For most of the modern era a cross-border commodity transaction needed three institutions. A trader took title and bridged the price. A consultant advised. A bank financed the flow. All three still do their work well. What changed is the transaction itself. Three forces have redrawn what a deal must contain before it can exist, and none of them is cyclical. The obvious conclusion is that a fourth party is needed to cover the space between the three. I think that conclusion is too small. The space is not a gap waiting for another intermediary. It is an architecture that has to be built, and whoever builds it changes who the trader is, where the margin settles and how far a producer can reach. The model I have in mind descends less from the Western trading desk than from the Japanese trading house, adapted to a market the sogo shosha never faced.
The forces were visible before they were named
In May 2020, at the bottom of the first COVID shock, Aluminium International Today published my essay The dark forest of COVID-19: Survival of the fittest business. I named four trends emerging from the crisis. National protectionism was growing, and the pandemic had exposed how weak the existing alliances and associations were at protecting their members. Competition was tightening under falling demand, pushing producers to look for alternative suppliers of raw materials and alternative markets. Industrial dependence on financial institutions and government support was deepening, and the struggle for access to institutional, investment and public finance was about to get tougher. And the metabolism of business itself was accelerating, with forced digitalisation stripping out unnecessary operating links.
The essay framed survival in wildlife terms. A company would need sharper sensory organs, meaning live operational control over geolocation, custody and product data across the entire chain, certified by an independent third party. It would need secure access to sources of food, including capital raised outside the conventional lending channel. And it would need socialisation in new market prides. I was not forecasting disruption. I was describing a system re-forming around three pressures that have since hardened into the conditions every cross-border flow now meets. Regulation that defines the transaction. Fragmentation that prices its origin. Finance that will only touch what it controls.
I. Regulation stopped following the transaction and started defining it
Compliance used to come after commerce. A deal was negotiated on price, grade and delivery. Documentation followed. A customs broker cleared it. For any flow touching the European Union that sequence has been inverted, and the inversion is written into the instruments themselves.
Take the Carbon Border Adjustment Mechanism, Regulation (EU) 2023/956, in its definitive phase since 1 January 2026. It does not tax a transaction after the fact. It prices the embedded emissions of cement, iron and steel, aluminium, fertilisers, hydrogen and electricity at the border, against emissions data that must exist, verified, before the goods move. An importer without authorised-declarant status does not pay more. It does not import. Add the certificate-surrender calendar, the quarterly holding rule from 2027 and the stepwise withdrawal of free allocation under the EU ETS through 2034, and the cost curve of a physical flow becomes a regulatory schedule as much as a freight rate.
The same inversion runs through the rest of the current wave. The EU Deforestation Regulation, Regulation (EU) 2023/1115, applying to large and medium operators from 30 December 2026, makes geolocation-backed due diligence a condition of placing goods on the market. The Corporate Sustainability Due Diligence Directive, Directive (EU) 2024/1760 as amended by the Omnibus I Directive (EU) 2026/470, applies from 26 July 2029 and extends liability along the chain of activities. RED III, Directive (EU) 2023/2413, decides through its greenhouse-gas thresholds and Union Database entries whether a biofuel cargo can be sold as sustainable at all. The Critical Raw Materials Act, Regulation (EU) 2024/1252, sorts projects and supply relationships into strategic and non-strategic, with direct consequences for permitting and finance.
The practical meaning is severe. The regulatory design of a transaction now comes before its commercial design. A deal negotiated first and documented later is built in the wrong order.
II. Markets fragmented along origin
The second force is geopolitical. Ten years ago origin was a logistics detail. Today it is a pricing category and increasingly a legal status. Five sanctions regimes operate simultaneously across the metals trade - EU, US, UK, UN and Switzerland's SECO - and they do not perfectly overlap. The same tonne of metal can be deliverable in one jurisdiction, warehousable in a second and untouchable in a third. Exchange rules have followed. Origin now determines what may be warranted, financed and delivered against contract on the world's principal metals venues.
