Real Estate Development Only Works When the Numbers Work

Hello everyone! 👋

One thing I keep noticing as I learn more about real estate development is how often the exciting part comes second.

The building is exciting. The design is exciting. The idea of transforming a site into something useful and valuable is exciting.

But before any of that really matters, the numbers have to make sense.

Real estate feasibility comes down to a single test — does the stabilized value of the finished project exceed the all-in cost of getting there? — and the pro forma, NOI, cap rate, and land residual are the instruments that answer that test honestly. Everything else in real estate finance, including the capital-stack and leverage math in the companion post, is downstream of whether the value-side numbers actually work.

That distinction matters because most under-rigorous deals fail the value-side test, but the developer doesn't notice until financing falls apart. By then the land is under contract, the architect has been paid, and the deal has its own gravity. The honest test is the one you run before any of that.

This post walks the value side: feasibility, pro forma mechanics, NOI, cap rate, the land residual, and what each of those instruments is actually telling you. The capital-stack side — debt, equity, leverage, and lender ratios — is the companion post.

Real estate feasibility starts with a simple question

At the center of all of this is one basic idea:

Does stabilized value exceed all-in cost?

That sounds technical at first, but it is actually a very grounded question.

A developer is asking whether the finished, income-producing project will be worth more than everything it takes to get there — land, construction, indirect costs, financing costs, and the profit required to justify doing the deal in the first place.

I like this framework because it cuts through a lot of noise.

You can love the concept. You can love the site. You can love the neighborhood story.

But if the stabilized value does not exceed the all-in cost in a way that supports the capital and the risk, then it is not really a deal. It is just an expensive idea.

That single question only works if "stabilized" is defined honestly — and honesty about stabilized is where most pro formas first start lying.

Stabilized does not mean perfect

"Stabilized" is a deliberately modest word — it concedes that the building will never be perfect and asks you to underwrite reality, not the brochure.

It does not mean 100% full forever.

In real life, buildings have friction. Tenants leave. Space turns over. Leasing takes time. Collections are not always perfect. So stabilized value is based on a realistic operating condition, not a fantasy version of full occupancy with zero problems.

That matters because a lot of bad assumptions start when people underprice vacancy, overprice rents, or pretend the property will operate with no drag.

The exact labels can vary a little depending on the underwriting style, but the broader idea stays the same: the project has to be tested under real operating conditions, not just best-case storytelling.

Once you accept that stabilized income lives in the messy middle, the question becomes how to write that messiness down on paper without flinching.

The pro forma is where the project becomes honest

The pro forma is one of those tools that sounds dry until you realize what it actually does.

It takes a project idea and forces it into structure.

Instead of saying, “This seems promising,” the pro forma asks:

• What income do you actually expect?
• What assumptions are driving that?
• What will operations cost?
• What financing costs show up?
• What cash flow is really left over?

That is where a deal stops being conceptual and starts becoming accountable.

What I like about this is that it makes feasibility feel less mysterious. A pro forma is really just a disciplined way of turning assumptions into consequences.

Inside every pro forma, one line is doing most of the work.

NOI is one of the key numbers that holds everything together

NOI — Net Operating Income — gets its authority from being the one number that survives every change in ownership, financing, and tax structure. That is why every other party in the deal eventually agrees to argue about it.

You can think of it as the property's operating earning power: what the building itself produces before any of the ownership-side complexity is layered on top.

That is why lenders care about it. That is why investors care about it. That is why valuation often starts with it.

And that leads directly to one of the most common valuation tools in real estate: the cap rate .

Cap rate is simple, but it changes how you see value

One of the cleanest ideas in the whole topic is this:

That’s it.

Of course, the hard part is not memorizing the formula. The hard part is choosing a reasonable NOI and a reasonable cap rate. But the logic itself is very intuitive: value depends on how much income the property generates and what return the market demands for that kind of asset.

And the inverse relationship matters a lot too.

If NOI stays the same and buyers bid the price higher, the cap rate falls. If NOI stays the same and the price drops, the cap rate rises.

That is one of those ideas that feels small at first, but once it clicks, a lot of real estate valuation starts making more sense.

