Who Sets the Rules in Blue Finance

In April 2023 the federal cabinet approved Roberts Bank Terminal 2, a three-berth container terminal to be built on an artificial island in the Fraser delta, its land base raised by dredging and infilling roughly 168 hectares of intertidal and subtidal habitat. The independent review panel had found significant adverse effects in about 30 areas, among them wetland function and juvenile Chinook salmon, where it judged the harm high in magnitude, permanent in duration and irreversible. The Minister of Environment and Climate Change determined the project was likely to cause significant adverse environmental effects, the Governor in Council decided those effects were justified in the circumstances, and the Decision Statement set 370 legally binding conditions. Two sets of rules describe a port project like this one. One is international, voluntary and already written. The other is Canadian, voluntary, and being written now, with the criteria that would reach a port still unpublished.

The document an issuer works from is called Bonds to Finance the Sustainable Blue Economy: A Practitioner's Guide. It appeared in September 2023 and ten people wrote it, working for the Asian Development Bank, the International Capital Market Association, the International Finance Corporation, UNEP FI and the UN Global Compact. The guide keeps the words blue bond inside quotation marks on every page, and a footnote gives the reason: blue is a theme under the existing bond principles rather than a class of its own. A green bond whose proceeds all go to ocean projects can be called a blue bond, and the issuer decides.

The Sustainable Blue Economy Finance Principles are older, launched in 2018 and running to fourteen points. UNEP FI hosts them but did not write them. They came out of the European Commission, WWF, the World Resources Institute and the European Investment Bank. A bank signs by declaring that it will endeavour to direct lending toward projects contributing to Sustainable Development Goal 14, the ocean goal, and will endeavour to report on how that is going. The signatory declaration then states plainly that the principles are voluntary, create no rights or liabilities, and leave every investment decision with the institution signing.

The guide sets out eight project categories, and it bounds them with distances. A coastal adaptation project has to sit within 50 kilometers of the coast or in the water. Move inland past 100 kilometers and a wastewater plant stops qualifying. For runoff from farmland the line is 200 kilometers from the coast, or 50 kilometers from a river that reaches the ocean. The guide does not say where the distances come from. Ports are category seven, filed under the Green Bond Principles heading for clean transportation.

Alongside the categories runs a list of what cannot be financed. The definition of a sustainable blue economy excludes non-renewable extractive industries and names offshore oil and gas, dredging and deep-sea mining. The reference behind that line is UNEP FI's work on marine extractives, where dredging means seabed aggregate extraction as an industry, so the exclusion does not by itself reach construction dredging for a port. The port category carries its own exclusions and they are the ones that bite. A port with air pollution fines is out, so is one that loses IUCN red-listed habitat in its development, and so is building on a greenfield site. No one has proposed a blue bond for Roberts Bank. The terminal would rise on an artificial island built over intertidal and subtidal habitat, which is the closest thing in the guide to a disqualifying fact, although the guide never defines greenfield.

All of this is voluntary, and the word covers more than one thing. Adopting the guide is optional. Its own steps are recommendations, since it recommends an external review before issuance and recommends annual reporting on allocation and impact. An issuer decides for itself whether a bond is green or blue. After adoption the picture firms up, though unevenly. An issuer publishes a bond framework and reports against it, and whether any of that carries a contractual consequence depends on the terms of the particular instrument rather than on the guide. What the guide cannot supply is a way to close the loop. Allocation reporting establishes where the money went. Impact reporting happens and is recommended, but the guide concedes that no globally accepted list of impact metrics exists for these projects, since they run from seafood to tourism to marine protected areas. A bond can report allocation and impact in full without anyone establishing that the ocean is in better condition than it would otherwise have been.

The amounts are small. Transactions labelled blue reached USD 5 billion between 2018 and 2022, against USD 2.2 trillion in cumulative green issuance by the end of that period. What a blue label delivers is set out in four parts, carried over from the Green Bond Principles: a commitment on where the proceeds go, a process for choosing the projects, a method for managing the funds, and reporting on both allocation and impact. That discipline can make financing more transparent without establishing that the ocean is better off.

