Real estate investment analysis is the work of turning a deal packet into figures a team can check, stress, and argue about. A one-page flyer can start that work. It cannot finish it. The rent roll, the trailing twelve-month statement, and the offering memorandum still have to be read against each other before anyone treats a return as useful.
Create an account and start underwriting
This guide is for acquisition teams, brokers, owner-operators, and fund analysts who already get those files. It walks the stages from source documents to a decision-support packet. It is not a buy or sell recommendation, and it is not tax, accounting, or investment advice. A model can be wrong. A person still has to compare it with the files it came from.
What real estate investment analysis is for
The job is narrower than "is this a good investment?" A useful analysis answers a smaller set of questions. What does the property earn and spend today? Which of those figures are in the documents, and which are assumptions? What changes if occupancy, expenses, or the exit price move? Who has to sign off, and what did they actually see?
Ranking pages for this topic often stop at cap rate, cash-on-cash return, and a neighborhood tour. Those metrics matter. They are also easy to compute from a flyer that nobody has reconciled. A cap rate built on a broker's projected net operating income is a different fact from a cap rate built on a checked trailing twelve months.
Keep the two layers separate:
- In-place results come from the rent roll and the trailing twelve-month statement. They describe what the property has been doing.
- Forward assumptions cover growth, vacancy, capital items, financing, and the exit. They describe what the team thinks happens next.
- The decision is a person's judgment about whether those assumptions are acceptable for this mandate. Software does not make it.
Appraisal practice draws a similar line. The Appraisal Institute's Standards of Professional Practice set requirements for developing and reporting analyses, opinions, and conclusions, including a distinct standard for review. An acquisition memo is not an appraisal. The habit is still the right one: show the development of the figures, then let a reviewer test them.
Start with the source documents, not the summary
Most multifamily packets arrive in the same order. A flyer or teaser comes first. An offering memorandum follows, often with a rent roll and a trailing twelve-month operating statement. Leases, a tax bill, or a capital schedule may show up later, or not at all.
Treat the flyer as a starting point. It usually names the price, the unit count, and a headline return. It rarely shows the unit mix, the loss-to-lease, or which expense lines were left out. The offering memorandum explains the story. The rent roll and the trailing twelve are where the story has to match the numbers.
A practical first pass asks four questions before any model is built:
- Do the unit count and the rent-roll rows describe the same property?
- Does in-place rent on the roll support the income line on the trailing twelve?
- Are concessions, bad debt, and vacancy shown, or buried in a net rent figure?
- Which expense lines are missing, thin, or obviously one-time?
If the answer to any of those is "we cannot tell," the analysis is not ready for a return. It is ready for a request. Ask the broker for the missing page. Do not invent the line to keep the model moving.
Slung's underwriting path is built for this intake. A team can forward a flyer, an offering memorandum, a rent roll, or a trailing twelve to sam\@slung.ai, or upload Excel, CSV, and PDF files in the app. Documented uploads include exports from systems such as Yardi, RealPage, AppFolio, Entrata, MRI, and Rent Manager. A flyer can open a draft. It does not fill in rent-roll or trailing-twelve figures that were never sent. New email senders may be guided to sign up, and email underwriting can be limited by plan.
The product documentation is the place to confirm how upload, email intake, and review actually work. Marketing lines about "any format" or "minutes later" are not guarantees. Scans and handwriting need a closer look, and unclear files should be sent again.
Reconcile every figure back to a page
Extraction is not the same as agreement. A model can pull a number from a PDF and still put it on the wrong line. The review step is where someone opens the source and checks the classification, the unit mix, and the period.
Work in a fixed order so the review does not wander:
- Identity. Property name, address, unit count, and statement period should match across the flyer, the rent roll, and the trailing twelve.
- Income. Gross potential rent, loss-to-lease, vacancy, concessions, bad debt, and other income should each have a home. A single "net rental income" line hides the bridges a buyer needs.
- Expenses. Payroll, utilities, repairs, insurance, and taxes should be tied to a named line, not a leftover bucket. Watch for a tax or insurance figure that is a partial year.
- Below the line. Capital items, debt service, and reserves are not operating expenses. Leaving them inside operating expenses makes net operating income look worse than the operation, and the reverse mistake makes it look better.
Human review is required. Slung tells users to correct classifications and unit mix, and to check extracted figures against the source, especially when a scan is faint or a note is handwritten. The terms also say evaluations can be inaccurate and are not professional financial, tax, or investment advice. That limit belongs in the workflow, not in a footnote nobody reads.
Recordkeeping has a separate, narrower job. The IRS explains that rental owners must report rental income and generally may deduct associated expenses, and it publishes tips on rental income, deductions, and recordkeeping. An acquisition model is not a tax return. It should still keep the income and expense support a later owner, lender, or tax adviser would ask to see. Do not treat a model total as a tax figure.
Separate what happened from what you assume
Once the in-place numbers agree with the files, the analysis can move forward. This is the step most flyers skip. They publish a stabilized return as if it were a historical result.
Label each driver before you change it:
- Observed. In-place rent, current occupancy, and trailing expenses, each tied to a page.
- Normalized. A documented adjustment, such as removing a one-time legal bill or annualizing a tax bill that covers ten months. Write the reason next to the change.
- Assumed. Rent growth, expense growth, vacancy, renovation cost, loan terms, hold period, and exit cap rate. These are choices. They are not facts in the offering memorandum just because the broker used them.
Slung's property dashboard is where a team edits those choices after extraction. Documented controls include cap rates, growth, financing, and renovations, with a re-run after the edit. Expense benchmarks and rent comps can sit beside the deal as context. They are inputs to a judgment, not a verdict that the rents or costs are "right."
