How is a commercial real estate pro forma different from a residential one?

A residential pro forma usually models a single rent stream against a flat operating-expense ratio. A commercial pro forma models a rent roll — multiple tenants, each with their own commencement date, escalation schedule, recovery treatment, free-rent period, and renewal option. The expense side splits into recoverable and non-recoverable buckets, the capital side tracks tenant improvements and leasing commissions on every rollover, and the cash flow gets stress-tested against debt coverage on a quarter-by-quarter basis. Lumping retail templates onto a 200,000 SF office building is where deals get mispriced.

What software do institutional CRE teams use for pro formas?

For commercial assets with lease-level complexity, Argus Enterprise is still the institutional default — most lenders, appraisers, and IC packages expect Argus output. For multifamily, hotel, and self-storage, Excel models dominate because the lease structure is simpler and the underwriting customizations are heavier. Most institutional shops at the Blackstone, KKR, Brookfield, and Nuveen level run both: Argus for the DCF, Excel for the sensitivity work, sources-and-uses, and waterfall. RealPage, Yardi, CoStar, and RCA sit upstream of the pro forma supplying market data and operating benchmarks.

What's the difference between a pro forma and a DCF in commercial real estate?

The pro forma is the projected cash flow stream — year-by-year revenue, expenses, NOI, capital, and reversion. The DCF is the valuation engine that consumes that cash flow stream and discounts it back to present value at a target rate. Argus produces both inside the same file, which is why people use the terms interchangeably. They are not the same. The pro forma is the assumption set. The DCF is what those assumptions are worth today.

What does a development pro forma include that an acquisition pro forma doesn't?

Five things: a sources-and-uses table showing how the capital stack funds construction; a development budget with hard costs, soft costs, financing costs, developer fee, and contingency; a construction draw schedule with capitalized interest; a lease-up absorption curve from certificate of occupancy through stabilization; and a yield-on-cost calculation comparing untrended stabilized NOI to total project cost. Development pro formas also typically include a refinance or sale assumption at stabilization to take out the construction loan.

What are the biggest errors in commercial real estate pro formas?

Five recurring ones. Underwriting market rents without checking the lease roll schedule. Assuming flat exit cap rates instead of building in 25-75 basis points of decompression. Ignoring real estate tax reassessment at the buyer's purchase price. Treating TI and leasing commissions as below-the-line capital instead of recurring drags on cash flow. And failing to model co-tenancy clauses, exclusive use provisions, or termination options that materially change the income stream. None of these are exotic. All of them are missed regularly.

How accurate is a 10-year CRE pro forma?

It depends what you mean by accurate. The line items are almost never right in any individual year — rents do not grow linearly, expenses do not inflate at exactly 2.5%, and vacancy does not stay flat. What a good pro forma does is establish a defensible mid-case that the deal can be evaluated against. The real test is whether the deal still clears the IRR threshold under a downside case with 150 basis points of rent growth haircut, six months of lease-up delay, and 50 basis points of exit cap expansion. If it only works in the base case, the margin is illusory.

Can AI build a commercial real estate pro forma?

Not the whole thing, and not the parts that matter. The judgment calls — vacancy assumptions, rent growth, expense load, exit cap, capital reserves — still belong to the underwriter. Where AI changes the workflow is upstream: extracting the rent roll into structured data, parsing the T-12 into normalized operating line items, abstracting each lease for the terms that affect the income stream, and flagging items that don't match what the seller is representing. That's where most of the analyst-hours on a pro forma actually go, and it's where most of the errors enter.

By Jeff Axelrod · May 19, 2026

Commercial Real Estate Pro Forma: The Practitioner's Guide

How institutional CRE pro formas actually get built — line items, a 10-year office worked example, Argus vs. Excel, and where AI changes the workflow.

