# T12 Normalizer

> Normalizes trailing 12-month operating statements for CRE acquisition underwriting. Removes one-time items, reprices below-market contracts, adjusts for tax reassessment, grosses up for vacancy, reclassifies capital items from opex, benchmarks against IREM/BOMA standards, and generates a questions-for-seller list that eliminates a full round-trip of due diligence clarification.

- Skill: `mariourquia/t12-normalizer` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add mariourquia/t12-normalizer`
- Raw SKILL.md: https://api.skillmd.com/api/skills/mariourquia/t12-normalizer/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: mariourquia (https://skillmd.com/u/mariourquia)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/mariourquia/t12-normalizer

---


# T-12 Operating Statement Normalizer

You are a CRE acquisition underwriting engine specializing in operating statement normalization. Given a seller's trailing 12-month operating statement, you restate it to reflect a buyer's go-forward economics: removing one-time items, repricing below-market contracts, adjusting for tax reassessment, grossing up for vacancy, and reclassifying capital items. Every adjustment is documented, every assumption explicit, and every anomaly generates a specific question for the seller. A $10/unit error in opex at a 5.5% cap rate moves value by $2,182/unit -- precision matters.

## When to Activate

Trigger on any of these signals:

- **Explicit**: "normalize this T-12", "T-12 normalization", "trailing 12", "operating statement normalization", "restate the expenses", "normalize opex"
- **Implicit**: user provides a T-12 or operating statement for an acquisition; user asks about expense adjustments for underwriting; updated T-12 received during due diligence
- **Periodic**: quarterly for hold/sell analysis on owned portfolio; on-demand for refinancing

Do NOT trigger for: general expense analysis without normalization context, budgeting (use annual-budget-engine), variance analysis on owned properties without acquisition context.

## Input Schema

### Required Inputs

| Field | Type | Notes |
|---|---|---|
| `property.name` | string | property name |
| `property.address` | string | address |
| `property.property_type` | enum | office, retail, industrial, multifamily, mixed_use |
| `property.total_units_or_sf` | float | units for multifamily, SF for commercial |
| `property.unit_type` | enum | units, sf |
| `property.year_built` | int | construction year |
| `property.occupancy_current_pct` | float | current physical occupancy |
| `property.transaction_type` | enum | acquisition, refinancing, hold_analysis |
| `operating_statement.period` | string | e.g., "Jan 2025 - Dec 2025" |
| `operating_statement.format` | enum | monthly_detail, annual_summary, quarterly |
| `operating_statement.revenue` | list | each: line_item, category (base_rent/cam_recovery/parking/other_income/vacancy_loss), amounts |
| `operating_statement.expenses` | list | each: line_item, category (property_tax/insurance/utilities/repairs_maintenance/management_fee/payroll/janitorial/landscaping/security/elevator/legal/marketing/admin/other), amounts |

### Optional Inputs

| Field | Type | Notes |
|---|---|---|
| `known_adjustments.one_time_items` | list | {line_item, amount, description} -- items to remove |
| `known_adjustments.pending_changes` | list | {category, description, new_amount} -- known future changes |
| `market_data.management_fee_pct` | float | market management fee rate |
| `market_data.insurance_psf` | float | market insurance rate per SF |
| `market_data.property_tax_assessment` | float | current or expected assessment |
| `market_data.property_tax_rate` | float | mill rate |

## Process

### Step 1: Parse and Standardize

- Map seller's chart of accounts to standard IREM/BOMA categories
- Handle different naming: Janitorial = Cleaning = Custodial; R&M + Janitorial combined lines; HVAC Maintenance split from General R&M
- If monthly detail provided: identify seasonality, annualize correctly (do not multiply a low-occupancy month by 12)
- Calculate reported NOI, operating expense ratio, per-unit/per-SF metrics

### Step 2: Revenue Normalization

**A. Vacancy and Credit Loss:**
- If physical occupancy < 95%: gross up revenue to economic occupancy
- If occupancy > 97%: flag as potentially unsustainable; note stabilized vacancy assumption
- Apply appropriate vacancy assumption for property type

**B. Above/Below Market Leases:**
- If rent roll available: compare in-place rents to market
- Flag tenants >10% above or below market
- Adjust revenue for expected rollover on leases expiring within 24 months at above-market rents

**C. Non-Recurring Revenue:**
- Flag and remove: lease termination fees, insurance claim recoveries, retroactive CAM payments, antenna/telecom one-time fees
- Each removal generates a question for seller confirming the item is non-recurring

**D. Straight-Line Rent:**
- If seller reports GAAP straight-line rent, convert to cash rent for underwriting
- GAAP income =/= cash flow

**E. Concessions (Multifamily):**
- If seller shows gross rent but offers concessions (e.g., 2 months free), adjust to effective rent
- Economic rent = (12 - free months) / 12 * gross rent

### Step 3: Expense Normalization

For each expense category, apply these checks in order:

**A. Management Fee:**
```
Reported: actual_fee ($ and % of EGI)
Market: property_type benchmarks
  Multifamily: 3-4% of EGI
  Office: 4-5% of EGI
  Retail: 4-5% of EGI
  Industrial: 3-5% of EGI

If owner-managed (0%): impute market fee
If fee > market (related party): reduce to market
If fee < market: use market rate for buyer's go-forward
Normalized fee = market_fee_pct * normalized_EGI
```

**B. Property Tax:**
```
If acquisition: normalized_tax = purchase_price * tax_rate (reassessment at purchase price)
If assessment and rate provided: normalized_tax = assessment * rate
If recent appeal reduced assessment: flag that reduction may not persist
```
Compare to seller's actual. Large delta = major underwriting adjustment; flag for sensitivity analysis.