That formula looks like arithmetic, but the cap rate itself is anything but.

Cap rates are not just math — they reflect expectations

What also stood out to me is that cap rates are not random numbers floating around in spreadsheets. They reflect expectations.

They carry assumptions about:

• income reliability,
• future growth,
• resale expectations,
• risk,
• and what investors think they need to earn for holding that asset.

So even though the formula looks simple, the cap rate itself carries a lot of market judgment inside it.

That is why two properties with similar income can still be valued differently. The market is not only pricing what the asset is producing now. It is also pricing how it feels about the future.

All of that — the income, the cap rate, the implied value — still only describes one half of the test.

Feasibility is where value meets cost

A project can have a projected stabilized value. Fine.

But then you still have to ask what it costs to actually get there:

• direct construction,
• indirect costs,
• leasing and tenant improvements,
• financing,
• land,
• and developer profit.

That is where financial feasibility becomes less about “what is it worth?” and more about “what is left over after the deal is built honestly?”

And the cost side rarely surprises in your favor. As Brian Potter has documented in Construction Costs Rarely Fall, construction outpaces general inflation across almost every period and almost every cost index — which means the cost side of the equation tends to grow faster than the rent side that has to absorb it.

This is also where the land residual method becomes so useful.

Because instead of starting with the seller’s asking price, it starts with the economics of the actual project.

The land residual is basically a discipline test

The land residual flips the negotiation: instead of starting with what the seller wants, it tells you the highest price the finished project can survive paying — and that price is rarely what the seller wants.

You estimate the project’s stabilized value. You subtract the cost to build. You subtract the profit needed to justify the risk. And what remains is the most you should be willing to pay for the land.

That is powerful because it shifts the conversation away from emotion and toward discipline.

A site is not worth whatever somebody hopes to get for it. It is worth what the finished deal can support.

And once again, that does not mean a developer can never pay more. But if they do, they should know exactly why and exactly what has to go right to make that decision survivable.

The numbers behind the decision

At some point, site evaluation stops being qualitative and starts being quantitative. The rest of this post walks through the key financial formulas and gives you small calculators to play with.

A simple feasibility check

At a high level, one of the first questions is whether stabilized value exceeds all-in cost.

All-In Cost = Land Cost + Direct Costs + Indirect Costs + Financing Costs + Developer Profit

If stabilized value does not exceed all-in cost by enough, the deal probably needs to be reworked — or left behind.

From rent assumptions to real income

A project usually starts with potential income, but the more useful number is the income you expect after real-world friction.

PGI = Rentable Area × Market Rent, then subtract vacancy and add miscellaneous income to get Effective Gross Income.

Estimating NOI

Once effective gross income is estimated, the next step is to subtract the ongoing costs of operating the property. NOI tells you what the property itself is producing before ownership-level returns are split up.

Turning income into value

A common real estate shortcut for value is the cap rate method:

.

For example, if a project has an NOI of $3,100,000 and the market cap rate is 6.5%, then the value is about $47.7M.

How much is the land really worth to this deal?

Instead of starting from the seller’s asking price, developers often work backward from what the completed project can support.

Return on cost

A quick way to compare value and cost:

But what about deals that beat the residual?

There is a sharp objection to all of this. A strict residual buyer would never have funded Hudson Yards, where Related and Oxford signed a 99-year, $1 billion air-rights lease with the MTA — over an active rail yard, before a single column was poured. No static residual blesses a billion-dollar land basis on a parcel where the "land" is sky. And yet the project, projected at roughly $25 billion all-in, happened, and it works. Or look at industrial 2018–2022: CBRE's H1 2022 Cap Rate Survey shows institutional industrial trading at cap rates clustered between 3% and 5%, in some cases below the cost of borrowing. Anyone who refused those deals on residual grounds missed one of the largest cap-rate-compression cycles in modern commercial real estate.

It helps to actually see the rail yard the deal was built on top of. The B1M's walkthrough below makes the conviction-vs-residual gap visceral — a billion-dollar land basis above an operating commuter line sounded absurd until it didn't.