Every commercial real estate deal I worked on in 10 years on the buy side at Stockbridge Capital — closing more than $2B over that stretch — lived or died on a pro forma. The broker sent one in the offering memorandum. Our analyst tore it apart and rebuilt it from the rent roll up. The lender ran their own version against ours. The appraiser produced a third. By the time the deal reached investment committee, four people had built variants of the same spreadsheet, and the IC discussion was almost entirely about which assumptions were defensible.

That is what a CRE pro forma actually is. It is not a forecast. It is a structured argument about what an asset will produce, built line by line from assumptions each of which has to survive being questioned by someone who has done this longer than you have.

This guide is the commercial-specific complement to our general real estate pro forma guide. That piece covers the universal mechanics — gross potential rent, vacancy, NOI, reversion — across multifamily, office, retail, and industrial. This one drills into what changes when the asset has commercial leases: recoverable expenses, tenant improvements and leasing commissions, co-tenancy, lease-by-lease modeling, and the Argus-vs.-Excel tooling question every commercial team has to answer. If you want the multifamily worked example, read that first. If you want the commercial-specific layer, you’re in the right place.

What a commercial pro forma actually models

A commercial pro forma is a multi-year projection of the cash flows produced by a property whose income comes from non-residential leases. The mechanics differ from multifamily in three structural ways:

Income is rent-roll-driven, not unit-driven. A multifamily pro forma can model 300 units at an average rent and a vacancy factor. A commercial pro forma has to model each tenant — commencement date, expiration date, current rent, contractual escalations, free-rent period, recovery treatment, renewal option terms, and tenant-specific risks — because each lease behaves differently when it rolls.
Expenses split into recoverable and non-recoverable. A commercial tenant in a triple-net lease pays its proportionate share of taxes, insurance, and CAM. A modified-gross tenant pays only the increases over a base year. A full-service-gross tenant pays nothing directly. The same operating expense line shows up differently in the cash flow depending on which lease pays for what.
Capital is a cash-flow drag at every rollover. When an office tenant’s lease expires, the building owner typically funds tenant improvements (TIs) and leasing commissions to re-tenant the space. On a Class B office building, that can be $40-80 per SF of TI plus 4-6% of total lease value in commissions. Multifamily pro formas reserve $250-400 per unit annually. Commercial pro formas absorb $5-15 per SF of leasing capital depending on the rollover schedule, and the cash flow profile gets lumpy.

The result is that a commercial pro forma is structurally a rent-roll-by-year model with overlays for vacancy, expense recovery, and capital. Argus Enterprise was built specifically for this; Excel can do it with sufficient discipline. Multifamily, hotel, and self-storage rarely justify Argus because the lease structure doesn’t require it. Office, retail, and industrial almost always do.

The line items that matter

The skeleton of every institutional commercial pro forma looks similar. The asset class changes the weighting, not the structure.

Revenue side

Base rental income. Rent from every tenant on the rent roll, escalated according to each lease’s contractual schedule. In Argus, this comes out of the lease-by-lease build. In Excel, it’s an annualized roll of monthly rent times 12 plus contractual bumps. Step rents, percentage rent provisions, and CPI-indexed escalations all live here.
Absorption and turnover vacancy. Vacancy from tenants leaving plus the lease-up period for vacant space. Argus separates “general vacancy” (a global percentage haircut) from “absorption and turnover vacancy” (the actual gap between one tenant leaving and the next signing). Excel models often blur the two; the more rigorous shops keep them separate.
Free rent and concessions. The lease-by-lease schedule of months of free rent granted at signing. On office, six to twelve months of free rent on a ten-year lease is normal in soft markets. Pro formas that net this against base rent without showing the month-by-month timing flatter the interim cash flow.
Expense reimbursements / recoveries. The dollar amount each tenant pays toward operating expenses based on their lease type and pro-rata share. In a triple-net deal, recoveries can equal or exceed base rent. In a modified-gross deal, they recover only escalations over base year. The recovery treatment lives in the lease, and this is the single line where commercial pro formas most often go wrong.
Percentage rent. Retail tenants paying a percentage of sales above a natural breakpoint. Models breakpoint dollars, sales reporting, and any unnatural breakpoints negotiated into the lease.
Other income. Parking, signage, storage, rooftop antenna leases, vending. Smaller in dollar terms but real.
Effective gross income (EGI). Sum of the above. This is the top-line number the rest of the model flows from.

