Deal facts
Property, ownership, rent roll, and the T-12 operating statement as extracted from source documents.
property · rent_roll · operating_statement Open standard for commercial real-estate underwriting
One verifiable underwriting record for every CRE deal.
UW Markdown is an open interoperability standard for underwriting. A .uwx.md record keeps typed deal facts, underwriting assumptions, calculation inputs, narrative, and provenance together in one plain-text file. Excel models, AI agents, underwriting applications, and data pipelines can all work from the same deal record.
AI can extract, classify, and explain. Deterministic code calculates and verifies the financial results.
RENT ROLL units + leasesT-12 operating statementASSUMPTIONS market + underwrittenDEBT terms + sizingCASH FLOWS + returnsNARRATIVE credit storyPROVENANCE who, what, when
Show me a deal
A plain-text record, and the underwriting it produces.
An excerpt from a 48-unit multifamily acquisition. The narrative explains the analyst's adjustments; the JSON block beneath it holds the same numbers in typed form, so any compatible tool can read and recompute them.
Parkview-Apts-Glendale-AZ.uwx.mdVALIDMETRICS RECOMPUTED
The record
plain text · excerpt---
uw_version: "3.0"
deal_name: "Parkview Apartments — Glendale, AZ"
asset_class: "multifamily"
quick_metrics:
purchase_price: 7200000
noi_underwritten: 396635
dscr: 1.109
ltv: 0.70
source_documents:
- "parkview_rent_roll_apr2026.xlsx"
- "parkview_t12_2025.pdf"
---
## Underwritten NOI
Vacancy is normalized to 7% against the
owner's reported 5%. The management fee
drops from 10% to 5.5% of EGI under the
new agreement. Taxes are reassessed to
the post-sale value.
```json uw:section=noi_model
{
"_meta": { "quality": { "confidence": "high" }, … },
"income": {
"vacancy_credit_loss": {
"value": 45864, "rate_applied": 0.07
},
"effective_gross_income": 630936
},
"expenses": { "total_operating_expenses": 234301 },
"net_operating_income": 396635
}
```What a compatible tool shows
underwriting viewParkview Apartments
48 units · 1987 vintage · Glendale, ArizonaPurchase price$7,200,000
UW NOI$396,635
DSCR1.109×
LTV70.0%
| Operating statement | T-12 | UW |
|---|---|---|
| Gross potential rent | $655,200 | $655,200 |
| Vacancy / credit loss | ($32,760) | ($45,864) |
| Other income | $18,000 | $21,600 |
| Effective gross income | $640,440 | $630,936 |
| Total operating expenses | ($228,344) | ($234,301) |
| Net operating income | $412,096 | $396,635 |
DSCR and LTV recomputed by the multifamily calculation pack
The pack derives DSCR, LTV, cap rate, and debt yield from figures stated in the record, using deterministic code rather than a model. The operating statement is stated, and validator cross-checks test it for consistency. The DSCR of 1.109× produces a financial flag in the warning band. Financial flags describe the deal; validation issues describe the record.
What the record holds
The underwriting you already do, written down in a form software can check.
Underwriting assumptions
Market and underwritten vacancy, expenses, growth, and cap rates, with the analyst's rationale beside each one.
noi_model · assumptions · market_analysisDebt and capital
Loan terms, sizing constraints, and capital stacks: the inputs behind DSCR, debt yield, and LTV.
debt_structure · preliminary_sizing · capital_stackCash flows and returns
Projected and dated cash flows, lease-up schedules, and distribution waterfalls behind IRR and equity returns.
dcf · lease_up_schedule · distribution_waterfallCredit narrative
Ordinary Markdown for the story a credit committee reads: thesis, risks, mitigants, and open questions.
Markdown prose · risk_assessmentProvenance
Each block records who or what wrote it, when, from which source, and how confident they were. Revisions supersede earlier blocks instead of erasing them.
_meta · append-only revisionsEach JSON block is a typed section with a published schema. Validators, calculation packs, converters, and conformance fixtures all work against those same sections. See UWX and UW Lite for the compact .uw.md summary representation.
AI never does the financial math
Models read documents. Code does the arithmetic.
The record sits between judgment and computation. Anything that interprets a document can write facts and narrative into it. Only deterministic code turns those facts into NOI, DSCR, or IRR.
- 01 · SourcesDeal documents
Rent roll · T-12 · lease abstracts · loan terms · market assumptions
- 02 · Analyst or AIExtract and explain
Pull facts from documents, classify them, flag gaps, draft the credit narrative
- 03 · Record
.uwx.mdFacts, assumptions, narrative, and provenance in one file both sides read
- 04 · EngineValidate and calculate
Check structure and rules, then compute the metrics a registered calculation pack or verifier defines
- 05 · ResultsRecomputed figures
DSCR · debt yield · LTV · cap rate · verified cash-flow returns, with an optional verification receipt
- 06 · ConsumersEvery downstream system
Excel · credit memos · underwriting apps · APIs · portfolio and data systems
AI is optional. A person, a parser, or an application can write the same record. Whoever writes it, financial results come from deterministic code, not from the model's output.
Receipts prove derivation, not truth. A verified receipt shows the stated results follow from the record under a named pack. It does not show that the rent roll or T-12 behind it is accurate.
Try it
Create and validate a record in two commands.
The published CLI runs with Node.js 18 or later. Nothing to clone, no account.
$ npx @uwmd/cli init --output deal.uwx.md
Created: deal.uwx.md
$ npx @uwmd/cli validate deal.uwx.md
Validation result: WARNINGS — deal.uwx.md
Stage Readiness (workflow completeness):
screening: incomplete
Issues (3):
[WARN] CC-13: the property section does not
state multifamily's primary size field …
[INFO] DQ-06: screening requires debt_structure
[INFO] DQ-06: screening requires validationStarter tools · non-normative
@uwmd/verify reads and verifies a record on its own. The reference tools help you work with it.
Open and vendor-neutral