Keep the pro forma assumptions list itself for the people who own that page of the process. This article only needs the boundary. If a figure is not in a source document, it is an assumption, and the packet should say so in the same row.
Run a few scenarios, then stop
Scenario review is where analysis earns its name. One case is a story. Three cases show what the story depends on. Ten cases usually show that nobody has decided which risk matters.
A small set is enough for a first investment-committee conversation. Read each row as one test, not as a combined forecast.
| Case | What you change | What you leave alone | What the packet should show |
|---|---|---|---|
| In-place. | Nothing beyond reconciled trailing figures. | Current rent, occupancy, and expenses. | The return if today simply continues. |
| Base. | The team's stated growth, vacancy, and exit. | The same debt and price as the ask. | The return the memo is actually proposing. |
| Downside. | One worse occupancy or exit cap, not both piled on at random. | The base case's other assumptions. | How far the return moves when that one bet is wrong. |
| Upside. | One better rent or expense outcome you can point to. | The base case's other assumptions. | What has to go right, and whether the files support it. |
Change one driver at a time when you are testing sensitivity. A downside that cuts rent, raises every expense, and widens the exit cap in the same breath is a mood, not a test. The reader cannot see which assumption did the damage.
Write the result in plain differences. A wider exit cap drops the levered return from X to Y, using the base case's rent growth. That line is more useful than a wall of internal rates of return with no trail. If you do not have the support for a number, leave the cell blank and say why.
Financing deserves the same restraint. Name the rate, the term, the amortization, and whether the proceeds are assumed or quoted. A quoted term sheet is a source. A rate typed in because it "looks current" is an assumption. Do not present either one as a commitment from a lender.
Build a decision-support packet, not a verdict
The last stage is packaging. A decision-support packet lets a colleague who was not in the model rebuild the conclusion from the files. It does not tell them what to do.
A packet that holds up in a meeting usually contains:
- The source list, with the file name and the period each statement covers.
- A short reconciliation: unit count, income bridge, and any expense line you reclassified, each with the page it came from.
- The in-place, base, downside, and upside cases, with one sentence on what changed.
- Open items. Missing leases, a stale tax bill, or a rent roll that does not match the unit count belong here, not in an appendix nobody opens.
- The limit. State that the figures are informational, can be wrong, and are not an appraisal or an investment recommendation.
Slung can help assemble the working file. After review, the dashboard model can be exported to an editable Excel workbook for offline scenarios, and exported again after the web model changes. That workbook is specific to the deal. It is not the free blank template. Report links can be view-only or editable, and paid plans can use a confidential link that hides selected assumptions so the recipient fills in their own. Inbound email in the document vault is not shared with recipients.
Brokers on a Principal plan can also use a DataRoom so an invited buyer supplies assumptions and receives a separate report. That buyer is a collaborator on the packet. They are not automatically a Slung customer, and the room's NDA step is a placeholder, not evidence of a signed agreement.
None of this is the decision. The committee, the buyer, or the owner still chooses. If the packet cannot be checked back to the rent roll and the trailing twelve, it is not ready to support that choice.
Where an underwriting aid fits, and where it stops
Use a tool where the work is repetitive. Reading a rent roll into rows, grouping expense lines, and keeping an editable model in sync with the latest file is that kind of work. Choosing the hold period, the renovation scope, or whether the basis fits the fund is not.
For the underwriting product, the documented path is straightforward. Create an account, choose Free Analyst, Principal monthly, or Principal annual, and upload the files or email them to Sam. A trial exists only when one is offered. Public pricing is "call for pricing." There is no published seat price or deal allowance to cite. And Principal annual is a real term, so "cancel anytime" is not a blanket promise. Paid subscriptions generally renew unless notice is given at least 30 days before the term ends.
Teams that want the analysis steps built into their own process, rather than a shared product workflow, are in a different conversation. Slung Enterprise scopes that work with the client: a first project, a build on the firm's actual deals and documents, then later improvements. Examples on the site, including a ledger turned into a trailing twelve, are examples. They are not a fixed package, and ownership follows the signed agreement.
A blank workbook is a third, smaller door. The free multifamily Excel model is a v53 file with blank property inputs. It is not the app, not a Principal plan, and not a finished analysis of any deal.
Request a demo if you want to see the underwriting workflow on a real packet, or talk with the team about pricing before you pick a plan.
Frequently asked questions
What is real estate investment analysis?
It is the staged review of a property's documents, in-place results, and forward assumptions so a person can judge the deal. It organizes evidence. It does not itself decide whether to buy, sell, or hold.
Which documents do you need before you trust a return?
Start with the flyer if that is all you have, then require a rent roll and a trailing twelve-month statement before you treat income or expenses as in-place. An offering memorandum explains the broker's case. It does not replace those two files.
How is this different from underwriting?
Underwriting is the modeling work inside the analysis: standardizing income and expenses, setting assumptions, and testing the result. Analysis is the full path from the source packet to a decision-support memo a colleague can review. The underwriting can be excellent and the packet can still be unfit if the sources were never checked.
Can software complete the investment decision?
No. Software can extract figures, keep scenarios in one model, and export an editable workbook. The user still has to correct classifications, compare the output with the source documents, and make the call. Generated figures can be inaccurate, and they are not professional investment advice.
How many scenarios are enough?
For a first review, show an in-place case, the base case you are willing to defend, and one downside that changes a single important assumption. Add an upside only when the files support the better outcome. More cases help only if each one names the driver that moved.
