The Impact Fraud Is Having On Commodity Trading.

In December 2022, someone at Trafigura's Rotterdam office did something almost embarrassingly simple. They opened a container.

For months, the world's largest independent metals trader had been buying and reselling nickel — hundreds of shipments, over $3 billion in trading volume, all moving through a network of companies controlled by an Indian businessman named Prateek Gupta. The nickel was LME-grade, 99.8% pure, the kind of material that trades on the London Metal Exchange without anyone blinking. Every shipment came with a bill of lading. Every bill of lading said nickel. Nobody had needed to look inside, because looking inside a shipping container mid-ocean is expensive, slow, and — this is the important part — unnecessary, so long as you trust the paper.

Someone finally looked. Inside: stainless steel. Iron briquettes. In some containers, material with no commercial value at all. By the time the High Court in London ruled on it in January 2026, Trafigura had been awarded roughly $500 million in relief — a figure the court increased to around $700 million weeks later. One hundred and seven trades, out of 542 total nickel transactions worth $3.3 billion, all resting on the same quiet assumption: that a piece of paper saying "nickel" meant there was nickel.

Here's the part of the story that gets missed, because it doesn't show up in the Trafigura judgment at all. Rewind eight years, to 2014, and a different Swiss-based trader — Mercuria — hit an almost identical wall in the Chinese port of Qingdao. Rewind six years past that same wall, and Mercuria is the company that later got hit with painted rocks in Turkey. Same company. Two entirely different frauds. A decade apart. This isn't a story about one bad actor or one unlucky trader. It's a story about a structural weakness the industry has known about for at least ten years and still hasn't closed.

Warehouse Receipt Fraud: The Warehouse That Wasn't There

In late May 2014, Mercuria discovered that a substantial quantity of aluminium and copper it believed it owned — stored in warehouses in Qingdao and nearby Penglai, worth roughly $270 million — might not exist at all, or might have been pledged as collateral more than once.

The mechanism was almost bureaucratically simple. A Chinese trading company called Decheng Mining had been taking the same physical metal, sitting in the same warehouse, and using it to secure multiple separate loans from multiple separate foreign banks — each believing its warehouse receipt represented exclusive claim to real material. Mercuria had structured its own financing as a "repo": selling the metal to Citibank with an agreement that Citi would sell equivalent metal back later. When Citi tried to deliver on that agreement, it could only hand over warehouse receipts — pieces of paper — because the warehouse operators never formally confirmed, in the language the law requires, that they were holding anything on Citi's behalf at all. The metal, the case makes clear, may never have been there to confirm.

What matters more than how the court eventually ruled is the mechanism it exposed: a warehouse receipt, like a bill of lading, is a promise about physical reality that nobody along the chain had actually re-verified against the reality itself. Multiply the same collateral across enough banks, and for a while, everyone is technically solvent — right up until someone asks to see the metal.

Six years later, in early 2020, Singapore watched the same failure mode detonate twice in the space of a few weeks — in two different companies, using two different tools, but the same underlying trick.

Agritrade International, a commodities trader dealing in coal and palm oil, collapsed in February 2020 owing more than $1.5 billion to creditors, including nearly $1 billion to secured lenders. Its former chief financial officer was later accused of deceiving a Chinese bank into disbursing $19.9 million against nineteen bills of lading that were, according to the charge, never validly pledged at all — paper claiming to represent cargo the bank never actually had rights to. She was eventually sentenced to twenty years. The company's founder fled Singapore just before police investigations began, and spent more than five years as a fugitive with an Interpol red notice attached to his name — until December 2025, when he was deported back from China and finally charged, over a separate scheme that had used fictitious invoices to draw $8 million from a digital lending platform back in 2019.

Weeks after Agritrade's collapse, Hin Leong — one of Asia's largest independent oil traders, run by a Singaporean businessman known across the trade as O.K. Lim — revealed it too had been concealing losses, this time using two fabricated oil contracts to secure $111.7 million in financing from HSBC. Lim was sentenced to seventeen and a half years in November 2024, later reduced to thirteen and a half on appeal.

And barely three months after that, ZenRock Commodities Trading collapsed after HSBC flagged what it called "extremely suspicious" financing practices — a scheme built on pledging the same cargo to more than one lender at once, the same underlying move Mercuria had run into in Qingdao six years earlier, just faster and lighter.

Three Singapore commodity traders. Three different fabrications — a fake bill of lading, a fake contract, a doubly-pledged cargo. All collapsing within roughly the same three-month window in 2020. When one company fails on fabricated paper, that's a scandal. When three fail the same way in one season, that's not bad luck. That's a market that had built its entire financing architecture on documents nobody was independently re-checking.

Three Fraud Economies, Not One

It would be tidy if all of this reduced to a single story — fake cargo, forged paper, end of lesson. It doesn't. The current cycle of litigation actually reveals three distinct fraud economies operating in parallel across the same industry, each exploiting a completely different weak point.

The first is asset and document fraud — Trafigura, Mercuria twice over, Agritrade, Hin Leong, ZenRock: sell the paper, not the product, or sell the same product's paper twice.

