How to Create a Pro Forma for Real Estate (Beginner to Institutional)

How to Create a Pro Forma for Real Estate (Beginner to Institutional)

Every pro forma answers the same question — what will this property earn? — but not every deal deserves the same size of answer. A thirty-second screen on a listing, a financing check before a loan application, and an institutional underwriting for an investment committee are all pro formas; they differ in how many assumptions they make explicit.

That is why this guide is organized as a ladder rather than a single recipe. There are four levels of pro forma, each one a strict extension of the last: same logic, more resolution. You will create Level 1 versions weekly and Level 4 versions rarely — and knowing which level a given decision requires is itself an underwriting skill. Climb as far as your deal demands.

All figures are illustrative examples, not market data. And if you would rather climb inside a working file, the free starter pro forma has the ladder's core levels pre-wired.

The Inputs You Need Before Any Level

Every rung consumes some subset of the same input list. Gather what your level requires:

• Revenue: scheduled rents (per unit or per lease), a vacancy assumption, other income
• Expenses: the operating cost lines — taxes, insurance, repairs, utilities, management, and the rest — ideally from a trailing-12 statement rather than a guess
• The deal: price, closing costs, any upfront capital budget
• Financing: loan amount or LTV, rate, amortization
• The future (Levels 3–4 only): growth rates, hold period, exit cap rate

The quality hierarchy for sourcing them never changes: actual documents (rent roll, T-12) beat listing claims, which beat rules of thumb, which beat hope. A Level 1 screen built on a real rent roll outranks a Level 4 model built on a broker's pro forma.

Level 1 — The One-Page Screen (Ten Minutes)

The question it answers: is this deal worth another hour of my life?

The Level 1 pro forma is a single stabilized year, top to bottom:

Gross scheduled rent → less vacancy allowance → plus other income → effective gross income → less operating expenses → NOI → and one valuation check: NOI ÷ price = the going-in cap rate.

Illustrative: a fourplex asking $520,000, renting 4 × $1,350: gross $64,800; less 5% vacancy → $61,560; plus $1,440 other income → EGI $63,000; less $27,800 of expenses → NOI $35,200; cap rate 6.8%. Ten minutes, and you know whether the price is even in the conversation for the market.

Level 1's two discipline rules: never skip the vacancy line (100%-occupied screens are how bad deals get second dates), and never model expenses as a lazy percentage if any real expense data exists. Even at one page, the management fee goes in.

What Level 1 cannot see: financing, taxes, capital costs, growth, exit — everything after this single year. It screens; it does not decide.

Level 2 — Add the Financing (Thirty Minutes)

The question it answers: does this deal work with debt on it — and what does the equity earn while I hold it?

Level 2 extends the same page below the NOI line: annual debt service from the loan terms, producing cash flow after debt service, plus the two derived numbers that matter — DSCR (NOI over debt service, the lender's gate) and cash-on-cash return (cash flow over total cash invested, the owner's yield).

Continuing the fourplex: 75% LTV is a $390,000 loan; at an illustrative 6.9% on 30-year amortization, debt service runs about $30,800. Cash flow: $35,200 − $30,800 = $4,400. DSCR: 1.14x — below the 1.20x–1.25x range many lenders want, which is Level 2 doing its job: this price probably doesn't finance at this leverage. Cash invested (25% down + $10,000 costs = $140,000) puts cash-on-cash at 3.1%.

Level 2 is where most rental-property analysis legitimately ends: for a stabilized small property with no exit thesis, the annual snapshot with financing is the decision-relevant model. Its blind spot is time — anything the deal does over years, Level 2 cannot price.

Level 3 — Add the Years and the Exit (An Afternoon)

The question it answers: what is the return over a full hold, including the sale?

Level 3 turns the column into a table: the Level 2 structure projected across the hold period (five to ten years) under explicit growth assumptions — rents and expenses grown at independent rates — with a reversion in the final year: forward NOI capitalized at an exit cap rate, less selling costs and the loan payoff. From the resulting equity cash flow series come the full-hold metrics: IRR (via XIRR) and the equity multiple.

Three creation rules keep Level 3 honest, and they are where beginner multi-year models fail:

1. Reference every assumption from an inputs block. The moment a growth rate is typed inside a formula, the model stops being auditable and starts rotting.
2. Exit conservatively. The convention is to assume the exit cap at or above the going-in cap, so the return is carried by income and the plan, not by a market-timing bet.
3. Grow expenses on their own rate. Symmetric growth silently expands margins forever — an artifact, not a forecast.

Level 3 is the minimum honest level for any deal whose thesis lives in the future: value-add plans, appreciation stories, anything you will hold through a cycle. The complete cell-by-cell construction of this level — revenue stack, expense build, debt module, reversion, returns — is the subject of our full tutorial on building a real estate pro forma from scratch.

Level 4 — The Institutional Model (Buy or Build Over Weeks)

The question it answers: everything a lender, partner, or investment committee will ask — including how wrong can we be?