Fragmentation reprices the chain itself. Premia and discounts attach to provenance. Buyers demand proof of custody from mine to port, not as a sustainability preference but as sanctions defence. Who produced this, where has it been, who has held it. Those questions used to be answered with an invoice chain assembled after loading. They no longer can be. Chain of custody became a commercial category, and the documentary spine of a transaction became part of the price.
For a producer outside the large trading blocs this is the decisive fact of market access. The material may be excellent and the counterparty willing. If the origin story cannot be evidenced to the standard of the importing regime and its banks, the transaction does not clear.
III. Bankability became the threshold, not the afterthought
The third force runs through finance. After the commodity-finance losses of 2020 and 2021 the European banks re-underwrote the sector. Fewer names, harder collateral standards, structures over signatures. Uncommitted transactional lines gave way to borrowing bases, collateral management agreements and controlled payment waterfalls. The question a credit committee asks is no longer who the borrower is. It is what exactly the bank controls, at which moment, evidenced by which document.
That moves the financing question to the front of the deal. Title transfer points, document flows, custody arrangements, insurance assignments, settlement mechanics - all of it has to be engineered before the term sheet, because it is the term sheet. A transaction assembled commercially and brought to a bank afterwards will be restructured or declined. Either way months are lost. The transactions that close are the ones designed to be financed from the first draft.
Unbundling the trader
Set the three forces against the classical division of labour and it is the trader's position that no longer holds together. The trader took title, carried the price risk and earned the spread. In exchange the producer surrendered the customer relationship, the margin and control over what happened to the material next. That bargain made sense when the trader's book was the only bridge across price, geography and credit. It no longer is. The regulatory shape of a flow, its origin evidence, its financing structure - the real work of making a modern transaction executable - cannot be solved from a trading book. They need an architecture.
Unbundle the trader and its function splits in two. One part is genuine risk-taking: warehousing tonnes, running a position, absorbing basis. That part stays, and where a book is truly needed the trader remains a counterparty. The other part is everything the trader did around the position. Connecting a producer to a distant buyer. Assembling the finance. Standing behind the paper. None of that requires taking the producer's title, the producer's customer or the producer's margin. It requires building the structure through which the producer keeps all three and reaches the buyer directly.
This is the pivot. The model that answers the three forces does not fill the space between trader, consultant and bank. It takes over the trader's architectural function without the trader's position, and it hands the producer what the trader used to take. The end customer. The margin. The capital. And capital that stays with the producer sits where development actually happens - in the mine that expands, the plant that modernises, the balance sheet that carries the next cycle.
The Japanese reading
There is a precedent for this, and it is not Western. The Japanese trading houses were never built around a proprietary book the way a Glencore or a Trafigura is. They were built to organise trade for the industries behind them. I pointed to this in the same May 2020 essay, in the section on socialisation in new market prides. Nine postwar houses - Mitsubishi Shoji, Mitsui Busan, Itochu Shoji, Marubeni, Sumitomo Shoji, Nissho-Iwai, Toyo Menka, Nichimen and Kanematsu-Gosho - came to carry more than 60% of Japan's foreign trade, with some twelve thousand companies working alongside them. Their first task was to put national capital in control of the country's foreign trade and use it to modernise the economy. Over a few decades the share of Japanese firms in that trade rose from 1% to 80%. And the houses went far beyond sales. They organised the supply chains, built distribution warehouses in the consumption centres abroad, financed the flows, and kept close ties to the Ministry of Trade and Industry, the Ministry of Finance and the Export-Import Bank. A trading house was a market buffer between producer and consumer, and infrastructure for its industrial partners rather than a rent extracted from them.
The essay closed with a prediction. New trading houses would be created and existing ones strengthened, they would work in conjunction with financial institutions, and they would gravitate toward jurisdictions offering stable financial and tax systems. Market centres such as Switzerland, I wrote, were "likely to receive an additional impetus for development". A year and a half earlier I had already placed that bet myself.
I did not know the houses only from print. Over the years I worked with them at principal level - Sojitz, Toyota Tsusho, Mitsui, Marubeni, Chori - and the operating logic was always the same. The trading company earns by making its counterparties stronger, not by standing between them and their customers. There is even a direct line inside that list. Two of the nine postwar houses, Nissho-Iwai and Nichimen, later merged into Sojitz.