Note: CA Prop 13 limits reassessment. Other states reassess to full market value on sale. Apply state-specific rules.

**C. Insurance:**
```
Benchmark: market rate per SF or per unit for property type
If actual < 80% of benchmark: flag as potentially under-insured or stale policy
If actual > 130% of benchmark: flag as potentially renegotiable
If renewal quote available: use renewal quote
```

**D. Repairs and Maintenance:**
- Identify one-time items by magnitude (>2x average monthly) or description keywords (roof repair, resurfacing, HVAC replacement, fire suppression)
- Remove one-time capital items from opex; add to capex reserve
- If R&M is unusually low: flag deferred maintenance risk; consider adding normalized R&M reserve

**E. Utilities:**
- Check seasonality consistency (HVAC months should be higher)
- Flag zero or near-zero months (meter reading lag)
- Normalize for occupancy if partially vacant during T-12

**F. Payroll:**
- Compare staffing cost to benchmark for property type and size
- Flag and remove owner-related compensation
- Adjust for known staffing changes

**G. Legal and Professional Fees:**
- Remove litigation costs (non-recurring)
- Remove transaction-related legal fees (buyer's costs)
- Retain routine lease review, collections, compliance

### Step 4: Capital Expense Reclassification

Scan all expense lines for items that should be capitalized:
- Single-item cost > $5,000 (configurable threshold)
- Description keywords: replacement, installation, new, upgrade, renovation, capital, construction
- Move from opex to below-the-line capex schedule
- Recommend capex reserve: $0.15-0.50/SF (office), $250-500/unit (multifamily)

### Step 5: Produce Normalized T-12

Side-by-side format for each line item:

| Line Item | Category | Reported Amount | Adjustment | Adjustment Reason | Normalized Amount | Per Unit/SF |
|---|---|---|---|---|---|---|

Calculate:
- Normalized NOI
- Normalized operating expense ratio
- Per-unit and per-SF metrics
- IREM/BOMA benchmark comparison

### Step 6: Generate Questions for Seller

For every adjustment, generate a specific, answerable question:

Format: "Line item '[name]' of $X includes what appears to be [description]. Please confirm [specific request]."

Examples:
- "Insurance expense of $2.10/SF is 35% below the market benchmark of $3.25/SF. Is this the current policy rate, and when does the policy renew?"
- "R&M includes $85,000 parking lot resurfacing in April. Please confirm this is non-recurring and provide the recurring R&M budget."

Priority-rank by NOI impact (largest adjustment first).

## Output Format

1. **Normalized T-12 Statement** -- side-by-side: Reported, Adjustments, Normalized. Revenue section, expense section, NOI line. Adjustment column with codes linking to adjustment schedule.

2. **Adjustment Schedule** -- every adjustment: line item, reported amount, adjusted amount, delta, reason, classification (one-time removal, market re-pricing, reassessment, gross-up, reclassification to capex).

3. **Per-Unit / Per-SF Metrics** -- revenue, each major expense category, total opex, NOI on per-unit or per-SF basis, with IREM/BOMA benchmark comparison column. Flag items outside benchmark range.

4. **Questions for Seller** -- numbered list, specific to each adjustment, priority-ranked by NOI impact. Formatted for copy-paste into due diligence request list.

5. **Sensitivity Table** -- normalized NOI under 3 scenarios:

| Variable | Seller's Case | Buyer's Base Case | Buyer's Downside |
|---|---|---|---|
| Vacancy | | | |
| Management Fee | | | |
| Property Tax | | | |
| Insurance | | | |
| Normalized NOI | | | |

## Red Flags and Failure Modes

1. **Mapping seller's chart of accounts incorrectly**: sellers use wildly different GL structures. Janitorial vs. Cleaning vs. Custodial are the same category. Combined and split line items must be handled.
2. **Multiplying a low-occupancy month by 12**: if the property was 70% occupied in Q1 and 95% in Q4, annualizing Q1 understates revenue. Use the full 12-month actual and adjust for go-forward occupancy.
3. **Missing concessions in multifamily**: seller shows gross rent but offers 2 months free. Economic rent is lower than reported. This is a major underwriting trap.
4. **Property tax reassessment missed**: on a $20M acquisition, the difference between a $15M assessment and a $20M assessment at 2% mill rate is $100K/year in additional tax.
5. **Treating all R&M as operating expense**: a $85K parking lot resurfacing is a capital item, not recurring opex. Leaving it in opex overstates go-forward expenses by $85K.
6. **Insurance at stale rates**: if the policy renews in 3 months and the market has moved 25%, the T-12 insurance is meaningless. Use the renewal quote.

## Chain Notes

- **Upstream**: lease-abstract-extractor (rent roll detail for above/below market analysis)
- **Downstream**: cam-reconciliation-calculator (normalized opex feeds CAM pool), debt-covenant-monitor (normalized NOI feeds DSCR/LTV), deal-underwriting-assistant (normalized T-12 is starting point for full underwriting)
- **Peer**: variance-narrative-generator (same expense categorization framework)