The CapEx Waterfall: How Lenders Prioritize Draws, Contingencies, and Change Orders—and What It Means for Real Execution Risk

The Structural Problem: Capital Deployment Under Constraint

Every development and repositioning loan contains a CapEx budget. Every budget contains line items, contingencies, and the inevitable change order. But the real story is not the budget itself, it is the waterfall that governs when lenders release capital, which draws they withhold, and what happens when costs exceed the plan. This architecture shapes execution risk in ways that formal loan documents rarely make transparent.

The CapEx waterfall is a constraint that mirrors a deeper structural reality: lenders cannot observe all project risks in real time, so they use capital sequencing as a mechanism of control. By controlling the order and timing of draws, lenders embed assumptions about project progress, cost discipline, and sponsor reliability directly into the cash flow. When those assumptions break, the waterfall breaks with it, and the project absorbs the friction.

This is not new. It is a consequence of information asymmetry and the lender's structural exposure to construction and lease-up duration variance. The mechanics have hardened over decades because the underlying constraint is permanent: lenders cannot fund a project they cannot observe, and they cannot observe a project whose costs are not yet incurred.

The Formal Waterfall: Draws, Retainage, and Contingency Release

Most construction and value-add loan agreements establish a draw request protocol. The sponsor submits evidence of costs incurred (invoices, lien waivers, architect certification). The lender verifies that the costs are within budget, that liens are released, and that progress aligns with the construction schedule. If all conditions are met, the lender releases funds.

But the waterfall does not distribute capital equally across all line items. Instead, lenders embed a hierarchy. Hard costs, including labor, materials, and subcontractor work, are typically funded first, subject to lien waivers and evidence of payment. Soft costs, including design, permits, and professional fees, are often funded second and sometimes withheld entirely until substantial completion. Contingencies, the cushion meant to cover scope creep and unforeseen conditions, are typically the last to be released, if at all.

Retainage is the mechanism that enforces this discipline. Lenders hold back 5 to 10 percent of each draw request, releasing it only after the general contractor certifies completion of that trade or phase. This holdback serves two purposes: it guarantees that subcontractors are paid (reducing lien risk to the lender), and it gives the lender leverage to demand corrections before releasing the final tranche.

Contingency release is where the waterfall reveals its true function. A typical construction loan sets aside 5 to 15 percent of total CapEx as contingency. But lenders rarely release this capital until the project demonstrates actual cost discipline. In practice, contingencies are released only when the project has spent down the base budget with minimal overrun, or when the sponsor demonstrates that the overrun is the result of a documented change order that the lender has approved.

Change Orders as a Structural Constraint on Flexibility

Change orders are the formal mechanism for expanding the scope of work and the CapEx budget. But the change-order process itself is a waterfall. The sponsor or general contractor identifies a cost variance, either because the original scope was incomplete or because field conditions require modification. The sponsor submits documentation of the cost impact. The lender reviews the change order against the project's contingency reserve and the loan agreement's change-order threshold.

Most loan agreements set a threshold, often 2 to 5 percent of total CapEx, below which the sponsor can approve change orders unilaterally, as long as the total does not exceed the contingency reserve. Above that threshold, the lender must approve the change order before capital is released. This creates a staggered system of control: small variances flow through the sponsor's discretion, while large variances require lender consent.

The structural tension emerges when field conditions drive repeated small change orders that aggregate above the threshold. The sponsor has already drawn down the contingency reserve, so there is no room for further variance without lender approval. The lender, observing that the project is burning contingency faster than the schedule anticipated, becomes more cautious about approving subsequent change orders, even if they are legitimate.

This is where execution risk becomes real. The project is not delayed because the work is not being done, it is delayed because capital is not being released. The general contractor cannot proceed without payment authorization, and the lender cannot authorize payment without confidence that the change order is within scope and cost discipline is being maintained. The waterfall freezes.

The Holdback as a Hidden Equity Call

Retainage and contingency holdbacks function as embedded equity calls on the sponsor. When a project burns contingency faster than anticipated, the sponsor has three choices: fund the overrun from retained earnings, reduce scope, or request that the lender approve a change order and release additional contingency capital.

In practice, most sponsors choose the first option: they fund the gap from their own capital. This is not a surprise. It is by design. Lenders structure the contingency and retainage mechanics to incentivize sponsors to maintain cost discipline and absorb small variances from their own resources. The holdback is a mechanism that converts sponsor capital into lender collateral without formally calling it a capital injection.

This structure becomes acute in deals where the sponsor's capital is already fully deployed. In value-add and development deals, the sponsor often has limited dry powder by mid-construction. When contingency is exhausted and change orders are pending lender approval, the sponsor faces a choice between waiting for lender consent (which delays the project and extends the interest carry) or injecting additional capital (which dilutes returns and increases sponsor leverage). The waterfall forces this decision.

Over cycles, this constraint has shaped how sponsors underwrite contingency. Many sponsors now budget contingency at 10 to 15 percent, not because the risk is higher, but because they expect to fund the gap themselves and want to maintain enough headroom to avoid a capital call or a lender dispute. This is path dependence in action: the structural constraint of the waterfall has become embedded in sponsor assumptions about what contingency should be.

What Execution Risk Really Means

Execution risk in a development or value-add deal is not solely a function of construction quality, labor availability, or supply-chain disruption. It is also a function of capital release timing, contingency adequacy, and the sponsor's ability to absorb variances without triggering lender intervention.

A project with excellent contractors and clear scope can still experience execution friction if the CapEx waterfall is structured too tightly. Conversely, a project with higher inherent construction risk can navigate it more smoothly if the waterfall allows adequate contingency release and change-order flexibility.

The lender's perspective is rational: they cannot observe all risks, so they use capital sequencing as a lever of control. But this mechanism has a cost. It transfers execution risk from the lender to the sponsor and the contractor. The sponsor must absorb small variances from contingency or retained earnings. The contractor must wait for lender approval of change orders, which delays cash flow and can disrupt labor scheduling. The project absorbs the friction in the form of schedule delays, cost creep, and stakeholder tension.

Over the cycle, this structural constraint has not changed. Lenders still cannot observe all project risks. Sponsors still cannot fund all variances without constraint. The waterfall still governs the release of capital. What has changed is the transparency with which deals acknowledge this tension and the specificity with which loan agreements define the change-order process, contingency release, and retainage mechanics.

The deals that navigate this most effectively are those where the sponsor and lender have explicitly negotiated the waterfall upfront, agreeing on contingency thresholds, change-order procedures, and retainage release schedules before closing. The deals that struggle are those where the waterfall is treated as boilerplate and the sponsor discovers its constraints only when the project encounters its first major variance.

What I'm watching next

• How construction inflation and labor scarcity are forcing sponsors to increase contingency budgets, and whether lenders are responding by tightening change-order thresholds or loosening contingency release mechanics.
• The emergence of draw-request automation and real-time lien tracking systems, and whether these tools reduce the friction in the waterfall or simply make the holdback mechanism more efficient.
• The role of interest reserves in funding CapEx overruns: as interest carry duration extends, whether sponsors are using interest reserves to absorb contingency burndown rather than drawing additional CapEx capital.