In 2024, I visited the Norwegian Embassy in Copenhagen with a banker who is a senior vice president of DNB’s energy division. He has a background in petroleum engineering rather than purely finance, and it tends to show in the way he patiently smiles when giving you a detailed answer to a simple question. His ability to connect details that few others would notice is probably one of the reasons he ended up responsible for evaluating project financing for CCS and other large-scale energy investments.

At one point during the evening, he mentioned a detail I had not given much thought until the ongoing critical minerals embargoes and the closure of the Strait of Hormuz: a strange mechanism in European banking regulations that is designed to make the financial system safer, but is making it more expensive to build the infrastructure Europe has decided it needs.

The mechanism is called Basel IV, and it was designed in part to prevent another 2008 by standardising risk weights across the European banking sector. But can standardised risk measures inadvertently raise the cost of financing the industrial infrastructure Europe needs for strategic autonomy? The answer depends on what happens when banking regulation and sectoral liability law collide.

We have seen this before in a different form in the Netherlands where the mining law directly applies to offshore carbon storage. The thing is, offshore carbon storage is much safer than onshore mining, so the risk model on the capital side gets completely disconnected from the actual physical risks of the operation. At the end of the day, this raises interest rates and ultimately the costs of carbon storage in the CCS space.

Banking risk

Banks have to maintain enough reserve capital to absorb losses when loads default, even large ones. How much capital that requires is determined internationally in Basel, Switzerland, by the Basel Committee on Banking Supervising (BCBS), ¹ which operates under the Bank for International Settlements (BIS). ² The standards they set are not legally binding in themselves, but national regulators implement them; in the EU through the Capital Requirements Regulation. ³

Basel IV

The current regulatory standards are a result of dealing with very different global economic situations since 1973. In 1988, Basel I introduced a simple risk floor where banks had to hold capital equal to a minimum of 8% of their risk-weighted assets (that includes interest bearing loans). Then in 2004, Basel II allowed banks with sufficient in-house modelling capabilities to use their own internal models to evaluate credit risk and calculate how much capital any given loan required on an individual basis to be rated within safety limits. It seems logical that a local bank with detailed knowledge about its borrowers can price risk more accurately than a regulator working from a broad categorical perspective. At least, it allowed for lower capital expenditures, especially for large infrastructure projects.

When the 2008 financial crisis happened, banks had been using their internal models for about four years to justify holding small amounts of capital against financial instruments (like mortgage-backed securities and collateralised debt obligations) which turned out to be much risiker than modelled. Part of the issue was model design, they were not structured to anticipate just how correlated failures might be, so when the housing markets failed, everything began failing. ⁴ The diversification benefits the models had assumed did not work and so, Basel III developed immediately after the crisis, raising the reserve capital requirements overall. Aside from capital levels being too low, there was another problem of the internal models producing risk scores that were too variable to supervise or compare across institutions, which can be desirable from a broader regulatory perspective.

A good few years later, Basel IV introduces an output floor that still lets banks use their own internal models, but they cannot let the those models produce a capital requirement below 72.5% of what the sector’s standardised models would require. This approach came partially into force at 50% in the beginning of 2025 and will increase to 72.5% by 2030. However, for larger corporations with revenues over €500 million, the option to use internal models is completely removed; those exposures must use the standardised approach regardless of how well the bank actually knows the borrower. So, any unrated company or project of that scale attracts a 100% risk weight, meaning the bank must hold capital proportional to the full loan value, irrespective of the borrower’s actual credit history. For project finance covering infrastructure, energy storage, and large-scale industrial investments, the standardised weights are up to 130% before a project is operational, falling to 100% once it is running (European Parliament and Council of the European Union_{2024, article 122a}).

Standardising risk scores across nations makes sense on many different levels, but there is a cost that comes better comparability. To compare risk across institutions, you need a common unit, and the common unit the standardised approach uses is categorical, not historical. So in our case, a project with stable cash flows, a long operational life, and a government-backed revenue floor is assessed in the exact same way as a speculative one of the same size. It is practical in most cases, but the methods become expensive when the project carries a kind of liability the category was not designed to for, and that comes to light in a few different examples.

Dutch mining act

The Dutch Mining Act, the Mijnbouwwet, was passed in 2003, more than a decade before the Basel Committee began finalising Basel IV. It was written by people thinking about geology, seismicity, and long-term environmental liability, not about capital ratios. Its job is to regulate subsurface activity in the Netherlands: oil and gas extraction, salt mining, geothermal energy, and since a 2012 amendment, the permanent storage of CO₂.

