The sustainability discount, and what it is worth.

One controls where the money goes. The other watches what the company does.

A green loan is a use-of-proceeds product. The facility agreement names the eligible categories, the borrower draws against qualifying spend, and the money is tracked to those assets. If you are buying rooftop solar, replacing gas heating with heat pumps, retrofitting a warehouse or moving a fleet to electric vehicles, the loan is green because the spend is green, and the company’s wider carbon performance is not the test. Nothing about the rest of the business has to change for the facility to qualify.

A sustainability-linked loan works the other way round. The proceeds are unrestricted, so the money can refinance existing debt, fund an acquisition or sit in a revolver, and the sustainability content lives in a set of key performance indicators with targets attached to them. Emissions per unit of output, energy intensity, waste to landfill, accident frequency, a recognised third-party sustainability rating. Hit the targets and the margin steps down. Miss them and, in most mid-market structures, the step-down simply does not apply. The convention is two to four targets rather than a long list, which is a deliberate correction: early deals set many loosely-connected indicators, and the market has moved toward a small number that are material to the business.

Neither label is a legal status. Both are defined by voluntary frameworks published by the Loan Market Association with its American and Asia-Pacific counterparts, and a lender applies them by contract rather than by regulation. That matters more than it sounds, because a label that is granted by agreement can also be withdrawn by agreement, which is exactly what has started happening.

An external verifier, every year, at your expense.

The obligation side of the trade tightened in March 2025, when the Loan Market Association and its counterparts revised the Sustainability-Linked Loan Principles. Verification of performance by a qualified external reviewer, both before signing and after it, moved from “should” to “shall”. Independent verification of performance against each target is now mandatory at least annually, running until the last target date. The revision also clarified that information already assured through published annual reports or regulatory disclosure does not need verifying twice, which is a real saving for a company that already reports, and no help at all to one that does not.

Three costs follow from that, and none of them appear in the term sheet. The first is the verifier’s fee, paid annually by the borrower. The second is the internal work of producing data an external reviewer will sign off: for a company that has never measured its energy use by site or its emissions by scope, the first year is a project rather than a report. The third is the negotiation itself, because targets have to be calibrated at signing, and a lender that intends to grant a discount will want the trajectory to be demanding enough to be worth granting. Legal time spent on the schedule is legal time not spent on the covenants.

There is a fourth cost with no invoice attached. A published target is a public commitment, and a company that misses one has created a fact about itself that its lenders, its customers and its own board can see. That is manageable when the targets were set on real operational plans. It is corrosive when they were set to win a discount worth a few thousand pounds.

Smaller, and better built, at the same time.

The direction of the market is the single most useful fact for a borrower deciding whether to chase a label. Volumes have fallen hard. Global sustainability-linked loan issuance dropped from US$530bn in 2024 to US$418bn in 2025, and the contraction was concentrated in exactly this instrument rather than in sustainable debt generally. Part of that is the arrival of 2025 target dates, which turned aspirational trajectories into measurable outcomes. Part is simple arithmetic on the part of borrowers who concluded that a discount of a few basis points did not repay the reporting it required, and quietly dropped the label at refinancing.

What remains is better built. In August 2025 the Financial Conduct Authority published a follow-up letter on the market concluding that structures had matured since its 2023 review: targets are now more material to the borrower and fewer in number, banks are applying stricter criteria and declining poorly-structured deals, and some lenders are stripping the label from loans that fail to meet it. The regulator was equally clear about the limitation. Pricing remains a challenge, and the margin adjustment for meeting or missing targets is still minimal. Both halves of that finding matter: the label is harder to obtain than it was, and worth roughly what it was worth.

One practical consequence shows up in drafting. Where a borrower and lender want the option without the commitment, facility agreements increasingly carry a dormant provision, an agreement to agree a sustainability framework later, with the ratchet inserted by amendment once targets are settled. That keeps the door open at no cost, and it is a sensible default for a company that expects its reporting to improve over the life of the facility.

Take the discount you already qualify for. Do not build a reporting function to earn it.

The decision rule is close to arithmetic. If the company already measures what the targets would measure, because it reports under an existing regime, because a large customer requires the data, or because the board tracks it anyway, then the marginal cost of a ratchet is the verifier’s fee and some drafting. Against a saving of five figures a year on a £10m facility, that trade is usually positive, and the label costs nothing in credibility because the numbers are already being produced. If the company does not measure these things, the first year of measurement will cost more than the discount pays, and the honest answer is to decline the label and revisit it at the next refinancing.

Where the label is being taken, four terms carry the value. Insist the ratchet is downward-only, so that a missed target forgoes a saving rather than adding cost. Insist that failure to deliver a report switches the benefit off rather than creating a default or a drawstop, which is the mid-market convention and should not have to be argued for. Calibrate the targets against the operating plan the board has already approved, not against an ambition invented for the facility agreement. And fix the treatment of a change in the business, because an acquisition, a disposal or a site closure moves an emissions baseline sharply, and a recalculation mechanism agreed at signing is far cheaper than an amendment negotiated after the event.

The one thing not to do is let the label decide the lender. A sustainability ratchet from a lender whose margin is fifty basis points wide of the market is a discount on an expensive loan, and the arithmetic in the second chart above settles that comparison before the sustainability schedule is opened.