First-Loss Capital Is Not a Discount — Designing Tranches That Actually Mobilize Investors

A first-loss tranche does its job only when it absorbs precisely the risks that private capital refuses to price. In practice, many facilities are sized backwards: the concessional envelope is decided first, and the tranche is carved from whatever remains. The result is a structure that looks blended on paper but leaves the private investor holding the same risk they declined at the outset.

The starting point is a disciplined risk distribution: which risks are controllable by the sponsor, which are transferable through contracts and insurance, and which genuinely sit with the project from day one — construction risk in frontier logistics, offtake risk in merchant power, currency convertibility risk in restricted markets. Only the residual, untransferable tail justifies first-loss capital.

Sizing then follows from that tail, not from the budget. A tranche sized to the actual residual risk is typically smaller than the political appetite — and more credible to the investors being asked to come in behind it. Credibility is the point: institutional investors do not price the presence of a first-loss tranche; they price whether its sponsor will honor it under stress.

Instrument choice matters as much as size. A guarantee callable on defined triggers behaves differently from an equity-first-loss layer absorbing proportional write-downs, and the choice should track the dominant risk: guarantees for event-driven risks, equity layers for performance-driven ones. Mixing instruments to satisfy multiple concessional funders' mandates is the most common way bankable structures die in committee.

The facilities that mobilize real private capital share one property: the de-risking layer was engineered for the project's risk distribution, and everything else — funder mandates, reporting, governance — was arranged around it.