Offshore carbon storage is, in physical terms, considerably safer than onshore mining. Depleted North Sea gas fields are stable geological formations with proven containment records. Industry estimates put the probability of significant leakage at a well-managed offshore site below 0.01%. But the Mining Act does not apply different standards to offshore storage. It applies the conservative framework developed for onshore mining—stricter scrutiny on soil movement, seismicity, groundwater effects—to a context where those risks operate at a different order of magnitude entirely.

What this produces is a category of liability that is structurally unlike almost anything else a bank finances. A CO₂ storage licence under the Mining Act does not simply grant permission to inject. It creates obligations that do not end when injection stops, cannot be fixed at a known amount at the time the loan is made, and can be expanded by the regulator at any point if scientific understanding changes. Three provisions carry the weight of this.

Article 31h requires the minister to withdraw or amend a licence if there is leakage, non-compliance, or—critically—“new scientific findings and technological progression.” There is no temporal limit on this trigger. A storage operator whose site is performing well can still face a mandatory licence change because the science around geological storage has moved on. This is entirely reasonable as environmental policy. It is very difficult to price as a financial liability.

Article 31j means the licence cannot be transferred to the state until at least 20 years after injection has stopped, and only then if the operator has pre-funded 30 years of subsequent monitoring costs. Article 31k adds that even after state handover, the previous licence holder remains potentially liable for costs the state later incurs. The total liability tail runs to 50 years or more, with a final cost that is indeterminate at origination. ⁵

The Act also maintains what is known as an ALARP standard: risk must be reduced to a level where further reduction would be “considerably disproportionate relative to the benefit.” This standard is itself re-adjudicable as technology and science evolve. What was acceptable in 2025 may not be acceptable in 2035.

This is not bad law. For onshore mining in one of Europe’s most densely populated countries, conservative long-tail liability provisions are entirely defensible. The problem is not the Mining Act on its own terms. The problem is the same one Basel IV creates from the other direction: a risk model calibrated for one context, applied to another where the actual risk operates at a different order of magnitude entirely.

Higher interest rates

In practice, both frameworks apply simultaneously to the same project. Each independently overstates the risk; together, they compound it.

The project vehicle for a CCS hub is typically an unrated SPV—a special purpose vehicle created for the project, with no credit history of its own—which places it in the 100% risk weight bucket under the standardised approach, or 130% during the pre-operational phase. Basel IV’s category-based model sees an unrated project vehicle and assigns maximum capital, regardless of the project’s actual cash flow stability or revenue structure. The Mining Act, meanwhile, has already required the operator to post substantial financial security—insurance policies, parent company guarantees, monitoring funds—to cover the long-tail liability the licence creates. That security does not reduce the bank’s capital requirement, because the standardised approach’s collateral recognition rules were not built to accommodate regulatory compliance instruments of this kind.

The result is that the bank must hold maximum capital against a project already legally required to provision for its own worst-case outcomes. Each framework, taken alone, overstates the risk of the specific thing it is pricing. Together, they produce a financing cost that bears little relation to what the project actually involves.

Project risk

When the cost of financing exceeds what a project’s economics can absorb, the risk does not disappear. It migrates.

Porthos and Aramis

In the Netherlands this is visible in how CCS infrastructure actually gets built. Porthos, the first large-scale offshore CCS hub in the EU, reached a final investment decision only with a combination of a €102 million EU grant, SDE++ subsidies covering the gap between CCS costs and the EU carbon price, and additional state guarantees during permitting delays. A 2024 audit by the Dutch Court of Audit confirmed that the project’s economics depend on continued public support. The insurance and financial security costs required under the Mining Act—estimated by industry analysts as several orders of magnitude higher than the actual statistical risk—are absorbed into the overall project cost structure that subsidies then have to cover.⁶

Aramis, the larger follow-on hub, provides a clearer signal. Shell and TotalEnergies—two companies with investment-grade credit ratings, large balance sheets, and direct commercial incentives to decarbonise—scaled back their commitments after the initial project phase. In April 2025 the Dutch government stepped in with approximately €650 million in direct support to keep the project viable.⁷ If two of the largest energy companies in the world find the commercial terms unattractive, it functions as a reasonable proxy for what the wider bank lending market is pricing.