The guide is candid about what remains unsettled. Its own text notes that institutions adopt different scopes and definitions, and names where they diverge: extractive industries, sanitation, and water management. Freshwater is the clearest case. The IFC's Guidelines for Blue Finance cover freshwater and the ocean together, drawing on the clean water goal as well as the ocean one. Only ocean projects supporting Goal 14, Life Below Water, fall inside the joint guide, which acknowledges even so that blue bonds have been issued for freshwater under the IFC's rules. The IFC updated its guidelines to a second version in September 2025. Nothing in the joint guide has changed since 2023.

The OECD guidance has a different audience. It sets four goals for a sustainable ocean economy, covering economic development and resilience, equity, healthy marine and coastal ecosystems, and climate action, and addresses them to development co-operation providers, meaning the agencies that spend aid money abroad. Canada is one of those. The guidance speaks to Canada as a donor deciding where aid goes, and not to a Canadian bank deciding what to lend against at home. The difference is institutional rather than national, since Canadian issuers sit inside the joint guide's stated audience, which names sovereigns, sub-sovereign agencies, banks and companies.

Nothing requires any of it to be applied to Roberts Bank. The terminal is not financed through a labelled bond and no one has proposed that it should be. The Vancouver Fraser Port Authority is not a signatory to the principles. A Canadian bank lending against a port on this coast might find the guide useful and might treat the port exclusions as a sensible checklist, and nothing obliges it to open either document. What Canada has instead is a rulebook of its own, currently being written.

Canada's sustainable finance taxonomy has been in the open since July 9, 2026, when the draft methodology report went out for public comment through August 13. The federal government funded the work in December 2025, the Canadian Climate Institute writes it, and an independent council of 17 chaired by Marlene Puffer approves what goes out. It is also voluntary. Its potential reach comes from a federal mandate, sector-specific technical criteria, and the prospect of Canadian institutions using it to classify what they finance. Climate mitigation is the first phase's sole positive environmental objective. Six sectors come first: electricity, buildings, transportation, mining, manufacturing, and agriculture and forestry. They were chosen for their emissions, their potential to decarbonize or help other sectors do so, and investor interest. Ports are transportation, and transportation is in the first three.

The ocean is in the report twice, in different roles. It appears in a sentence about what follows the first phase, which says the taxonomy will be expanded to include climate resilience, water and marine resource protection, biodiversity protection, pollution control and the circular economy. No date attaches to any of them. It appears again inside phase one, not as an objective but as a harm test. The proposed do-no-significant-harm framework covers biodiversity and healthy ecosystems, water conservation, pollution, adaptation and the circular economy. Its generic criteria would be grounded in Canada's existing laws and policy, and the report's own example is an environmental impact assessment for large-scale infrastructure, with sector-specific harm criteria to follow alongside the technical screening criteria. The report asks for comment on whether any of it should go beyond existing law. What failing a harm criterion would cost is not settled in this draft. The framework proposes partial alignment, under which a user discloses which criteria are met and which are not, in preference to the binary approach the report attributes to the EU, where violating one criterion makes a project ineligible.

The closest thing I have to experience of this is LEED certification on new buildings. It was voluntary, it cost more to build to, and the market was not willing to absorb the difference. My heart was with LEED every time. The numbers usually did not support it, and a borrower who wanted it was choosing to spend money that would not come back in the value of the building. A voluntary label does not change a credit decision by itself. It starts to when a valuation, a lender's policy or a pricing difference makes it worth something.

The two frameworks ask different questions of the same berth. The international guidance is ocean-specific from the start. It sets out eligibility criteria for blue projects, and its port exclusions speak to habitat loss and to building on greenfield sites. The Canadian taxonomy starts from climate. Environmental harm enters through safeguards grounded in existing Canadian law, with sector-specific criteria still to come. At Roberts Bank existing Canadian law produced 370 conditions and a CAD 150 million guarantee covering the first three years of construction, which is a real instrument and a bounded one. Consultation on the transportation criteria runs this fall, with final criteria targeted for January 2027. The separate question of which environmental objectives should follow climate mitigation was put to the public in the methodology consultation, which closed on August 13, 2026. The report commits to adding water and marine resource protection. What has not been set is when marine protection becomes an objective in its own right.