Skip to content

Lesson 04 · 14 min read

Summarizing Offering Memorandums and Market Reports

Turn an offering memorandum into a standard deal sheet with AI, flag what the marketing leaves out, cross-check the numbers against the rent roll, T-12 and leases, and summarize market reports without mixing facts and forecasts.

An offering memorandum is a sales document. It is written for the seller, by the seller's broker, to get the best price. That does not make it dishonest, but it does mean the numbers on page 4 are the most favorable version of the deal that can be defended. AI can turn 50 glossy pages into a one-page deal sheet quickly. Your job is to make sure the deal sheet shows the deal, not the brochure.

This lesson walks through a four-step OM workflow: get the file readable, extract a standard deal sheet, flag what is missing, and cross-check against the rent roll, T-12 and leases. Then it covers market reports, where the risk is mixing a publisher's forecast with measured facts. A worked example on a Clermont retail center shows an OM NOI that does not match the T-12.

Education, not investment advice. AI summaries and the examples here are teaching tools. Verify every figure against source documents and get advice from your lender, CPA and attorney before you make an offer.

What an OM is and is not

Most OMs carry a disclaimer that says, in substance, the information was obtained from sources believed reliable but is not verified or guaranteed, and the buyer must do its own investigation. Take that literally.

What an OM usually contains:

  • Asking price or a call-for-offers process
  • A stated NOI and cap rate, often labeled "in-place," "year 1," "pro forma," or not labeled at all
  • A rent roll summary, sometimes simplified
  • Tenant profiles and lease abstracts written by the listing team
  • Location, demographic and traffic data
  • Photos, maps and aerials

What it often leaves out or softens:

  • Actual trailing operating history, as opposed to a budget or pro forma
  • Upcoming lease expirations without renewal notice, and termination rights
  • Deferred maintenance and capital needs
  • Below-market or above-market leases
  • Tenant sales or health, for retail
  • Real estate tax reassessment after sale
  • Debt terms if a loan must be assumed, including fees and lender approval

Pro Forma vs. Actuals explains how broker math gets creative. This lesson shows how to use AI to find those spots faster.

Step 1: get the file in and readable

Before any summary, confirm the tool can see the whole document.

  • Upload the PDF to a tool your brokerage approves for the material. Many OMs come under a confidentiality agreement, so read it first (Lesson 7).
  • Ask: "List every section heading with its page number, and tell me the last page you can read."
  • Compare the answer to the actual table of contents. If pages are missing, or the rent roll page comes back empty, the PDF may be scanned images. Run OCR (optical character recognition, which converts page images to text) or type key tables in yourself.
  • Charts and maps rarely extract well. If a number only appears in a chart, read it yourself.

Step 2: extract a standard deal sheet

Use the same deal sheet on every OM. A consistent format makes deals comparable and gaps obvious.

FieldWhat to captureLabel to require
Asking pricePrice, or "unpriced" with offer dateSource page
NOI as statedThe OM's headline NOIIn-place, year 1, or pro forma
Cap rate as statedThe OM's headline cap rateWhich NOI it uses
OccupancyLeased SF ÷ total SFPer rent roll, not marketing text
TenantsName, suite, SF, lease start and end, optionsSource page
RentCurrent rent per tenant, escalationsSource page
Lease typeNNN, modified gross, gross; what landlord paysPer tenant
ExpensesTaxes, insurance, CAM, management, reservesActual, budget or pro forma
RecoveriesWhat tenants reimburse, any capsPer tenant if stated
DebtAssumable loan, balance, rate, maturity, feesOr "none stated"
PhysicalBuilding SF, land, year built, roof and HVAC agesSource page
OtherRights of first refusal, purchase options, pending itemsOr "none stated"
ROLE: Act as an acquisitions analyst for a buyer.

SOURCE: Use only the attached offering memorandum. No outside knowledge.

TASK: Fill in this deal sheet. One row per field. Columns: Field | Value | OM page | Label.

