via Property Form (taxes, insurance, HOA, maint., capex)	User-provided
Down Payment %	User input via Property Form	User-provided
Interest Rate	User input via Property Form	User-provided
Loan Term	User input via Property Form	User-provided
Vacancy Rate	User input via Property Form (default 5%)	User-provided
Management Fee %	User input via Property Form (default 8%)	User-provided
Land Value %	From `UserProfile.land_value_pct`	User-provided
Investment Targets	From `UserInvestmentTargets` model	User-provided

Derived Metrics¶

Metric	Formula	Service
NOI	Effective Rent - Total Expenses	`investor_app/finance/utils.py`
Cap Rate	NOI / Purchase Price	`investor_app/finance/utils.py`
Cash-on-Cash	Annual Cash Flow / Down Payment	`investor_app/finance/utils.py`
DSCR	NOI / Annual Debt Service	`investor_app/finance/utils.py`
GRM	Purchase Price / Annual Rent	`investor_app/finance/utils.py`
After-Tax IRR	NPV-based with exit cap scenario	`core/services/projections.py`
Depreciation	27.5-yr straight-line on improvements	`investor_app/finance/utils.py`

BRRRR Calculation Inputs¶

Used by the BRRRR Calculator (core/services/brrrr.py and templates/brrrr_calculator.html).

Data Point	Source	Notes
Purchase Price	User input
After-Repair Value (ARV)	User input	Estimated market value post-rehab
Rehab Cost	User input	Total renovation budget
Closing Costs	User input	Optional, defaults to 0
Refinance LTV %	User input	Default 75%
Monthly Rent	User input	Post-rehab expected rent
Property Tax	User input	Annual
Insurance	User input	Annual
Other Expenses	User input	Monthly
Down Payment %	User input
Interest Rate	User input	Purchase loan rate
Refinance Rate	User input	Cash-out refi rate

BRRRR Derived Metrics¶

Metric	Formula	Location
ARV Estimate	`median(comps_ppsf) × subject_sqft`	`investor_app/finance/utils.py`
Rehab Cost Estimate	`cost_per_sqft[level] × sqft`	`investor_app/finance/utils.py`
Max Refi Loan	`ARV × LTV%`	`investor_app/finance/utils.py`
Cash Left in Deal	`purchase + rehab + closing - refi_loan`	`client-side JS + brrrr.py`
BRRRR CoC Return	`annual_cf / cash_left` (∞ if ≤ 0)	`investor_app/finance/utils.py`

External Integrations (Planned)¶

Integration	Purpose	Phase
ATTOM API	Property details, comps, ownership	Phase 4
HUD / FHA	Government-owned property lists	Phase 4
Zillow / Redfin	Public listing data, Zestimate	Phase 4
Census API	Demographics, population trends	Phase 2 (partial via manual data)
BLS API	Employment, wage data	Phase 2 (partial via manual data)

Data Quality Notes¶

User-provided data is assumed accurate — prei performs no verification against external sources.
Market data comes from public, free-tier government APIs or manual entry. Refresh timing depends on data source update frequency.
BRRRR ARV is user-estimated or derived from comparable sales; a formal CMA engine is planned for Phase 4.
All currency values use Python Decimal — no floats for financial data.