Ground-Up Development: Process and Capital Stack

Introduction

A feasibility model tells a developer that a project should work. Getting it built is a different problem. It takes two to four years, a layered capital structure stacked from senior debt up through sponsor equity, and a sequence of risks that change character as the project moves from a drawing to a permitted site to a leased building. This article covers the mechanics of ground-up development from the process and equity side: the timeline from raw land to stabilization, the capital stack that funds the build, the gap-filling layers between senior debt and common equity, the merchant-build-versus-build-to-core exit decision, and how RE IB fits around all of it.

It is the companion to development feasibility and residual land value, which covers the metrics that decide whether to break ground. This article covers what happens once the decision is yes. The construction debt itself is treated in depth in bridge and construction lending on the lender's side; here the focus is the whole capital stack and the development process that wraps around it.

Most ground-up projects run 24 to 36 months from land acquisition to stabilization, though the full cycle can stretch from under a year for a simple shell to six years for a complex, entitlement-heavy urban project. The timeline moves through four phases, and the defining feature of development is that the risk does not just shrink as the project advances, it changes character entirely at each phase.

The practical consequence is that a single project is really three different risk profiles stacked end to end, which is why development commands the widest return premium of any real estate strategy and sits at the opportunistic end of the spectrum alongside the return targets that compensate for it.

Stabilization

The point at which a newly developed or repositioned property reaches and sustains a target occupancy (typically 80 to 90 percent) over a defined period, marking the transition from development risk to ordinary asset-management risk. Stabilization is the trigger for replacing short-term construction debt with permanent financing and for beginning distributions to equity, and it is the moment a development is first valued like a standard operating asset rather than a project under construction.

Where the Risk Actually Lives

Three risks define the development period, and they do not arrive at the same time. Entitlement risk is binary and front-loaded: either the project can be approved and permitted as designed, or it cannot, and a denial can strand the predevelopment equity entirely. Construction-pricing risk dominates the build phase: a budget set at groundbreaking has to survive two years of materials and labor inflation, and as covered in the feasibility article, an overrun flows straight through the yield on cost and the development spread. Market-timing risk is the cruelest because it is the least controllable: a developer underwrites a market two to three years before delivery, and the rents, cap rates, and demand that existed at groundbreaking may not exist when the building opens.

These risks also shape the capital structure of the earliest phase. Predevelopment and entitlement are funded with the most expensive money in the project, sponsor and high-risk equity, precisely because no lender will finance a project that does not yet legally exist. The capital that takes the binary entitlement risk demands the highest return, which is why developers try to control land with options and contingent contracts rather than buying it outright before they know the project can be approved.

No single source funds a ground-up project. The money is assembled in a stack, ordered from the safest, cheapest capital at the bottom claim to the riskiest, most expensive at the top. Each layer has a different position in the repayment waterfall, a different cost, and a different risk appetite. A representative 2026 stack for a mid-sized development looks like this.

The senior construction loan is the cheapest capital because it has the strongest collateral and the first claim on repayment. Lenders size it to a percentage of total cost, the loan-to-cost ratio, rather than to value, because there is no stabilized value yet. In 2026, most construction lenders are underwriting 50% to 60% loan-to-cost, a meaningful tightening from the cheap-debt era, which is itself part of why marginal projects stall.

Loan-to-Cost (LTC)

The ratio of a construction or development loan to the total cost of the project (land plus hard, soft, and financing costs), as distinct from loan-to-value, which measures the loan against a finished property's market value. Construction lenders size to loan-to-cost because a project under construction has no stabilized value to lend against. A 55 percent loan-to-cost ratio on a project costing 100 million dollars supports a 55 million dollar construction loan, leaving the remaining 45 million to be filled by gap debt and equity.

How a Construction Loan Funds

A construction loan does not arrive as a lump sum. It is a draw-based facility: the lender releases funds in stages as construction milestones are completed and certified, and the borrower pays interest only on the balance actually drawn. That structure has a direct effect on the feasibility math, because interest accrues on a balance that grows over the whole build, and that capitalized carry lands inside total development cost. The lender typically holds back a retainage on each draw until work is verified, and on most ground-up loans the sponsor signs a completion guarantee, often with recourse during construction that burns off once the building is delivered and stabilized. Because the building earns nothing while it is being built, the loan almost always carries an interest reserve, a pool of loan proceeds set aside to pay interest as it accrues; the project is in effect borrowing to pay the interest on its own loan, which is one more reason a longer build erodes the yield on cost. These facilities are short and floating-rate, typically three to five years priced over a SOFR base, with extension options sized to cover the lease-up period. When the project stabilizes, a permanent loan (agency debt for multifamily, a CMBS or balance-sheet loan for other types) refinances the construction loan and takes the lender out.