Expense side

Real estate taxes. Reassessed to the buyer’s purchase price. Sellers almost always present the current tax bill. The buyer’s year-1 taxes will be materially higher in most states, with Georgia’s 299C freeze, California’s Proposition 13, and Florida’s portability rules each producing buyer-specific outcomes.
Insurance. Mark to market based on a fresh quote from a broker. Coastal, wind, wildfire, and convective storm exposures repriced dramatically between 2021 and 2026 — underwriting the seller’s legacy policy is one of the most common errors.
Utilities, repairs and maintenance, contract services, payroll, marketing, G&A. The operational line items. In a commercial pro forma, these are split between recoverable (passed through to tenants) and non-recoverable (eaten by the landlord). The split depends on the lease language.
Management fee. 2-3% of EGI for commercial, sometimes 4% for management-intensive retail. Even if the seller is self-managing at zero cost, every buyer should model a market management fee.
Total operating expenses. Sum. Expressed as expense ratio (OpEx / EGI). Office 35-45%, industrial 20-30%, retail 25-35% depending on lease type mix.

Capital side

Tenant improvements (TI). Capital spent to build out space for incoming tenants. Modeled on a per-SF basis per the lease. New office tenants in 2026 typically demand $50-100 per SF of TI on a 7-10 year lease.
Leasing commissions (LC). Broker fees on signed leases. Typically 4-6% of total lease value, split between leasing agent and tenant rep. Office and retail leases carry LC; industrial sometimes does, sometimes doesn’t.
Capital reserves. Recurring building-level capital (roof, parking, mechanical systems). Office: $0.30-0.50 per SF annually. Industrial: $0.15-0.30 per SF. Retail: $0.20-0.40 per SF.
Non-recurring capex. Specific projects — re-skinning a facade, replacing a chiller plant, repaving a parking lot — modeled in the year they hit.

Financing side

Debt service. Interest plus principal. Interest-only periods modeled separately. For floating-rate debt, model the rate path and any cap structure.
Refinance assumptions. If the deal contemplates a refi at year 3 or 5, model the new loan sizing (LTV or DSCR-constrained), the new rate, and the cash-out proceeds.
Debt coverage ratios. Year-by-year DSCR and debt yield. Most lenders today require 1.20-1.30x DSCR and 9-10% debt yield. Pro formas that violate covenants in year 1 don’t get financed.

Terminal value (reversion)

Year N+1 NOI. The forward-looking NOI used to capitalize value at sale.
Exit cap rate. Almost always 25-75 basis points above going-in. The era of compressing exit caps ended in 2022; pro formas that show flat or compressed exits in 2026 are not credible.
Selling costs. Typically 2-3% of gross sale price.
Net sale proceeds. Gross sale value minus selling costs minus loan payoff.

That’s the institutional template. Argus generates it from a lease-by-lease build. Excel generates it from explicit assumption tabs. The math is the same.

Worked example: 10-year office pro forma

Let’s put numbers on it. Assume a 125,000 SF Class B office building in a suburban Southeast market. Four tenants on the rent roll: a 40,000 SF anchor at $24 PSF NNN through year 6, a 35,000 SF tenant at $26 PSF modified gross through year 4, a 25,000 SF tenant at $22 PSF NNN through year 8, and a 15,000 SF tenant at $28 PSF NNN through year 3. Currently 92% leased; 10,000 SF vacant. Purchase price: $22M. Going-in NOI: $1.45M, implying a 6.6% going-in cap. Buyer’s business plan: stabilize, re-tenant the year-3 and year-4 rollers at market ($30 PSF NNN), refresh the common areas, exit year 7.

