Germany Blocks COSCO's 80% Bid for Hamburg Operator Zippel on Supply-Chain Grounds

The Analysis

Germany has stopped COSCO from buying eighty percent of Konrad Zippel. The stated ground, given by the Economy Ministry, is that the acquisition would deepen dependencies and put German and European supply-chain resilience at risk. That is what we know. The Bundeskartellamt cleared the same purchase in February. Those two sentences belong next to each other, and most of what has been written in the last few days assumes they contradict each other. They do not, and understanding why they do not is the difference between reading this as a political story and reading it as a legal one.

Zippel is a Hamburg company that does not own a quay. It moves containers between Hamburg and Bremerhaven and the German hinterland by road, rail and inland waterway, with roughly two hundred trucks of its own, about three hundred and fifty employees, and around two hundred and five thousand TEU handled in 2024. Its corridors serve Berlin, the Halle-Leipzig chemical belt, and Bundeswehr logistics. COSCO already holds twenty-four point nine nine percent of the Tollerort terminal, a minority position that went through years ago. Put those four figures side by side and the shape of this decision stops looking like a ban on Chinese capital in European ports.

I am not offering a view on whether the decision is sound policy, and I am not going to pretend to a prediction about what happens next. What I can do is draw the perimeter, which is the useful part of any screening decision: what exactly was decided, under which instrument, what the stated ground needs to be true, which transactions sit outside it, and under what conditions every conclusion currently being drawn from it collapses. Start with the two rulers, because nearly all the confusion comes from using one as though it were the other.

Germany runs two separate reviews over a transaction of this kind, and they have nothing in common except the target company. One is merger control, carried out by the Bundeskartellamt. The other is foreign investment screening, carried out by the Economy Ministry. Strictly speaking they ask different questions. Merger control asks whether a concentration will significantly impede effective competition: how the relevant market is defined, what shares the parties hold in it, whether enough credible alternatives remain afterwards. Foreign investment screening asks whether an acquisition by a non-domestic investor impairs public order or security in the acquiring state. Same share purchase, same closing date, two tests with two different objects of review.

It follows that there is nothing odd in February's outcome followed by this one. A transaction can be entirely innocuous on competition grounds and still fail the security test, and it can be blocked on competition grounds and pass a security test with no difficulty. The two assessments are not sequenced versions of each other, they are parallel. Anyone reporting the February clearance as evidence that Berlin changed its mind has misread which instrument did what. I am deliberately not giving you section numbers or quoting provisions here, because I have not read the decision itself and I have no intention of reconstructing a legal basis I have not seen. The structure is what matters, and it is not in dispute.

Now to the stated ground, which is where the analysis actually has to happen. The ministry said the transaction would deepen dependencies and endanger resilience. There is a premise embedded in that sentence, and the premise is forward-looking. It does not say that anything has been harmed. It says that acquiring this company would increase dependence, and that increased dependence carries risk. Here is an exception worth holding on to: that formulation is very hard to disprove, because the harm is hypothetical and future, and administrative practice almost everywhere grants the executive wide latitude on predictive security judgments. That is not a criticism of Germany specifically; it is how these instruments are built. The practical consequence for you is that the reviewable surface here is small. An applicant arguing "no harm has occurred" is answering a question nobody asked.

So which elements carry the weight. Reading the facts in the report, three are present that together make this decision comprehensible. The target controls hinterland capacity on Hamburg's inland corridors, by truck, rail and barge, which is the physical link between quay and factory door. Those corridors touch Berlin and the Halle-Leipzig chemical belt, so the link runs into nationally significant industrial geography. And the target handles Bundeswehr logistics, which puts it adjacent to defence supply. Remove any one of the three and the argument stands less well. Remove the defence element in particular and what remains is an acquisition of a mid-sized road-rail-barge operator, the sort of transaction that closes in Germany most weeks without anybody outside the parties noticing. I am not prepared to say which element was decisive, because I have not seen the reasoning. I am prepared to say which one the outcome most plausibly turns on, and it is the one that has received the least attention.