Who provides that construction loan has shifted. Regional and money-center banks were the traditional construction lenders, but bank balance-sheet retrenchment after 2023 pulled many of them back, and debt funds and private-credit lenders stepped into the gap, often at wider spreads but with more flexibility on leverage and structure. The result in 2026 is a construction-finance market that is open for well-capitalized sponsors and well-located projects but expensive and selective for everyone else, reinforcing the same screening the development spread imposes on the equity side. The full lender universe is mapped in debt funds and private credit.

A senior loan at 55% loan-to-cost and a sponsor contributing 5% to 10% of cost still leaves a large gap, often 30% to 40% of the budget, that has to be filled. Two instruments compete for that slot. Mezzanine debt is a loan secured not by the property but by a pledge of the equity interests in the property-owning entity, sitting behind the senior loan and requiring an intercreditor agreement with it; it is priced in the 12% to 20% range. Preferred equity is a passive equity investment that sits above the common equity but below the senior debt, earning a fixed preferred return, usually 10% to 14%, paid before the sponsor sees any distribution.

Senior construction lenders generally prefer preferred equity to mezzanine debt, because preferred equity, structured as a passive equity position, does not threaten the senior lender's claim the way a competing secured loan does. That preference has made preferred equity the default gap-filler on development deals, typically landing between 10% and 20% of total cost and stretching toward 25% for experienced sponsors in strong markets. The deeper mechanics of these layers, and how they interact in a default, are covered in mezzanine and preferred equity on the debt side.

The common equity below both sits last in the waterfall and absorbs the first loss. It splits between the sponsor's GP equity (typically 2% to 10% of cost, the "skin in the game") and the LP equity that provides the bulk of the dollars and targets a 15% to 25%+ IRR. The split of profits between them, including the sponsor's promote above return hurdles, runs on the same waterfall and promote mechanics that govern any real estate equity partnership.

Stacking these layers carries a cost the feasibility model feels directly. Each layer above the senior loan is more expensive than the one below it, so the more a project leans on 12% to 20% mezzanine or 10% to 14% preferred equity to fill the gap, the higher the blended cost of the entire stack climbs. A development that needs only modest gap capital pencils at a far lower hurdle than one relying heavily on expensive subordinate layers, which is why the thinner senior loans of 2026, at 50% to 60% loan-to-cost rather than the 70%+ of the prior cycle, quietly raise the return a project must clear to justify itself.

At stabilization, the developer faces a strategic fork that defines the entire business model. A merchant builder develops, leases up, and sells the finished asset, locking in the development profit (the manufactured value in the development spread) and recycling the capital into the next project. A build-to-core developer develops, leases up, and holds, converting a completed development into a long-term income asset and earning the ongoing yield rather than a one-time gain.

The choice is partly philosophy and partly cycle. Merchant building maximizes velocity and crystallizes the spread immediately, but it depends on a liquid exit market at completion. Build-to-core trades the quick profit for durable income and is the natural play late in a cycle, when selling into a soft market would waste the development spread; holding lets the developer wait for a better exit while collecting cash flow. The two strategies have converged at the largest platforms: firms like Hines, Prologis, and Blackstone routinely develop to institutional-quality standards and then decide, asset by asset, whether to sell or fold the building into a long-term core vehicle.

The economics are stark in numbers. A merchant builder who develops a $100 million project to a $130 million stabilized value sells, books the roughly $30 million development profit, and recycles the capital into the next deal, accepting that the gain is taxed and the income stream is gone. A build-to-core developer holds the same building, forgoes the immediate $30 million, and instead collects the $6.5 million of annual NOI on an asset that cost $100 million, a 6.5% yield on cost that beats the 5.0% going-in yield a buyer of a comparable finished asset would accept. Holding lets the developer keep earning the development spread as income rather than cashing it out at once, which is why late-cycle strategy so often tilts toward build-to-core.

A third path sits between them. In a forward sale (or forward funding), a developer agrees to sell the project to an institutional buyer before or during construction, often with the buyer funding construction draws, which de-risks the developer by locking the exit early while letting the buyer acquire a brand-new asset at a negotiated price. Most institutional build-to-core buyers will only commit once a project is fully entitled and permitted, so that groundbreaking can follow immediately before market conditions shift.

Development is not core RE IB the way REIT M&A or a REIT IPO is, but bankers and capital markets advisors interact with it constantly, and candidates are expected to understand where the work sits. Four touchpoints matter.