A skeleton 10-year pro forma at the major-line-item level (all figures in thousands of dollars, rounded):

Line item	Yr 1	Yr 2	Yr 3	Yr 4	Yr 5	Yr 6	Yr 7
Base rental income	2,820	2,890	2,920	3,180	3,610	3,720	3,990
Absorption/turnover vacancy	(150)	(95)	(260)	(380)	(95)	(290)	(110)
Free rent	(40)	(20)	(180)	(240)	(60)	(190)	(75)
Expense recoveries	760	790	810	855	925	960	1,005
Other income	35	36	37	38	39	40	41
Effective gross income	3,425	3,601	3,327	3,453	4,419	4,240	4,851
Real estate taxes	(385)	(395)	(405)	(415)	(425)	(436)	(447)
Insurance	(95)	(98)	(101)	(104)	(107)	(110)	(114)
Utilities	(180)	(185)	(190)	(195)	(200)	(206)	(212)
Repairs & maintenance	(175)	(180)	(185)	(190)	(195)	(200)	(205)
Contract services	(140)	(143)	(147)	(151)	(155)	(159)	(163)
Management fee (3%)	(103)	(108)	(100)	(104)	(133)	(127)	(146)
G&A and other	(95)	(97)	(100)	(102)	(105)	(108)	(110)
Total OpEx	(1,173)	(1,206)	(1,228)	(1,261)	(1,320)	(1,346)	(1,397)
NOI	2,252	2,395	2,099	2,192	3,099	2,894	3,454
Tenant improvements	(50)	(25)	(900)	(1,250)	(100)	(650)	(175)
Leasing commissions	(15)	(10)	(220)	(310)	(35)	(175)	(50)
Capital reserves	(50)	(51)	(53)	(54)	(56)	(57)	(59)
Unlevered cash flow	2,137	2,309	926	578	2,908	2,012	3,170
Sale proceeds (yr 7 exit)							45,800
Less selling costs (2.5%)							(1,145)
Total CF to all capital	2,137	2,309	926	578	2,908	2,012	47,825

A few observations the model surfaces:

NOI dips in years 3 and 4 as the year-3 and year-4 leases roll. Free rent and TI hit the cash flow simultaneously. This is the structural lumpiness commercial pro formas show that multifamily never does.
The exit assumes year-8 NOI of approximately $3.55M capitalized at 7.75% ($45.8M gross). The going-in cap was 6.6%; the exit is 115 basis points wider. Some IC committees will push back on the magnitude; the modeler’s job is to defend it.
Years 3-4 unlevered cash flow drops below 50% of stabilized. A lender modeling DSCR will see covenant pressure unless the loan is structured with interest-only through year 5.
TI plus LC totals approximately $4.0M over the hold — roughly 18% of purchase price. This is the capital drag commercial pro formas have to surface explicitly. Hide it below the line and the levered IRR looks fictitious.

This is what a real commercial pro forma looks like at the summary level. The lease-by-lease build underneath it — each tenant’s escalation schedule, free rent, recoveries, and rollover assumptions — is where Argus earns its license fee, and where Excel modelers spend most of their time.

Argus vs. Excel: where each tool fits

This is the question every commercial team works through.

Argus Enterprise is the institutional standard for lease-by-lease DCF modeling. It handles complex recovery structures, percentage rent, market leasing assumptions, and absorption modeling natively. Lenders, appraisers, and most institutional LPs expect Argus output as the underwriting deliverable on commercial deals. The downside: it’s expensive (low-five-figures per seat annually), the learning curve is real, and it forces you to model the way Argus wants you to model. For a 40-tenant office building, that’s a feature. For a 4-tenant industrial deal, it’s overkill.

Excel dominates on multifamily, hotel, self-storage, and small commercial. It also dominates for the wrap-around analysis — sources and uses, capital stack, waterfalls, sensitivity tables, IRR distributions — that Argus is genuinely bad at. Most institutional shops run Argus for the DCF and Excel for everything around it.