FIELDS: asking price; NOI as stated; cap rate as stated; what NOI the cap rate uses;
occupancy per rent roll; each tenant (suite, SF, start, end, options); current rent
and escalations per tenant; lease type per tenant; each expense line and whether it is
actual, budget, or pro forma; recoveries and caps; debt terms; building SF; land;
year built; roof and HVAC age; ROFR, ROFO or purchase options.

RULES:
- In the Label column use IN-PLACE, PRO FORMA, BUDGET, SELLER CLAIM, or UNLABELED.
- If the OM's marketing text and its rent roll disagree, report both and flag CONFLICT.
- Show the math for any figure you calculate, such as occupancy.
- Write NOT IN SOURCE where the OM is silent.

The CONFLICT instruction is important. Marketing pages say "fully leased" or "100% NNN" while the rent roll on page 18 tells a different story. You want the model to surface that, not smooth it over.

Step 3: flag what is missing

Once the deal sheet is filled in, ask a separate question: what should be here and is not?

Based on the deal sheet and the OM, list:
1. Items a buyer would need to underwrite this property that the OM does not provide.
2. Any figure labeled PRO FORMA or UNLABELED, and what document would show the actual.
3. Leases expiring within 24 months of today's date: [date].
4. Any statement the OM makes that a buyer should verify independently, with page.
Do not speculate about values. Just list the gaps.

Give the model today's date. It may not know it, and lease timing depends on it.

Step 4: cross-check against the source documents

The deal sheet is a list of claims. Each claim has a document that proves or disproves it. Request those documents early, often after an LOI is accepted, and run the same AI extraction on each. Then compare.

OM claimDocument that tests itWhat to compare
NOI and expensesT-12 operating statement, prior-year statementsLine by line, NOI to NOI
Rent and occupancyCertified rent rollTenant by tenant, SF and rent
Lease terms and optionsThe leases and all amendmentsDates, rent steps, options, rights
RecoveriesLeases plus CAM reconciliationsWhat tenants actually reimbursed
TaxesCounty property appraiser and tax collectorCurrent bill and likely reassessment
Physical conditionProperty condition report, roof and HVAC recordsAges and capital needs
DebtLoan documentsBalance, rate, maturity, assumption fee

AI is useful here as a comparison engine. Upload the OM deal sheet and the T-12 and ask for a line-by-line reconciliation with every difference explained or marked UNEXPLAINED. Then verify the explanation yourself. How to Read a Rent Roll and The T-12 and Normalized NOI show what to look for, and Financial Due Diligence covers the full audit.

In-place versus pro forma: read the label

The single most common OM issue is a headline cap rate built on pro forma NOI. Pro forma NOI may include vacant space at assumed rent, scheduled rent increases that have not happened yet, expense cuts, or a lower management fee than the property pays today.

Every dollar in the spread needs an explanation, and every explanation has a cost or a risk. Leasing a vacant suite takes time, tenant improvements, and a leasing commission. A scheduled bump depends on the tenant staying. An expense cut depends on someone actually achieving it. None of that is wrong to model. It is wrong to price as if it already happened without saying so.

Market reports: summarize with citations, separate facts from forecasts

Brokerage research reports, data-platform reports and economic development publications are useful context. AI can summarize them well if you control three things.

Separate reported facts from forecasts. A report might state measured vacancy for last quarter and also predict that rents will rise next year. The first is a fact about the past, as measured by that publisher. The second is an opinion. In a client update, they need different language: "The report states..." versus "The report forecasts..."

Keep definitions and boundaries attached. Publishers define markets differently. One firm's "Orlando industrial" may include different counties, building sizes or property classes than another's. A vacancy figure without its definition can mislead. Ask the model to capture the report's market definition, building inventory threshold and time period alongside every number.

Do not merge reports. Asking AI to "combine these three reports into one vacancy number" produces a figure no publisher stated. If reports disagree, show each one with its source.

SOURCE: Attached report: [publisher], [title], [period].

TASK:
1. State the report's market definition, geography, and property types covered (page).
2. List reported facts for the most recent period: each figure exactly as stated, with page.
3. List forecasts, predictions, and interpretations separately, with page.
4. List any figures that are for a different period or geography than the headline.

RULES: No numbers that are not in the report. No blending with other sources.