MSE is an interpretation of that model, not a copy. The sogo shosha worked in a world of open trade, cheap capital and light regulation, and it carried the position itself. Today's world is the opposite on every count. Regulation defines the flow. Origin is a legal status. Banks finance only what they control. In this world the position has turned from an asset into a liability, because a party that holds title also holds a conflict, and both a bank's counsel and a sanctions officer can see it. The adaptation is to keep the trading house's architectural role - market access, structuring, finance, logistics, information - and drop the one thing the modern market punishes. What remains is the sogo shosha's function delivered by a party that takes no title and runs no book, and is therefore acceptable to producer, buyer, regulator and bank at once.
Why a commission agent - and why Swiss
That party has a legal form, and Swiss law has held it for over a century. The commission agent - the Kommissionaer of Articles 425-438 of the Code of Obligations - contracts in its own name but for the principal's account. Title does not pass to it. The goods and the customer remain the principal's. Its compensation is a mandate fee fixed in advance, not a spread whose size the principal never sees. And it owes the principal the duties of an agent under mandate law: loyalty, care, account.
The form fits the role exactly. It gives the architect standing to act - to contract, to instruct logistics, to present documents, to face the bank - while the statute itself guarantees that title, customer and margin never leave the principal. It is the trading house's function with the position stripped out, written into law. A bank's counsel can verify the allocation of interest in an afternoon. A sanctions officer can see that no proprietary position is being taken. A producer can see that the relationship and the economics stay on its side of the table. None of this is a marketing position. The allocation of interest is statutory and enforceable in a Swiss court.
Seated in that chair, one firm can build what the classical triad cannot build together: the regulatory shape of the transaction, its origin evidence, its financing architecture, and then the execution itself - contracts, documents, logistics, settlement - while every relationship and every basis point stays with the principal. The trader remains a counterparty where a book is genuinely needed. The consultant remains a source of analysis. The bank remains the lender, and it receives a transaction already built to its collateral standard. The model does not try to beat the triad at its own game. It puts the producer at the centre of the system, where the intermediary used to stand.
The system, not the trade
None of this stops at a single producer-to-buyer flow. The same architecture runs through the whole field MSE works in, because the principle is the same everywhere: build the structure, and let value settle with the party that creates it. In mining and resource investment it is the difference between a headline grade and a financeable asset - the diligence, the offtake and the vehicle that turn a deposit into capital a board can commit. In M&A and special situations it is the structure through which an asset changes hands without leaking its value on the way. In terminals and logistics it is the move from treating custody as a cost line to running it as a controlled, surveyed, insurable instrument - the point where a warehouse receipt becomes bankable collateral and a supply chain becomes transparent enough to finance.
I made this argument in a different register in Sol et Luna. The commodity value chain is one system. Sector, logistics and trading infrastructure only produce a complete picture when they are held simultaneously, and value that stays invisible while they are treated separately becomes visible, and capturable, once they are built as one. Seen this way, the execution gap was never a hole in the market. It was the market telling us that the old division of labour had stopped serving the producer, and that the architecture which once belonged to the trading house had to be rebuilt. Without the position. On the producer's side of the ledger.
What this means in practice
I incorporated Metal Supply Experts GmbH in Zug in December 2018 on this reading of the market. CBAM did not yet have a regulation number. The four trends of 2020 were still a bet that regulatory mass, fragmentation and financing discipline would converge and force the trader's function to come apart. They converged, and it came apart.
The method follows from the thesis. Every mandate starts with the regulatory calendar and the financing architecture, ahead of the commercial terms. Every flow is built with its evidence spine - origin, custody, emissions, sustainability certification - as load-bearing structure. And every engagement is priced as a mandate, because the moment an adviser's income depends on a spread, the adviser has become the trader this model exists to replace.
The forces are not relenting. The CBAM certificate calendar runs to 2034. EUDR's operator obligations begin this December. CSDDD applies from 26 July 2029. Each deadline adds documentary mass to the front of the transaction and widens the distance between a deal as negotiated and a deal as executable. The classical institutions will keep doing what they do well. The architecture that holds a modern flow together, and keeps its value with the producer who created it, is where MSE stands.