Development Feasibility: Yield on Cost and Land Value

Introduction

Ground-up development is the one corner of real estate where value is manufactured rather than purchased. Instead of paying a market cap rate for an income stream that already exists, a developer commits two to four years and a large construction budget to create a brand-new income stream, and the whole exercise only makes sense if the finished building is worth more than it cost to build. Feasibility analysis is the discipline that answers a single question before a shovel hits the ground: does this project create enough value to justify the cost and the risk? Real estate investment bankers do not run feasibility models day to day (that is the developer's job), but they need to read them fluently, because the value of a developer client, a REIT's development pipeline, and a land contribution into a joint venture all rest on the same handful of metrics.

This article covers the core feasibility toolkit: yield on cost (the return a developer earns on every dollar spent), the development spread (how that return compares to what finished buildings sell for), residual land value (the maximum a developer can pay for the dirt), and profit on cost (the bottom-line margin). These four metrics sit alongside the property-level DCF and IRR and equity multiple as the analytical backbone of any development decision. The companion article on the ground-up development process and capital stack covers the timeline and financing structure that turn these numbers into a built asset.

Yield on cost, also called development yield or return on cost, is the stabilized NOI a completed project will generate divided by the total cost to build it. It is the development world's answer to the cap rate. Where a cap rate measures the yield on a building's current market price, yield on cost measures the yield on what it costs to create that building from scratch.

Total development cost is the all-in number, not just construction. It bundles four buckets: the land, the hard costs (the physical construction, including materials, labor, and site work), the soft costs (architecture and engineering, permits and impact fees, legal, marketing, and financing fees), and the carry (the interest and capitalized cost of capital during the construction and lease-up period before the building produces income). The carry bucket is the one developers most often underestimate. Interest on the construction loan is drawn and capitalized over the full build, so a longer timeline or a higher rate compounds directly into total cost and drags the yield on cost down before a single tenant signs. On a two-year build, financing carry alone can run 8% to 12% of total cost, which is why time to completion is itself a cost rather than just a schedule. A development that costs $100 million all-in and stabilizes at $6.5 million of NOI carries a 6.5% yield on cost. The same metric anchors the data-center development math in data center capex and time to power, where yield on cost is computed per megawatt rather than per square foot.

Yield on Cost

The ratio of a development's stabilized net operating income to its total all-in development cost (land plus hard costs plus soft costs plus financing carry). It expresses the unlevered return a developer earns on every dollar invested in creating the asset, and it is the development equivalent of a going-in cap rate on an existing building. A higher yield on cost means a more profitable project at a given exit cap rate.

The reason yield on cost matters more than a simple construction budget is that it lets a developer compare a project they would build against a comparable building they could simply buy. If finished assets of the same type and quality trade at a 5.0% cap rate, and the developer can build to a 6.5% yield on cost, the project is worth doing: the developer creates an asset yielding 6.5% for a price that the market values at 5.0%. That gap is the entire economic rationale for development, and it has a name.

Trended Versus Untrended Yield on Cost

There is a subtle but important choice buried in the numerator. Untrended yield on cost uses today's market rents to compute stabilized NOI, as if the building could lease up at current rates the moment it opens. Trended yield on cost uses rents grown forward to the projected delivery date, two to four years out, to reflect the rent inflation expected over the construction period. Trended numbers always look better, because they capture rent growth the developer has not yet earned and may not get.

This distinction is where many development underwritings quietly fall apart. A project that pencils only on a trended basis is betting on rent growth to bail it out, and that bet is exposed precisely when it is most fragile, in a softening market. Disciplined lenders and equity partners underwrite to the untrended yield on cost and treat any trended uplift as upside rather than base case. The risk is asymmetric because costs are also moving: the Engineering News-Record Building Cost Index rose 4.2% in 2025, with structural steel up 11.9% and embedded tariff costs running roughly $15 to $25 per square foot on mid-rise multifamily. If hard costs trend up while rents stall, the trended yield on cost a developer underwrote two years earlier can evaporate by delivery.

The development spread is the difference between a project's yield on cost and the exit cap rate at which the finished, stabilized building would sell. It is the single most important number in ground-up underwriting.

Using the earlier example, a 6.5% yield on cost against a 5.0% exit cap rate produces a 150-basis-point development spread. That spread is the developer's manufactured value: build at a 6.5% yield, sell at a 5.0% cap, and the asset that cost $100 million is worth roughly $130 million on completion ($6.5 million of NOI capitalized at 5.0%). The spread, not the absolute yield, is what tells you whether a project is worth the risk, because it isolates the value created over and above simply buying a comparable finished building.

How much spread is enough depends on the property type and the risk of the build. Spreads required to make a deal financeable have widened as construction costs and rates rose, and they vary meaningfully by sector and timeline.

Industrial illustrates the forgiving end of that table. A simple distribution shell that builds to a 6.5% yield on cost against 5.0% Class A caps clears only a 150-basis-point spread, but the capital is exposed for a twelve-to-eighteen-month build, so a modest spread is tolerable. An office tower demanding 300-plus basis points is not being greedy; it is pricing four years of construction, deep tenant-improvement spend, and an uncertain lease-up at the end.

