Honduras didn't pay its generators. It refinanced them.

Honduras's state utility owes private generators roughly $655 million for power already delivered, about L17,385 million as of March 2026, with payment delays running four to seven months, per Energía Estratégica. The government's response is not to pay that bill. It is to bond it.

ENEE buys around L1,600 million of electricity a month and has for years spent more doing so than it collects in tariff revenue, a gap that shows up every few months as a fresh headline about near-bankruptcy. By May 2026 other reporting put the generator debt closer to $1.05 billion, a later and separately sourced figure that at minimum confirms the direction: the backlog kept growing even as the state announced a rescue plan.

Most of the coverage treats the rescue itself as the story: bonds get issued, generators get paid, the lights stay on. We think the more useful question sits one layer down, in what kind of claim a generator is actually left holding once an unpaid invoice gets turned into a fifteen-year bond.

From receivable to duration risk

A normal power purchase agreement extends the offtaker something close to thirty-to-sixty-day trade credit. That is a working-capital risk, and a credit committee can underwrite it: a defined collection cycle, reserve accounts sized to a few months of billings, liquidated-damages language if the offtaker breaches. Four-to-seven-month arrears already stretch that model past where it was designed to work.

Honduras's fix stretches it further. Sefin authorized ENEE to issue up to L7,152.7 million in bonds earmarked specifically to pay down generator arrears. ENEE placed L3,510.5 million of that in May 2026 and auctioned a further L2,200 million in June, at tenors of five to fifteen years, per El Heraldo and Tu Nota.

A receivable that clears in sixty days is a working-capital risk. A receivable that clears in fifteen years is a sovereign bond, and nobody repriced the collateral when it changed shape.

The generator's balance sheet now holds Honduran government paper instead of an account receivable from a utility. The credit backing that paper is not "ENEE collects tariffs and settles its bills." It is Sefin's own projection that ENEE will need L35,434 million in financing between 2026 and 2029, including L9,833.4 million in 2026 alone, much of it to cover exactly this debt. The same sovereign that let the arrears build is the counterparty on the instrument meant to resolve them.

The stress is already operational, not modelled

This is not a scenario five years out. One generator, Lufussa, reportedly cut output this year for lack of funds to buy its own fuel, per El Heraldo. Outages tied to the broader supply strain have had concrete costs, including a fire linked to a power cut at a market in Juticalpa, reported by La Prensa in July. A system where a generator can't afford fuel because its buyer hasn't paid it is not a tail case anyone needs to stress-test. It already happened.

What we think follows

Lenders financing Honduran private generation should stop treating ENEE arrears as a working-capital receivable the moment delay crosses a defined threshold, say ninety days, and start modelling the unpaid balance as sovereign duration exposure instead: shorter effective tenor assumptions, bigger reserve accounts, and security packages, trust structures, escrowed collections, that do more than assume payment arrives "eventually."

There's a fair counter here. A dated instrument with a stated coupon and maturity is at least a legible claim, better than open-ended arrears sitting on no instrument at all. Some generators are probably taking the swap precisely because a bond, however long its tenor, beats an invoice nobody will commit to a date on. That is a real improvement over the status quo. It is not the same thing as getting paid.

Our own pipeline sits in Peru, not Honduras, where COES runs dispatch independently of the distribution companies and generators that actually sign PPAs, a different structure from a single utility that is buyer, dispatcher and now bond issuer at once. ZFC Energy Group's 870+ MW Peruvian pipeline, including Olmos Wind's roughly 135 MW with integrated battery storage targeting Ready-to-Build in the first quarter of 2028, is underwritten against that separation of roles, not around it.

The open question

Does the L7,152.7 million bond program, and whatever trust restructuring follows it, actually clear the backlog and put ENEE on a cash-pay basis by 2027, or does it just stack another liability onto Sefin's own balance sheet without touching why ENEE spends more buying power every month than it collects in tariffs? I don't have a confident answer, and Sefin's own multi-year financing estimate, which keeps growing release over release, doesn't read like one either.

Either way, anyone signing a new Honduran PPA this year should be pricing the bond, not the invoice.

See the full active portfolio or read more Insights.