Level 4 is Level 3 with resolution and accountability added, and it is defined by five capabilities:

• A real revenue engine. Unit-by-unit rent roll with in-place versus market rents for multifamily; lease-by-lease with recoveries and rollover costs for commercial. Revenue as a system, not a cell.
• An economic vacancy stack and a normalization bridge. Physical vacancy, concessions, and bad debt as separate lines; a documented bridge from the T-12 actuals to the pro forma, adjustment by adjustment.
• Dual-constraint debt sizing. Proceeds computed as the lesser of the LTV and DSCR tests, with an amortization schedule carried through the hold — the way lenders actually size loans.
• A sensitivity grid. The 5×5 matrix stressing the two assumptions the return is most exposed to, because a single-scenario model only tells you what happens if you are exactly right.
• Partnership logic where equity is shared. The GP/LP waterfall computing each party's returns from the same cash flows.

Plus the construction standards that make the file defensible: visually distinct inputs, no hardcodes, error-check tie-outs, and written methodology. To see a Level 4 model filled out on a complete deal, tab by tab, the worked pro forma example walks one line by line.

Level 4 is also where the create-versus-buy question becomes real. Building one is among the best educations in real estate finance available and costs weeks of construction and validation; buying one costs a few hundred dollars and an audit of the vendor's transparency. There is no wrong answer — but there is a wrong reason: building at Level 4 to save money usually spends more, in hours and in error risk, than it saves.

Mapping Documents to Inputs

The ladder consumes documents, and knowing which document feeds which input keeps the creation process from stalling:

• The rent roll → the revenue section. Unit-level scheduled rents at Levels 1–2; the full in-place-versus-market mix at Level 4. No rent roll available (a listing-stage screen)? Use the listing's claims, marked as unverified, and treat the eventual rent roll as a gate.
• The T-12 operating statement → the expense build. Actual line items beat every rule of thumb. Missing a T-12, build expenses from components — the assessor's taxes at your price, a live insurance quote, utilities from comparable buildings, management at market — and widen your sensitivity ranges to match the input quality.
• The loan quote or term sheet → the debt module. Rate, amortization, LTV, and the lender's DSCR minimum. Pre-quote, model a deliberately conservative placeholder and label it.
• Comparable sales → the cap rates. The going-in check at Level 1 and the exit assumption at Levels 3–4 both come from adjusted comps, never from the listing's own claimed cap.

The mapping enforces a useful humility: every input traces to a document or is flagged as a guess, and the flags tell you exactly how much to trust the output.

Choosing Your Level

The ladder's practical payoff is calibration — matching the modeling effort to the decision:

Two failure modes bracket the table. Under-modeling — deciding a value-add deal on a Level 2 snapshot — prices none of the plan and most of the hope. Over-modeling — a 10-tab DCF to screen a listing — burns hours and, worse, lends false authority to inputs that are still guesses. A Level 4 model on Level 1 inputs is precision theater; the ladder only works climbed with the data.

The Signature Mistake at Each Level

Each rung has its own characteristic failure, worth naming because they are how pro formas quietly lie at every scale:

• Level 1: the disappearing vacancy line and the missing management fee — the two omissions that make bad screens look like good deals.
• Level 2: testing interest-only payments on an amortizing loan, or computing cash-on-cash on the down payment alone instead of all cash invested — both inflate the yield.
• Level 3: the exit cap set equal to (or below) the going-in cap, and revenue growing faster than expenses forever — the two assumptions that manufacture returns out of convention.
• Level 4: precision theater — a beautiful model on guess-grade inputs, its ten tabs lending false authority to numbers no document supports.

Notice the pattern: the mistakes get subtler as the models get bigger. A Level 1 error is visible on the page; a Level 4 error hides in an assumption cell. Which is one more argument for climbing the ladder in order — you learn each level's honesty rules while they are still easy to see.

Frequently Asked Questions

What should a simple real estate pro forma include? At minimum (Level 1): scheduled rent, a vacancy allowance, other income, line-item operating expenses including management, NOI, and the cap rate against the price. Add financing (Level 2) the moment a loan is involved.

How do I create a pro forma for a rental property with no operating history? Build the revenue from market comps and the expenses from component estimates (taxes from the assessor at your price, an insurance quote, utilities from similar properties, management at market) — and label the whole file as estimate-grade. New-build or no-history pro formas deserve wider sensitivity ranges, not more decimal places.

What is the difference between a pro forma and a budget? A budget plans the operations of a property you own; a pro forma projects the investment performance of a deal — it includes the acquisition, the financing, and (at Levels 3–4) the exit. The operating budget is roughly the expense section of the pro forma, extended into management detail.

Can I create a pro forma in Google Sheets instead of Excel? Levels 1–3, comfortably — the formulas are identical. Level 4 files circulate among counterparties who expect Excel, and the institutional models you might buy or receive will be .xlsx; fluency there is the practical default.

How accurate does a pro forma need to be? A pro forma is not a prediction to be graded; it is a set of explicit assumptions to be tested. Its job is to be honest — sourced inputs, conservative conventions, stress ranges — so that the decision it supports survives the assumptions being somewhat wrong. Accuracy is the market's job; discipline is yours.

Start on the Ladder

Two ways in. The free starter pro forma gives you Levels 1 through 3 in one working file — drop in a deal and climb. When a deal demands Level 4, The Multifamily Sheet is the institutional implementation: the full rent-roll engine, T-12 bridge, dual-constraint debt sizing, 10-year DCF with reversion, GP/LP waterfall, and 5×5 sensitivity grid — fully unlocked, formula-transparent, with a documented methodology PDF.

This article is for educational purposes only and does not constitute investment, legal, or tax advice. All figures are illustrative examples, not market data. Consult qualified professionals before making investment decisions.