The spread is also a measure of how forgiving a project is. A wide spread absorbs cost overruns, schedule delays, and unexpected rent softness; a thin one does not. In mid-2026 Bay Area multifamily, spreads above 300 basis points were unusual enough that the rare ground-up project clearing that bar could still pencil and attract financing at a time when many comparable deals nationally could not. With Class A industrial cap rates near 5.0% and value-add office and select-service hotel caps at 8.5% or higher in 2026, the exit cap assumption alone can swing a project from financeable to dead.

What Moves the Development Spread

Three inputs drive the spread, and all three moved against developers across 2025 and 2026. Construction costs push the yield on cost down as total cost rises: with the Building Cost Index up 4.2% and steel up 11.9% in 2025, the denominator of yield on cost kept climbing. Interest rates hit the spread from both sides, raising the financing carry inside total cost and pushing exit cap rates wider at the same time. And the exit cap rate itself, the single most sensitive assumption, is a forecast made years before the sale: underwrite it too tight and the spread is fictional. When all three move adversely at once, spreads compress toward zero and ground-up starts stall, which is precisely the pattern that slowed multifamily and industrial deliveries after 2025 even with demand intact. The discipline is to test how far each input can move before the spread closes, rather than underwriting a single optimistic case.

The first three metrics tell a developer whether a project works at a given land price. Residual land value flips the question around: given everything else, how much can the developer afford to pay for the land and still hit the target return? It is the most important number in any land acquisition, because land is the one cost a developer negotiates rather than estimates.

GDV is the gross development value, the completed and stabilized value of the finished project (typically the stabilized NOI capitalized at the exit cap rate). From that, subtract all the non-land development costs (hard costs, soft costs, and financing carry) and the developer's required profit. Whatever is left is the maximum the developer can pay for the land. If the landowner will sell at or below that figure, the deal is feasible; if the asking price is higher, the developer must find more rent, cut costs, accept a thinner profit, or walk.

Residual Land Value

The maximum price a developer can pay for a site while still achieving a target profit, calculated as the gross development value of the finished project minus all non-land development costs minus the required developer profit. It is the dominant method for pricing development land because land rarely has reliable comparables, so its value is derived from what the completed project can support rather than from sales of similar parcels.

One mechanical subtlety matters here. If the required profit is expressed as a percentage of total cost, and total cost includes the land you are trying to solve for, the calculation becomes circular. The common fix is to express the required profit as a percentage of GDV for the residual calculation, which removes the circularity and lets the land value fall out cleanly.

Residual land value also explains why a developer's existing land basis can make or break a project. A sponsor who bought a site cheaply years ago, before costs and rates rose, may still clear the profit hurdle on a project that would never pencil for a buyer paying today's residual value. A sponsor who overpaid at the peak can be underwater before construction starts, with no way out except more rent or a thinner return. This is why land is marked to its residual value, not its historical cost, in any serious development analysis or pipeline valuation.

Profit on cost is the bottom-line development margin: the completed value of the project, less everything it cost to build, divided by that total cost.

In the running example, a $130 million completed value against a $100 million total cost is a $30 million profit, or a 30% profit on cost. Developers commonly target a profit on cost of roughly 15% to 20% as the minimum threshold to take on the construction and lease-up risk of a ground-up deal, with riskier asset types and longer timelines demanding more. Profit on cost is closely related to the development spread (both measure manufactured value) but expresses it as a margin rather than a yield differential, which makes it the cleaner number for a go or no-go decision.

The same margin can be quoted two ways, and the distinction trips up candidates. Profit on cost divides profit by total cost; profit on GDV divides the same profit by the completed value. Because GDV is larger than cost, profit on GDV is always the smaller percentage: a 30% profit on cost is roughly a 23% profit on GDV on the same deal. US developers and lenders usually speak in profit on cost, while profit on GDV is more common in UK and European development finance. Neither is wrong, but comparing one developer's profit on cost to another's profit on GDV is comparing two different things.

This is also the feasibility gate. A project clears when, at untrended rents and realistic costs, the development spread is wide enough and the profit on cost high enough to compensate for the risk. The metrics interlock: yield on cost feeds the spread, the spread and the profit target set the residual land value, and the land value the developer can actually negotiate determines whether the whole thing pencils. Move any one input (rents soften, steel prices jump, the exit cap widens) and the others shift in response. The discipline of feasibility is testing how much each input can move before the gate closes.

Bankers are not principals, but development feasibility shows up constantly in RE IB work, and fluency in these metrics separates strong candidates from weak ones. Three contexts matter most.

The construction-financing structure, draw mechanics, and completion risk that wrap around these metrics are covered in bridge and construction lending on the debt side and in the ground-up development process article on the equity and process side. Together they describe how a feasibility model becomes a financed, built, and stabilized asset, and how the value manufactured in the spread actually gets realized.