Earlier this year, the first paper from UCL Energy Institute’s Shipping and Oceans Research Group and Strider Carbon showed how transition-related stranded asset risks could manifest as traditional banking risks. Simplified liquidation scenarios in the paper mapped out losses for lenders and equity holders in typical ship finance structures. It also set out how specific characteristics of shipping debt structures could lead to system-level risks.
The second paper, published today, follows up with lessons from history. It examines four episodes in which capital-intensive assets lost value faster than their owners and financiers expected — the German and UK electricity transitions, the collapse of the German KG shipping funds in the early 2010s, and the bankruptcy of Hanjin Shipping in 2016. Strider Carbon led the work on the two shipping case studies, and this blog unpacks those further.
The questions that follow are the ones ship finance practitioners are likely to ask when they encounter the KG and Hanjin case studies and lessons drawn from them. And they deserve careful answers.
The KG collapse and Hanjin had nothing to do with climate. Why would they tell us anything about transition risk?
Causal mechanics: Risk convergence
KG and Hanjin crises came about from the convergence of several factors, external and internal to shipping. Transition risk, too, would arrive through converging factors: policy, technology, their timing, and the market cycle they land in. The triggers may be different but the way converging pressures translate into defaults and devaluations may not.
Cascade: Impact across capital stack
These are historical recordings of how devaluation events and losses travelled from ships to financiers; through equity first but not stopping there. Understanding that path could point to where protections thin out.
Transmission mechanics: Specific to financial structures rather than a crisis
Leverage calibrated to ordinary shipping cycles meeting devaluations beyond them, failed refinancings, maritime liens outranking mortgages, and collateral values that fall further because many lenders need to sell — all of these operate the same way regardless of who the borrower is or why they fail.
What happened in the KG collapse? And why did losses reach banks that provided senior and secured debt?
The KG system channelled roughly EUR 30 billion of retail equity into single-ship funds between 1993 and 2009. Around 440,000 private investors were involved, financing at one point more than 20% of the international container fleet. High fees in the KG structures meant thinner buffers to absorb losses than the headline equity suggested. When charter rates collapsed after 2008 (with the tax advantages already curtailed), equity absorbed the first losses, as it is designed to do.
But the losses did not stop there. Vessel values had collapsed alongside charter income, so the ships securing the loans were often worth less than the debt outstanding. Seniority meant banks were hit later, not that they weren’t hit. And the banks’ remaining option, which was to ‘enforce and sell’, worked for one fund but not for hundreds: selling into a market where everyone was selling would deepen the very devaluations the banks were trying to avoid. Caught between recognising losses and deepening them, most deferred, and the workout ran for years.
The scale tells the story. Over 350 funds declared bankruptcy in 2012 and 2013; by 2016, a fifth of the containerships owned by German funds were insolvent. The division of Commerzbank partly responsible for ship financing lost close to EUR 850 million in 2009 alone. HSH Nordbank (then the world’s largest ship financier) doubled the amount it set aside for expected loan losses to EUR 1 billion in 2009, against around EUR 35 billion of ship loans on its books, many of which its own auditors considered at risk. Provisioning at roughly 3% of the loan book showed a recognition of the risk. It is unclear what share of the at-risk loans this covered, but what followed indicates that it was insufficient: the bank cut its shipping exposure by 40% within four years after this.
Hanjin’s banks were financial owners of the vessels in the bareboat charter structures. Why didn’t that protect them?
This is one of the more uncomfortable findings. When Hanjin filed for bankruptcy in 2016, it was the world’s seventh-largest container line. It was operating 196 vessels, of which it owned five outright. Korean banks were exposed both as lenders and as beneficial owners of vessels under BBCHP (bareboat charter with purchase option) structures. In practice, maritime liens for crew wages and bunker debts ranked ahead of their claims in many jurisdictions; ports refused vessels for fear of non-payment; $14 billion of cargo sat at sea, affecting 8,300 cargo owners. Creditors eventually filed claims totalling $10.5 billion. Around $220 million (roughly two cents on the dollar) was recovered. Ownership of collateral, under system-wide stress, turned out to be a weaker protection than it appeared on a term sheet.
So, were banks careless?
No. None of this is an argument that banks are doing something wrong.
In both collapses, the record shows institutions behaving the way well-run ones do. German banks deferred enforcement to avoid crystallising losses through fire sales: individually reasonable. Banks lent at conservative loan-to-value ratios against real assets: prudent. Individual lenders kept tenors short and expected to refinance or exit before trouble arrived: rational. The failure was not in any single decision but in what the decisions added up to when several institutions behaved similarly under similar circumstances: simultaneous refinancing needs, correlated enforcement choices, and collateral markets that stayed liquid only while few participants needed them. Transaction-level resilience and system-level resilience turned out to be different things, and the second did not follow from the first. That is why the useful response is architectural — shared scenarios and coordinated standards, which individual institutions can then tailor to their own strategies and risk appetite.