Two more cautions:

  • Licensing. Some reports are free and public; some come from paid subscriptions with use limits. Quoting a figure with attribution may be fine, while reproducing tables or charts may not. Check the terms.
  • Dating. Always state the period a figure covers. A report from two quarters ago is not "current." For current market levels, combine published reports with comps from your own brokerage and conversations with lenders and appraisers.

A worked example: the Clermont retail center

Illustrative numbers. Not market data.

A buyer's agent receives an OM for a four-tenant neighborhood retail center in Clermont. Total building area is 12,000 SF. The asking price is $4,250,000. The OM headline reads "7.00% cap rate on NOI of $297,500" with the label "Year 1."

Deal sheet extraction. The AI deal sheet reports:

  • Asking price $4,250,000 (OM p. 2)
  • NOI $297,500, labeled UNLABELED, since "Year 1" does not say in-place or pro forma (OM p. 4)
  • Cap rate 7.00% on that NOI (p. 4). Check: $297,500 ÷ $4,250,000 = 7.00%
  • Occupancy: CONFLICT. The executive summary says "fully leased" (p. 3). The rent roll (p. 11) lists Suite D, 1,300 SF, as "pro forma tenant, $22.00/SF NNN"

That conflict is the first flag. Marketing text calls the center full. The rent roll shows a vacant suite carried at an assumed rent.

Missing items. The gap prompt lists: no trailing operating statement in the OM, Suite B's lease expires in seven months with no renewal notice stated, the roof age is not given, and the taxes shown are the current bill with no note on reassessment after a sale.

Cross-check. After the LOI, the agent receives the T-12. It shows NOI of $268,900. The agent uploads the OM deal sheet and the T-12 and asks for a line-by-line reconciliation. The AI attributes the difference to Suite D's assumed rent. The agent checks the math:

The reconciliation holds: the entire gap is the vacant suite. Other expense lines match within rounding, which the agent confirms on the T-12 itself rather than relying on the AI's statement.

MeasureOM headlineIn-place (T-12)
NOI$297,500$268,900
Cap rate at $4,250,0007.00%6.33%
OccupancyMarketed as full10,700 of 12,000 SF leased, 89.2%

Occupancy check: 12,000 − 1,300 = 10,700 SF, and 10,700 ÷ 12,000 = 89.2%.

What the buyer does with it. The agent reports both numbers to the client with the source for each. Simple math shows that $268,900 at the OM's own 7.00% would be about $3.84 million. That is not a recommended offer. It is a way to show the client how much of the asking price depends on leasing Suite D, which will also take downtime, tenant improvements and a commission. Suite B's near-term expiration goes on the diligence list for an estoppel and a conversation with the tenant. The buyer's offer strategy comes from comps, the lender's view, and the client's goals.

AI produced the deal sheet, the gap list and the reconciliation in minutes. The agent's checks turned them into facts.

Common mistakes

  • Copying the headline cap rate into a client email without the NOI basis.
  • Letting the summary say "fully leased" when the rent roll does not.
  • Skipping the T-12 request because the OM "already has the numbers."
  • Trusting an AI reconciliation without checking the lines yourself.
  • Blending market reports into one figure no publisher stated.
  • Calling a forecast a fact in client communication.
  • Uploading an OM under a confidentiality agreement to a tool the agreement does not allow.

What to take away

  • An OM is marketing; its numbers are claims to verify, not facts
  • Confirm the tool can read the full file, and type in tables it cannot read
  • Use one standard deal sheet for every OM, with a source page and a label for every figure
  • Require the model to flag CONFLICT when marketing text and the rent roll disagree
  • Ask a separate question for what is missing, and give the model today's date
  • Cross-check the deal sheet against the T-12, rent roll, leases, tax records and loan documents
  • Every dollar between pro forma and in-place NOI needs an explanation and a cost
  • Summarize market reports with page citations, separating reported facts from forecasts
  • Keep each report's definitions and period attached to its numbers, and never merge reports into a new figure

Next lesson: Lease Abstraction with AI.

Get Market Insights Delivered

Weekly Central Florida CRE updates — cap rates, new listings, market trends, and investment opportunities. No spam, unsubscribe anytime.