Where AI changes the workflow. Neither tool addresses the upstream problem: producing the rent roll, T-12, and lease abstract that feed the model. That’s where the analyst-hours go. The rent roll arrives as a PDF in a format Argus doesn’t ingest. The T-12 comes out of Yardi in a chart of accounts that doesn’t match the standardized operating-statement template. Each lease has to be read and abstracted by hand for the terms that affect the income stream. Two to three days of analyst work, per deal, before any modeling actually starts.

Atlas (DDee.ai) sits upstream of Argus and Excel. It extracts the rent roll PDF into structured data, parses the T-12 into normalized operating line items, abstracts each lease for the material terms — base rent, escalations, free rent, recoveries, options, exclusivity, co-tenancy — and flags variances against the seller’s representations. Every output cites back to the source document so the analyst can verify in five seconds, not five minutes. The model doesn’t change. The defensibility and the speed of the inputs do.

The errors that keep showing up

Across the deals I reviewed during my buy-side career and the conversations I have now with acquisition teams at PE shops, REITs, and lender credit committees, the same six errors keep appearing in commercial pro formas:

Underwriting market rents without modeling the rollover schedule. Market rent assumptions matter only when the lease rolls. A building 90% leased with five-year average remaining term doesn’t get to “market” for several years, and the pro forma should reflect that.
Flat or compressed exit caps. Anything tighter than 25 basis points of cap rate expansion from going-in to exit in 2026 is not defensible without a specific argument.
Real estate tax under-reassessment. Sellers show the current tax bill. The buyer’s year-1 taxes will be higher in most states. This single line can shift NOI by 5-15%.
Below-the-line capital. TIs and LCs that should appear in the cash flow get netted against sale proceeds, which inflates the interim levered cash flow and the IRR.
Recovery treatment errors. A modified-gross tenant modeled as if they pay a triple-net share, or vice versa. The reimbursement line can be a major component of EGI; getting it wrong moves the whole model.
Missing lease provisions. Co-tenancy clauses on retail, termination options on office, exclusive use restrictions, expansion rights, kick-out clauses. These don’t change the base rent, but they change the credibility of the cash flow stream. The pro forma needs to be built on the lease abstract, not on the rent roll.

None of these are sophisticated errors. All of them get past first-pass review on a regular basis. Most of them get caught in diligence, but only if the diligence team has time to reconcile the model against the lease file — which on a portfolio acquisition is almost never the case without modern tooling.

How sophisticated buyers stress-test a CRE pro forma

The first thing a real acquisitions team does with a seller’s pro forma is rebuild it. The second is run it through scenarios. Standard institutional practice:

Vacancy haircut. Add 200-500 basis points to the seller’s vacancy assumption. Compare against the CoStar submarket average for the asset class.
Expense ratio sanity check. Compare OpEx ratio to RCA, CoStar, and recent broker comps. If the seller’s pro forma shows a materially lower expense load than the comp set, one of you is wrong.
Tax reassessment. Apply the local jurisdiction’s reassessment math to the purchase price.
Insurance mark-to-market. Fresh quote, not the seller’s policy.
Rent growth reality check. Defensible against the five-year market forecast for the submarket.
Exit cap decompression. At least 25-50 basis points.
Three-scenario downside. Base, downside, bear. In downside, lease-up slows by six months, rent growth drops 150 basis points, expenses inflate 100 bps faster than base. If the deal still clears the IRR threshold in downside, real margin exists. If it only works in base, the margin is illusory.

The bottom line

A commercial real estate pro forma is the financial spine of the investment thesis. It’s also where deals get won, lost, and quietly mispriced. The structure is well understood — base rent, vacancy, recoveries, expenses, NOI, capital, reversion. The errors are well understood too — under-reassessed taxes, compressed exits, below-the-line capital, missing lease provisions, recovery mis-modeling.

