cre-broker-brief
Build a commercial-real-estate brief for an income property (apartments, retail centre, office, industrial) that a broker hands a seller or markets to buyers — NOI-over-cap valuation benchmarked to lo
SKILL.mdskills/cre-broker-brief/SKILL.md
GitHubCRE broker brief (income property)
Value is NOI ÷ cap rate, not revenue × multiple, and there are usually no reviews worth
mining. Write for the broker's chair (win the listing, or market it to buyers), never as
advice to the owner. Pair with flip-deal-feed for the underwriting scripts and
broker-prospecting-dossier for the pre-listing cold approach.
Read the request first: intent drives the deliverable
- "Help me win the listing" -> a valuation / why-now pitch to the seller.
- "Already signed, show buyer upside and comps" -> a buyer-marketing brief.
- A "revenue" figure on a CRE request is usually NOI. Compute
ask ÷ NOI= going-in cap and lead with it.
There is no fixed template. The sections below are ingredients; the brief is the best answer to what was asked.
Data-gathering playbook (the repeatable value)
- County property records: the spine. Most counties run a public assessor app: verified
units, year built, class, land and building size, the county's income-approach value, owner
entity and mailing address, and assessed-value history. Find the county's portal; recipes do
not port between counties.
- Assessed-value history is a trade detector: a 2-3x jump in one year usually means a sale and reassessment. A rename or "new management" dates it.
- Grounded web research (
gemini-web-searchor equivalent) for what records lack:- the last trade (trade press, sold listings). Often paywalled; record it as a diligence lead, never print an unverified buyer name as fact.
- the buyer pool: active local and regional investors in this asset class, named, sourced.
- peer asking rents by unit type: the mark-to-market check.
- submarket cap rates, occupancy, rent growth, supply pipeline from brokerage research.
- Imagery: Street View and Static Maps (Google Maps Platform key). Check that the Street View frame actually shows the building.
The walls: be honest, never fake
- Non-disclosure states (e.g. Kansas, Texas, Utah and others) do not publish sale prices. Use the county's income-approach value as a conservative anchor and point the broker at their own CoStar for trades.
- Many county apps are address-only, so an owner's portfolio can't be enumerated free. The owner
mailing address is the clustering key (see
broker-prospecting-dossier). - Surface the downside. If asking rents already sit above the submarket, say so: the upside is a renovation premium, not blanket increases. Buyers trust the honest read.
The analysis
- Cap-rate football field. Value at a range of local market caps (e.g. 6.0 / 6.5 / 7.0 / 7.5%) against the ask. If the ask's cap is wider than where the submarket trades, the ask is below market value: often the headline.
- Use the right cap for the class and vintage. Older Class-C walk-ups trade well above institutional Class-B caps; using the low figure can double the apparent value created.
- Levered return: state your own assumptions (LTV, rate, amortisation), compute DSCR, cash-on-cash and the value-add equity created. When caps sit below the debt rate, leverage hurts at ask; show the buy-below-ask + cheaper-debt scenario and say plainly that the money is made on price.
- Pre-empt the county value. The buyer will pull the assessed value; explain it (tax-basis floor, non-disclosure, peers assessed lower).
- Value-add leverage: every $1 of NOI growth = $1 ÷ cap of value. Who pays which utility is often a bigger lever than rent (sub-metering / bill-back). Label scenarios illustrative.
- What to request next: rent roll (in-place vs market), trailing-12 operating statement, sold comps from the broker's own data.
The deliverable: hand-authored HTML -> PDF with real charts
| Chart | Type | Carries |
|---|---|---|
| Price per door | horizontal bars | ask vs live value-add comps |
| Cap-rate football field | horizontal bars | value at local caps vs the ask |
| Assessed-value history | line | the trade-detection step-up |
| Valuation anchors | horizontal bars | county floor -> ask -> value-add target |
| Submarket | donut + stat tiles | occupancy (true part-of-whole only), rent growth, pipeline |
| NOI -> value | line | leverage at the cap |
| Levered return | table | going-in vs stabilised (DSCR, CoC, equity) |
| Buyer pool | cards | named active investors, sourced |
Bars, lines and tables for comparisons; a donut only for a true part-of-whole; never a pie for
comps. Charts render only real, stated numbers. Embed images as data URIs so the HTML is
self-contained for the PDF renderer (WeasyPrint or headless Chrome). Version every revision
(v1, v2); never overwrite.
Done when
- It answers the stated intent directly.
- It reads as a broker document: no "you should fix" anywhere.
- Every figure is public data or the broker's own; sources footnoted; nothing invented.
- Valuation is NOI ÷ cap at local caps; levered assumptions labelled as yours.
- Charts match the prose; the property photo verifiably shows the building.
- The honest downside is in it.
- Optional: run it past a broker reader-persona (see
simulate-campaign) before sending.