The traditional position
Project finance assumed a fixed-price, date-certain contract with a single contractor bearing completion risk, backed by liquidated damages and security. Lenders sized debt against that certainty.
Contractors have retreated from it. Losses on large fixed-price projects led several major firms to withdraw from that model entirely, and those still offering it price the risk at a level sponsors frequently decline.
What is being used instead
Engineering-procurement-construction management. The contractor manages delivery for a fee, with the sponsor holding subcontracts and the associated risk. Cheaper, and the completion risk stays with the owner.
Target price with pain-share and gain-share. Cost overruns and savings are shared on an agreed formula, usually with a cap on the contractor's exposure. Aligns interests better than either extreme.
Split contracting. Separate contracts for major equipment, balance of plant, and installation, with the sponsor or a wrap provider bearing interface risk. Common where equipment supply is scarce and suppliers will not accept downstream obligations.
Progressive design-build. Two phases: a development phase where cost is established collaboratively, then a fixed or target price for construction. Reduces the pricing of unknowns into the initial number.
What lenders require when there is no wrap
- A larger contingency, funded and controlled, with clear release conditions
- Sponsor completion support: a guarantee or equity commitment covering overrun and delay
- An independent engineer with real authority over drawdowns
- Interface risk explicitly assigned, with a named party responsible where schedules collide
- Realistic liquidated damages caps assessed against actual exposure rather than a convention
Cost escalation and supply
Extended lead times for major equipment and volatile input costs have made fixed pricing harder to hold. Escalation mechanisms indexed to published measures are increasingly accepted. Where they are used, the sponsor's exposure should be modelled against a plausible escalation range, not the base case.
The judgement to make
Risk that a contractor will not accept does not disappear; it returns to the sponsor, to lenders, or to a contingency. The structures above are honest about where it sits. The failures in recent US construction have generally been projects where a contractor accepted risk it could not carry, and the resulting insolvency delivered a worse outcome than a candid allocation would have.