The transition will be gradual, not a crash. Doesn’t that change everything?
The paper notes that a decarbonisation-driven stranding may unfold more gradually than a market crisis, potentially giving stakeholders more time to respond. But gradual arrival only helps when institutions recognise early and already know how they will act under different scenarios. In the electricity case studies in the main paper, stranding began when renewables were still around 10% of generation, and credit rating agencies did not flag German stranded asset risk until 2014, nearly a decade after the signal was visible.
Whether shipping’s threshold sits at the same level is genuinely uncertain. What the record does not support is the assumption of clear and timely warning, or that players will act on the warning they get, or that individually sensible responses will add up safely at system level.
Why did the Korean banks come out of Hanjin relatively lightly?
Hanjin’s creditor banks cut support early, booked losses promptly and provisioned against nearly their full exposure. The cost of Hanjin’s failure to their balance sheets was contained to around 11 basis points of their combined risk-weighted assets. German banks, deferring enforcement to avoid fire sales, spent the better part of a decade working out the consequences. The comparison has limits — Hanjin was a single exposure, while the KG collapse ran through hundreds of funds at the core of the German banks’ shipping business, making interventions both structurally harder and necessary at scale. But early loss recognition could be a lever to help contain damage.
Ships can be retrofitted. Doesn’t that mitigate risk?
Yes, partly, and it is a real advantage. But retrofit options have expiry dates. Yard capacity is limited, costs rise with regulatory deadlines, and a retrofit deferred long enough converges with the replacement decision. And when global regulation impacts the entire fleet simultaneously — even if regional variations offer some optionality in the interim — it has the potential to create market conditions and synchronise lender behaviour, as in these historical instances.
What is the cost of not acting early for shipowners?
Owners’ capital is where the losses landed first
Equity absorbed the first and largest hits in both shipping collapses, so the cost of being wrong impacts owners more heavily than their lenders. The earlier paper’s liquidation scenarios showed this formally.
The costliest decisions were the ones made while assuming continuation
Ordering into a market already oversupplied, contracting long-term charters after rates had fallen. Prolonging the status quo did not preserve it; the continuation decisions themselves fed the prolonged downturn that followed.
Size changes how the waiting works, not whether it costs. Larger, diversified owners retain options while they wait. They have access to bank finance and room to reposition fleets gradually. Smaller owners with single-segment fleets face a narrowing version of the same choice: as lenders concentrate on diversified operators with documented transition plans, waiting can quietly become exclusion from conventional finance rather than a preserved choice. As Strider Carbon’s resilience post argues, waiting for certainty is itself a decision — and one that may mean far less control when it is made.
What could a ship finance institution take away from these historical cases?
Portfolio assessment: Recognise both supply- and demand-side risk at segment level
The KG collapse shows what concentrated exposure to a single segment can cost when the market turns. Supply-side risk (vessels losing competitiveness) and demand-side risk (cargoes losing volume) affect segments differently, and the historical record suggests these risks were recognised but their scale was underestimated. The earlier paper also found that banks’ mitigating actions lean heavily towards supply-side risk, so demand-side risk assessment may need deliberate attention.
Loss recognition: Decide posture and decision triggers in advance
While the KG and Hanjin scenarios are not directly comparable (being one company versus hundreds of funds), they do suggest preparedness matters. Institutions that already know how they would act in different situations, and what their decision triggers would be, are more likely to act early when one of those situations arrives. A bank that already knows it could take measures A, B and C when a risk extends to a defined share of its portfolio is positioned to act faster in a crisis.
Refinancing concentration: A portfolio-level risk that transaction-level checks can miss
Loan tenors of 5–7 years against vessel lifespans of 20–25 years mean that an individual bank can exit before risks materialise, but the system as a whole cannot. The question is not only whether individual transactions are protected. In these historical cases they largely were, on paper. But what would the portfolio look like if many institutions made similar decisions at the same time?
An interview with SeaNews, which explored these cases further, is here.
The full report is available here. As with the first paper, practitioner perspectives on the shipping case studies are welcome, particularly from anyone who lived through either collapse from inside a lending institution. These case studies pieced together information from peer reviewed papers, news articles, opinion pieces as well as grey literature; but institutional memory is richer, and likely has more to teach.