Size deserves a paragraph, because this is where intuition misleads. Two hundred and five thousand TEU in 2024. Take Hamburg's annual container throughput as roughly seven point eight million TEU — I am using round numbers for order of magnitude, do not quote my figure to four places — and Zippel represents around two point six percent. Two hundred trucks. Three hundred and fifty employees. There is nothing strategically enormous about this company measured by volume or by headcount, and it was blocked.

That is the observation nobody has made interesting use of, and it carries the practical lesson. If a screening authority prohibits a company handling two point six percent of a port's boxes, then the test is not running on size. It is running on position in the chain. Every due diligence model I have seen built on turnover thresholds and headcount screens would have cleared this target comfortably. Jurisdictional interest attaches to what an entity does, not how large it is. Two hundred and five thousand TEU is negligible against a terminal; it is the entirety of the last sixty kilometres into a chemical belt for whoever rides those corridors.

Which brings me to what the coverage has missed. Every headline I have read files this under geopolitics: China being pushed out of European ports, Berlin hardening its line, screening returning after a quiet spell. Not one of them has said the thing that distinguishes this decision from every other Chinese-investment-in-Europe story of the last decade. What was refused is not quayside access. It is the port-to-door leg. Zippel owns no berth, no crane, no concession. It owns trucks, rail slots, barge capacity and the inland relationship, which is to say it owns distribution reach rather than port capacity.

Read that next to Tollerort. A twenty-four point nine nine percent stake in an actual container terminal was permitted. Eighty percent of a hinterland carrier has been refused. Those are two data points, not a rule — the assets are different, the years are different, the political composition of the government is different — and I would not build a policy thesis on them. But the gap between them is where the actual boundary runs, and it sits somewhere other than where everyone is pointing. Quayside minority participation, tolerated. Majority control of inland distribution into industrial and defence-adjacent corridors, not tolerated. If you are an exporter planning European operations, that distinction will matter more to you than any amount of commentary about geopolitical posture.

What follows from that is a control test rather than a nationality test. Nothing in this decision turns on the flag of an investor in the abstract. Everything turns on control of a specific kind of asset. Hold eighty percent of a hinterland operator with defence-adjacent customers and you are one place. Hold twenty-five percent of anything with no governance rights and you are somewhere else entirely, and calling both "Chinese investment" does not make them legally comparable. This matters because the due diligence question changes shape. The relevant inquiry is not whether the counterparty has Chinese shareholders; it is whether the structure confers control, and control over what.

Now the routes back in, because nothing here is necessarily final. Carve out the Bundeswehr logistics activity and sell it to somebody else, and the element I described as load-bearing disappears; the same trucks, the same corridors, the same eighty percent may then produce a different answer. Restructure below control with no board rights and a different legal test applies to the start. On judicial review, this class of decision is reviewable in principle, but I have not seen the reasoning and I have no interest in predicting how much deference a court will show a ministry on a predictive security assessment — historically a great deal, and I will leave it there. Treat both directions as open. I would not tell you this is permanent, and I would not tell you it will be overturned.

Equally worth stating plainly: what this decision does not touch. It does not restrict chartering capacity from Chinese carriers. It does not restrict slot agreements, commercial agency arrangements, or ordinary service contracts between a European shipper and a Chinese or Chinese-owned inland operator. It says nothing about shipping line shareholdings in terminals beyond the position that already exists. It prohibits one acquisition, by one acquirer, of one company, on one stated ground. There is a widely available reading that concludes Europe has closed itself to Chinese logistics investment. That reading is not supported by this decision, and the twenty-four point nine nine percent sitting two lines away from it is the evidence against that reading.

On the practical side I stop, and I stop deliberately. Whether any of this changes how a particular shipment is declared, which inland carrier is named on which transport document, or how a transit movement needs to be registered, is execution, and it is not mine. Derek Xu does that work and does it properly. My competence ends at describing the boundary; I do not do filing advice and I am not going to improvise any.

The arithmetic, then, and I will keep every assumption visible. Assume a Chinese exporter moving forty forty-foot boxes a month into Germany through Hamburg, so four hundred and eighty a year. Assume the inland leg from quay to Berlin or down to Halle-Leipzig costs around four hundred and fifty euro per forty-foot box, which I am carrying into the sum at roughly four hundred and ninety dollars — that conversion is my assumption, substitute your own rate and the method still holds. Your annual spend on the German inland leg is then about two hundred and thirty-five thousand dollars. Assume seventy percent of that runs through a single hinterland provider, as most mid-sized exporters' arrangements do. That is around one hundred and sixty-five thousand dollars a year riding on one commercial relationship in the segment this decision was about.