What separates a defensible pro forma from a marketing exercise is not Excel mastery or Argus certification. It’s the integrity of the inputs. The rent roll, the T-12, and the lease file. Spend less time formatting cells. Spend more time validating the data that flows into them.

That’s the workflow we built DDee.ai around. The pro forma sits downstream. We work on getting clean inputs into it.

Request a Demo →

Frequently Asked Questions

How is a commercial real estate pro forma different from a residential one?: A residential pro forma usually models a single rent stream against a flat operating-expense ratio. A commercial pro forma models a rent roll — multiple tenants, each with their own commencement date, escalation schedule, recovery treatment, free-rent period, and renewal option. The expense side splits into recoverable and non-recoverable buckets, the capital side tracks tenant improvements and leasing commissions on every rollover, and the cash flow gets stress-tested against debt coverage on a quarter-by-quarter basis. Lumping retail templates onto a 200,000 SF office building is where deals get mispriced.
What software do institutional CRE teams use for pro formas?: For commercial assets with lease-level complexity, Argus Enterprise is still the institutional default — most lenders, appraisers, and IC packages expect Argus output. For multifamily, hotel, and self-storage, Excel models dominate because the lease structure is simpler and the underwriting customizations are heavier. Most institutional shops at the Blackstone, KKR, Brookfield, and Nuveen level run both: Argus for the DCF, Excel for the sensitivity work, sources-and-uses, and waterfall. RealPage, Yardi, CoStar, and RCA sit upstream of the pro forma supplying market data and operating benchmarks.
What's the difference between a pro forma and a DCF in commercial real estate?: The pro forma is the projected cash flow stream — year-by-year revenue, expenses, NOI, capital, and reversion. The DCF is the valuation engine that consumes that cash flow stream and discounts it back to present value at a target rate. Argus produces both inside the same file, which is why people use the terms interchangeably. They are not the same. The pro forma is the assumption set. The DCF is what those assumptions are worth today.
What does a development pro forma include that an acquisition pro forma doesn't?: Five things: a sources-and-uses table showing how the capital stack funds construction; a development budget with hard costs, soft costs, financing costs, developer fee, and contingency; a construction draw schedule with capitalized interest; a lease-up absorption curve from certificate of occupancy through stabilization; and a yield-on-cost calculation comparing untrended stabilized NOI to total project cost. Development pro formas also typically include a refinance or sale assumption at stabilization to take out the construction loan.
What are the biggest errors in commercial real estate pro formas?: Five recurring ones. Underwriting market rents without checking the lease roll schedule. Assuming flat exit cap rates instead of building in 25-75 basis points of decompression. Ignoring real estate tax reassessment at the buyer's purchase price. Treating TI and leasing commissions as below-the-line capital instead of recurring drags on cash flow. And failing to model co-tenancy clauses, exclusive use provisions, or termination options that materially change the income stream. None of these are exotic. All of them are missed regularly.
How accurate is a 10-year CRE pro forma?: It depends what you mean by accurate. The line items are almost never right in any individual year — rents do not grow linearly, expenses do not inflate at exactly 2.5%, and vacancy does not stay flat. What a good pro forma does is establish a defensible mid-case that the deal can be evaluated against. The real test is whether the deal still clears the IRR threshold under a downside case with 150 basis points of rent growth haircut, six months of lease-up delay, and 50 basis points of exit cap expansion. If it only works in the base case, the margin is illusory.
Can AI build a commercial real estate pro forma?: Not the whole thing, and not the parts that matter. The judgment calls — vacancy assumptions, rent growth, expense load, exit cap, capital reserves — still belong to the underwriter. Where AI changes the workflow is upstream: extracting the rent roll into structured data, parsing the T-12 into normalized operating line items, abstracting each lease for the terms that affect the income stream, and flagging items that don't match what the seller is representing. That's where most of the analyst-hours on a pro forma actually go, and it's where most of the errors enter.