The honest concession here is that the residual is a snapshot, not a forecast. It assumes today's rents, today's cap rates, today's costs. A developer who only runs residual will misprice any deal whose thesis is appreciation, entitlement upside, or assemblage value — the way Hudson Yards' thesis was that anchoring office, retail, residential, and public space together creates value no individual parcel could justify alone.

But the same era produced the wreckage that vindicates the residual test. Tishman Speyer and BlackRock paid $5.4 billion for Stuyvesant Town in 2006 on appreciation assumptions about deregulating rent-stabilized units; the deal collapsed in roughly three years (Bagli, Other People's Money). Tides Equities built a $7 billion Sun Belt multifamily portfolio on floating-rate bridge debt at sub-4% caps; by 2024, around 20% of the book was in distress and roughly $1.5 billion in loans were maturing into a 5%+ SOFR world. Applesway Investment Group lost 3,200 units across $229 million of Arbor loans to foreclosure in April 2023 alone. Trepp's multifamily CMBS delinquency rate went from 1.33% in April 2024 to 6.57% one year later — roughly a 4× jump in twelve months. Even Toll Brothers, the disciplined builder, took $399 million of land impairments in fiscal 2008 and another $267 million in fiscal 2009 when its own residuals proved too optimistic.

You can feel the asymmetry yourself. Set over-pay to 20%, NOI growth to 4%, and the exit cap shift to −50 bps — the deal pencils. Now hold over-pay at 20%, drop NOI growth to 0%, and shift the exit cap +50 bps. The same purchase price, same building, same hold period — and the deal craters.

The pattern is asymmetric. Paying above residual on conviction can win — Hudson Yards did, industrial 2018–2022 did. But losing on conviction is permanent equity destruction, not a temporary mark. The residual and the replacement-cost check are the floor that tells you whether the downside is survivable. Sam Zell is honest about both sides in Am I Being Too Subtle? — he made his reputation buying at cycle bottoms when residual math said "no, but the cycle says yes," and he refused to chase deals where cap rates implied yields below replacement-cost economics. The discipline is not to never pay above residual. It is to know exactly how much above, and exactly what has to go right to make that decision survivable.

A few practical lessons I'm taking away

If I had to compress the value-side test into a few grounded takeaways:

• A project is not feasible just because the concept is attractive — stabilized value has to exceed all-in cost in a way the math actually supports, not in a way the spreadsheet can be cajoled into.
• NOI is the connective number — it sits at the center of operating performance, valuation via cap rate, and lender ratios. Most other real estate finance numbers are derivatives of NOI.
• Cap rate looks like simple division but carries a lot of market judgment. It encodes growth expectations, perceived risk, what comparable assets recently traded for, and the alternative-investment yield environment. Sam Zell makes the point bluntly in Am I Being Too Subtle?: he won't buy on cap rate alone — "there's nothing more relevant than replacement cost." Cap rates that drift below replacement-cost-implied yields are a warning, not a deal.
• The land residual is the discipline tool that catches optimistic land prices early. If the deal doesn't pencil at the land price the seller wants, the residual tells you what price the project can actually carry.
• Sensitivity testing matters more than point estimates. A project that pencils at 5.5% cap and a 1.10 lease-up reserve, but breaks at 6.0% cap or a 1.05 reserve, is a project that depends on the market not moving — which the market always does.
• The NCREIF NPI cap-rate series is a useful sanity check. As of Q3 2025, market value-weighted appraisal cap rates were around 4.60% and transaction cap rates around 5.62% across institutional commercial real estate — handy reality benchmarks when your model assumes something dramatically different.
• The honest test is the one you run before committing capital. Most of financial feasibility is the discipline of letting the numbers tell you no, not the creativity of finding ways for them to say yes.

What this gets you ready for

This is the value-side half of real estate feasibility. The other half — debt, equity, leverage, lender ratios, and recourse — sits in the companion post on the capital stack. A developer who masters one half but not the other will keep building deals that pencil on paper but can't be financed, or deals that get loans but never reward equity for the risk it took.

Sometimes the smartest financial decision is not how to make the deal work. Sometimes it is realizing that the deal only works if you keep lying to yourself.