Now price the optionality, which is the only part of this that is a decision rather than a fact. Take thirty percent of the volume, a hundred and forty-four boxes, and place it with a second operator at a five percent rate premium as the cost of keeping the second relationship warm and operationally real. That is three thousand five hundred and twenty-eight dollars a year. Set it against being forced to re-source the same thirty percent in a hurry, which in practice means spot, and assume spot at twenty-five percent above your contract rate: seventeen thousand six hundred and forty dollars. The gap is about fourteen thousand dollars. Those are my assumptions producing my number, not a market quotation, and every one of them is yours to argue with. The point of doing it is that the premium for keeping a second option open is small and the exposure is not.

Notice what has not appeared anywhere in that calculation: a single euro of tariff or freight increase caused by this decision. Nothing in the prohibition changes the availability of inland capacity in Hamburg tomorrow. If a broker's quote to you moves next week, the cause is somewhere else, and conflating the two is how companies end up paying a premium they were told was inevitable.

Here is the counterfactual that would settle it, and it is worth looking out for because it is coming. If a roughly comparable European hinterland operator — same size, same inland corridors, no defence-adjacent customer — is cleared for acquisition by Chinese capital within the next year or two, then my reading collapses and the decisive variable was security adjacency alone. If such a target is refused as well, then the operative test is about control of the inland leg generally and my reading was too narrow. Until one of those cases arrives nobody, including me, has enough evidence to generalise, and anyone doing so is writing about their priors.

So the conditions under which none of the above holds. It does not hold if the reasoning, once published, turns on factors nobody has reported — a specific dependency finding, a specific customer relationship, an issue with the acquirer rather than the target. It does not hold if this turns out to be a negotiated withdrawal dressed as a prohibition, which would put it in a different category altogether. And it does not hold as guidance: this is one decision under one national regime, and other member states run their own screening systems with their own thresholds and their own tolerance. Do not stretch it into a European rule, and do not build a strategy on the assumption that it is one.

There is a timing element buried in the sequence that deserves attention, because it converts directly into a planning assumption. The cartel office concluded in February; this was resolved in October. Two parallel tracks, yes, but it puts eight months between one cleared point and the other, and it puts the screening decision last. If you are modelling a European acquisition that touches anything in this area, do not model it as one review with one date. Assume merger control and investment screening as two sequential gates, assume the second is slower and less predictable, and back-solve your signing date from the worse of them rather than the better. Companies lose transactions on calendars long before they lose them on substance.

On retroactivity, which is the question I get asked first and the one the headlines invite: nothing in this decision disturbs completed transactions. The Tollerort holding is not mentioned as impaired, no order applies to assets already acquired elsewhere, and a prohibition operates forward on a pending acquisition, not backward on a closed one. That is the general architecture of screening. It is also why the reported figures, a twenty-four point nine nine percent stake still standing next to an eighty percent refusal, want to be read as two separate decisions rather than as two points on one sliding scale.

One more boundary, because it decides whether any of this reaches you at all. The Chinese exporters with real exposure here are the ones moving past a European import agent towards their own local distribution: taking a stake in a European inland haulier, buying into a bonded warehouse operator that runs its own fleet, setting up a joint venture in which they take governance rights over how boxes reach the door. The exporter who books through Hamburg, sells DDP and pays a European haulier a rate carries no screening exposure whatsoever, and should not read this as a caution against their existing model. The variable is control of a European inland asset, not the nationality of whoever happens to be shipping through Hamburg.

What I would put in writing with reasonable confidence is narrow. Zippel's trucks still run. Nothing in this decision makes any existing contract harder to perform, and nothing in it should trigger a scramble to renegotiate. What it does do is place a marker against one specific expansion route — acquiring control of European inland distribution that touches defence logistics or nationally significant industrial corridors — and that marker belongs in the file the next time such a structure comes across your desk. That is all it is. It is less than the headlines claim and more useful than they are.

↑ Back to the key points

— By Dr. Ingrid Voss