Danish CCS tender

In Denmark, the dynamic has taken a different form. CCS tenders there have placed post-closure liability onto public waste-to-energy companies; municipal utilities that were not designed to carry long-tail geological risk and whose ultimate backstop is the taxpayer. In a conventional state-backed project, the government bears the risk directly. In a conventional commercial project, a private entity bears it in exchange for adequate compensation. The Danish structure created a third category: quasi-public entities holding liability they lack the mandate or capitalisation to properly absorb.⁸ The pattern is the same in each case. The risk does not go away when commercial banks price it too high. It shifts onto public balance sheets, state subsidies, and entities not built to hold it.

Gift of foresight

Basel IV is a reasonable response to a real problem, and it was designed to address the mispricing of correlated financial risk in a highly interconnected banking system. But it is a calibrated for risks of 2008, and now Europe is facing an entirely different category of systemic risk.

Trade embargoes

The closure of the Strait of Hormuz and the critical minerals embargoes are not credit events. They do not appear in any risk weight. A bank holding a loan to a CCS infrastructure project in the Netherlands—a project enabling European heavy industry to decarbonise without relocating—faces higher regulatory capital requirements than a bank holding a loan to an unrated importer of Chinese industrial inputs, despite the geopolitical logic pointing in the opposite direction.

This is not an argument that the importer’s loan is riskier in credit terms. Under the standardised approach, they may carry identical risk weights, and the financial credit risk may genuinely be similar. The argument is narrower: the standardised approach’s unit of measurement is credit default probability, and credit default probability is blind to strategic dependency, supply chain concentration, and the long-term cost of industrial capacity erosion. These are real systemic risks. They just do not show up in the output floor.

The United States, through the Inflation Reduction Act, has partially addressed this asymmetry for its own industrial base. IRA production and investment tax credits create a government-backed revenue floor for qualifying projects, structured in a way that banks can factor into their underwriting. A bank financing a comparable CCS or battery storage project in the US is looking at a meaningfully different risk profile, because the revenue support is designed to be bankable. European equivalents—the SDE++ subsidy in the Netherlands, competitive tender schemes elsewhere—are valuable but less straightforwardly bankable as collateral. Basel IV’s standardised approach does not distinguish between these two situations.

Development finance

The natural objection is that development finance exists for exactly this. The European Investment Bank, KfW, and national promotional banks have long provided financing for complex infrastructure that commercial banks cannot efficiently price. If Basel IV pushes commercial banks toward the exit on CCS and industrial infrastructure, perhaps the answer is simply more development finance.

This argument has real force. Development banks play an essential role, and there is a defensible case that the division of labour—commercial banks for standard credit, development banks for complex long-dated infrastructure—is broadly sensible. The difficulty is scale. The Clean Industrial Deal, the Net-Zero Industry Act’s 50 million tonne CO₂ storage target by 2030, the Critical Raw Materials Act’s processing ambitions: these require a volume of financing that development bank balance sheets alone cannot supply. A market for strategic infrastructure that depends entirely on public institutions is itself a form of concentration risk: political constraints, sovereign borrowing limits, and the absence of the commercial discipline that comes from private capital at stake.

The question is not whether public finance should play a role; it clearly should and does. The question is whether the current framework is producing the right balance, or whether the unintended consequences of Basel IV are pushing it further toward public dependence than either the banking regulators or the industrial policymakers intended or would choose.

Conclusion

The banker was not making a political argument when he mentioned this in Copenhagen. He was describing a technical friction he encounters in his work: risk models that were carefully designed to measure one thing, being applied to something else, producing financing costs that nobody calibrating those models intended.

Basel IV was a reasonable response to a real failure. The Mining Act is a reasonable framework for managing the long-term risks of subsurface intervention. Both were designed by people trying to do their jobs responsibly. Neither was designed with the other in mind, and neither was calibrated for a moment in which Europe’s industrial capacity has become a question of strategic autonomy.

Risk frameworks, like maps, are drawn for the terrain that mattered when they were made. What they cannot capture is the terrain that had not yet become relevant. The cost of that gap does not stay abstract. It shows up in state subsidies for projects that should be commercially viable, in public utilities holding liability they were not built to carry, and in CCS hubs that reach construction only after governments step in to make the numbers work.

None of that appears in the Basel IV output floor. It is, in that sense, a risk that has been standardised out of sight.