Real Estate Underwriting: How to Pressure-Test a Deal Before You Commit
A model that looks good on paper is not the same as a deal that holds up in the real world. The distance between those two things is where money gets lost.
Underwriting is not data entry. Plugging numbers into a spreadsheet produces a spreadsheet. What it does not produce is an honest picture of whether a deal actually works under conditions that do not cooperate with your assumptions. That requires asking harder questions than most people are willing to ask before they fall in love with a deal.
On the acquisition side, the rent comps deserve more scrutiny than they typically get. Are they actually comparable — same finishes, same concessions, same market timing, or are they the most favorable examples someone found to justify a number they already wanted? The T-12 needs to be read critically, not accepted at face value. Deferred maintenance, owner-paid expenses, and revenue that was temporarily inflated all hide inside trailing financials that look clean until someone looks carefully. Expense benchmarks need to be tested against market reality, not optimism. Property taxes need to be modeled post-sale, because the assessment that worked for the seller's basis will not be the assessment you inherit after a recorded transaction. And the returns need to be stress-tested, what happens when interest rates move, when cap rates expand, when rent growth slows below projection? A deal that only works under best-case assumptions is not underwritten. It is hoped.
Development deals carry all of those questions plus a longer list. Hard and soft cost estimates need to come from real bids or verified comparable projects, not ballpark figures that will be corrected by a GC after you are already committed. Entitlement and permitting timelines need to be grounded in the actual jurisdiction and the actual project type, not an idealized schedule that assumes nothing goes sideways with the city. Lease-up velocity needs to be supported by absorption data from the submarket, not by the belief that a well-located project will fill itself. Exit cap assumptions need to reflect what nearby comparable deliveries are actually trading at, not what the market was doing when the proforma was first built. And every development model needs a clear answer to the question that matters most: what happens when costs go up or the timeline slips? Because they will. The only variable is by how much.
A properly underwritten deal produces more than a number. It produces a defensible position, a custom financial model built on sourced assumptions, investor-ready materials that can withstand a sophisticated buyer's questions, and a backup report that explains the logic behind every line item. That is the standard for raising capital seriously and for making acquisition or development decisions that hold up after the enthusiasm of the initial opportunity wears off.
The deals that perform are almost always the ones where someone did the uncomfortable work of pressure-testing the model before committing. The deals that blow up usually had a model that looked great and assumptions that nobody challenged.
If you are working through an acquisition or development project and need the numbers run correctly, that is exactly the kind of work I do. Start with fifteen minutes at calendly.com/jeph-reit.