The Modeling Test: Single-Property DCF Build

Introduction

The single-property DCF is the most common real estate modeling test, handed out in everything from a 45-minute timed exercise to a multi-hour case, and candidates routinely misjudge what it is grading. It is not a search for a perfect IRR to two decimals. It is a check on whether you can build a clean, auditable model, whether you know which handful of assumptions actually drive the returns, and whether you can defend your choices when an interviewer pokes at them. A messy model with the right answer loses to a clean model with a defensible one. Understanding the build sequence, the few inputs that matter, and the traps that quietly cost points is what separates a strong performance from a frantic one.

The interviewer is evaluating three things at once, and only one of them is arithmetic. First, structure: is the model laid out so that assumptions are separated from calculations, labeled, and easy to follow? Second, judgment: do you know that the exit cap rate, the rent growth assumption, and the leverage are what move the answer, while a rounding choice on reimbursements is noise? Third, composure: can you finish a coherent model in the time given rather than a beautiful half-finished one? A candidate who races to fill cells without a plan usually produces something neither clean nor complete.

This framing also tells you where to spend marginal minutes: on the assumptions that drive returns and on a working returns calculation, not on modeling vacancy unit-by-unit when a percentage will do. The conceptual underpinning of the property DCF is covered in the real estate DCF article, and the method-selection logic in how to value a single property.

The core of the model is a top-to-bottom build from gross revenue down to the cash that reaches equity. The sequence is always the same, which is what makes it learnable.

The NOI line deserves care because every downstream number scales off it; the mechanics of building it cleanly are in NOI and property cash flow. Keeping unlevered and levered cash flows on separate lines, rather than collapsing them, is also what lets you show both the property return and the equity return without rebuilding.

Two parts of the model carry more weight than their size suggests: how the debt is sized and how the exit is set. Lenders do not simply lend a fixed percentage; they size to the tightest of several constraints, and a strong model reflects that.

Debt Service Coverage Ratio (DSCR)

DSCR is net operating income divided by total debt service (interest plus principal), measuring how comfortably the property's income covers its loan payments. Lenders typically require a minimum of 1.20x to 1.30x, meaning NOI must exceed debt service by 20 to 30 percent, and the loan is sized down if that test binds.

In practice the loan amount is the lowest figure that satisfies all of the lender's tests at once, so a clean model checks each and takes the binding constraint.

The exit is the other high-leverage input. The sale price is the forward NOI capitalized at an exit cap rate, and the single most common error is to apply the cap rate to the current NOI rather than the year-after-sale NOI, or to forget that the exit cap rate is usually set above the going-in rate. The debt metrics and how they interact are covered in property debt metrics: LTV, DSCR, and debt yield.

With the cash flows and exit in place, the returns are mechanical, but knowing what each one tells you is part of the grade. The levered IRR captures the time-weighted return to equity including the exit; the equity multiple, or money-on-money, shows total dollars returned per dollar invested without regard to timing; and cash-on-cash shows the annual income yield on equity before any exit. A strong candidate quotes the IRR and the equity multiple together, because a high IRR on a quick flip and a high multiple on a long hold are very different stories. The full returns framework, including how a promote splits them between partners, is covered in IRR, equity multiple, and the real estate waterfall.

The way you work the test is itself being judged. Open by reading the prompt for the given assumptions and the required outputs, then build the skeleton end to end before refining. Label your assumptions in one block so the grader, and you, can change a driver and watch the answer move. When you finish, run two quick sanity checks: does the going-in cap rate implied by your purchase price look like a real market rate, and does the levered IRR sit in a plausible range for the risk?

In the end the DCF test is as much a communication exercise as a spreadsheet one. The grader wants a model they could pick up and audit, driven by a few well-chosen assumptions, ending in returns you can defend. That same discipline scales up to the REIT 3-statement and NAV modeling test at the entity level, and it is the habit that carries straight onto the desk, where every model you build is handed to someone else to check.