This post is the first of two on real estate financial feasibility in my ongoing series — Real Estate Development. The companion post covers the capital stack — debt, equity, leverage, and lender ratios. Earlier posts cover why development starts long before construction and why it starts with the site.

Sources

• Sam Zell, Am I Being Too Subtle? Straight Talk From a Business Rebel (Portfolio, 2017) — used for the cap-rate-vs-replacement-cost framing in the cap-rate section. https://www.amazon.com/Am-Being-Too-Subtle-Straight/dp/1591848237
• NCREIF, NCREIF Property Index (NPI) — Trends Report — quarterly cap-rate data series for institutional commercial real estate; used for the Q3 2025 cap-rate benchmarks (4.60% appraisal, 5.62% transaction). https://user.ncreif.org/data-products/property/
• William Brueggeman & Jeffrey Fisher, Real Estate Finance & Investments (16th ed., McGraw-Hill, 2018) — the standard textbook covering pro forma, NOI, cap rate, land residual, and feasibility. Recommended as the deep-reference companion. https://www.amazon.com/FINANCE-INVESTMENTS-Jeffrey-William-Brueggeman/dp/1260091945
• Break Into CRE / CRE Analyst — pro-forma walkthrough video resources — useful video walkthroughs of the same calculations covered here. https://breakintocre.com/ · https://www.creanalyst.com/fast-track
• CoStar / RealPage — multifamily cap rate transaction comps — paywalled comp data; CRE Daily's free summaries are a usable alternative for non-subscribers. https://www.costar.com/campaign/multifamily-property-data · https://www.credaily.com/briefs/cap-rates-stabilize-driving-momentum-in-multifamily-investment/
• Charles V. Bagli, Other People's Money: Inside the Housing Crisis and the Demise of the Greatest Real Estate Deal Ever Made (Dutton, 2013) — Stuy Town / Tishman Speyer / BlackRock $5.4B blowup; cited in the counterargument. https://www.penguinrandomhouse.com/books/308966/other-peoples-money-by-charles-v-bagli/
• Vicky Ward, The Liar's Ball (Wiley, 2014) — book-length account of the GM Building deal and the Macklowe-era pay-above-residual culture; useful background for the conviction-overpay archetype. https://www.amazon.com/Liars-Ball-Extraordinary-Building-Toughest/dp/1118295315
• William J. Poorvu & Jeffrey L. Cruikshank, The Real Estate Game (Free Press, 1999) — HBS reference text on the back-of-the-envelope feasibility analysis pattern this post mirrors. https://www.amazon.com/Real-Estate-Game-Intelligent-Decisionmaking/dp/068485550X
• CBRE U.S. Cap Rate Survey, H1 2022 — industrial cap rates clustered 3%–5%, the data point used in the counterargument. https://www.cbre.com/insights/reports/us-cap-rate-survey-h1-2022
• The Real Deal — "Arbor Realty Forecloses $229M Multifamily Portfolio in Houston" (April 10, 2023) — Applesway / Arbor figures used in the counterargument. https://therealdeal.com/texas/2023/04/10/arbor-realty-forecloses-on-229-million-multifamily-portfolio-in-houston/
• The Real Deal — "Tides Equities flew too close to the sun" (July 6, 2023) — Tides $7B Sun Belt portfolio context. https://therealdeal.com/national/2023/07/06/tides-equities-flew-too-close-to-the-sun-it-wasnt-alone/
• Multi-Housing News — CMBS Delinquency Rates tracker — Trepp multifamily series (1.33% Apr 2024 → 6.57% Apr 2025). https://www.multihousingnews.com/cmbs-delinquency-rates/
• Toll Brothers, FY2009 10-K (SEC EDGAR) — fiscal-2008 and fiscal-2009 land impairment figures ($399.1M and $267.4M). https://www.sec.gov/Archives/edgar/data/794170/000095012310115566/c06107e10vk.htm
• Global Infrastructure Hub — "Hudson Yards air rights monetisation" case study — $1B / 99-year MTA air-rights lease detail. https://www.gihub.org/innovative-funding-and-financing/case-studies/hudson-yards-air-rights-monetisation/