The second is corruption, and it doesn't touch a single shipping document. In November 2022, a UK court sentenced Glencore's UK trading arm to pay £280 million — at the time the largest penalty ever handed down in a UK corporate bribery conviction — after the company pleaded guilty to seven counts under the Bribery Act for paying roughly $29 million to officials across five African countries to secure favorable oil contracts. The sentencing judge didn't describe it as a few bad traders. He described a culture in which bribery was accepted as part of one desk's ordinary way of doing business, conduct that was endemic among the traders on it. Bribery, he said, spreads like a disease.

That wasn't isolated. Since 2017, the U.S. Department of Justice has worked through a parallel sweep of the same industry that reads like a roll call of its biggest names. Sargeant Marine pleaded guilty in 2020 to bribing officials in Brazil, Venezuela, and Ecuador. Gunvor pleaded guilty in March 2024 and paid over $661 million over bribery tied to Ecuador. Weeks later, Trafigura itself pleaded guilty in a separate Brazilian bribery matter, paying over $126 million after admitting its entities earned roughly $51 million in profit from a scheme run through offshore accounts and coded language in emails. Freepoint Commodities settled its own FCPA case the same week it settled with a different regulator entirely, for a related but distinct wrong. Add Vitol and Glencore's own US resolution, and the DOJ's six corporate resolutions and twenty individual convictions come to more than $1.7 billion.

The third fraud economy doesn't require a fake cargo or a bribed official at all — just information nobody else was supposed to have. Freepoint's separate $91.5 million CFTC settlement covered traders using bribed-for intelligence about a South American state oil company's shipping plans and its competitors' bids to structure their own trades ahead of the market. And in June 2024, the CFTC fined Trafigura — the same company that would go on to win the nickel judgment eighteen months later — $55 million for trading gasoline while knowingly holding misappropriated confidential information, manipulating a U.S. Gulf Coast fuel-oil benchmark to benefit its own derivatives book, and — a first for the regulator — writing employee non-disclosure agreements broad enough to discourage staff from reporting misconduct to the government at all. That third finding is the one worth sitting with longest. A company can build every other control correctly and still fail here, because this fraud doesn't happen on a dock. It happens in an employment contract, written by lawyers, designed to make sure nobody ever picks up the phone.

Same industry. Same decade. Three completely different playbooks, sometimes run by the same company in the same year. A due diligence program built only to catch the first kind — the fake cargo, the forged bill of lading — would have caught Gupta, Agritrade, and Hin Leong, and missed Gunvor entirely, and missed the non-disclosure clause that only surfaced because someone broke it anyway.

The Gap That Makes All of This Possible

None of this happens in a vacuum, and the number that explains why is bigger than any single fraud on this list.

Roughly 80% to 90% of world trade depends on some form of trade finance — letters of credit, warehouse financing, receivables lending, the machinery that lets a cargo move before the buyer has fully paid for it. Every two years, the Asian Development Bank surveys the world's trade finance providers and measures the gap between what businesses ask to borrow against their trade flows and what actually gets approved. In its most recent survey, covering 2025, that gap held steady at $2.5 trillion — roughly 10% of all global trade, and more than 60% wider than it was a decade earlier. Small and mid-sized businesses are rejected for financing at rates approaching 41%.

What a financing gap that size does to a market isn't complicated, even if it sounds abstract. It doesn't create fraud by itself — most firms operating under financing constraints are entirely legitimate. What scarcity does is change the cost of skepticism. When a cargo, an allocation, or a financing line might disappear if you take an extra 48 hours to verify a document, verification starts to feel like the thing you can't afford — rather than the thing you can't afford to skip. Every fraud on this list depended on someone, somewhere, deciding the deal was too time-sensitive to double-check. A market with a documented $2.5 trillion shortfall is a market that manufactures urgency constantly, for entirely legitimate reasons — which is precisely what makes it such fertile ground for the illegitimate ones hiding inside it.

The Regulators Finally Caught Up

Something changed in the years after these cases broke, and it's worth naming plainly: the law stopped treating "we followed our process" as a sufficient defense.

On the first of September, 2025, a new UK offense came into force — failure to prevent fraud, created under Section 199 of the Economic Crime and Corporate Transparency Act. It applies to any large organization whose associated person commits fraud intending to benefit the company, and the company can't show it had reasonable procedures in place to stop them. The organization doesn't even need to have profited. It's enough that someone tried to make it profit on the company's behalf.

Companies House moved on a parallel track. From November 18, 2025, every UK company director and person with significant control has to verify their identity — not just register a name, but prove it with real identification. It closes a door fraud in this sector has walked through for years: the shell company, the newly formed entity suddenly transacting at a scale wildly out of proportion to its staff or history. A dealer license or a certificate of incorporation has always felt like proof of legitimacy. It never was. It was proof a form was filed correctly — a different thing entirely.

The law has been moving on the documents themselves, too. The UK's Electronic Trade Documents Act, in force since 2023, requires that an electronic trade document exist in a system where only one party can ever exercise control of it at a time, where control is genuinely extinguished the moment it's transferred, and where the record can't be quietly copied the way a PDF of a bill of lading always could be. That's a direct legislative answer to Qingdao, Agritrade, and ZenRock alike — all three depended, at bottom, on the same underlying document existing, or appearing to exist, in more than one place at once.

Singapore didn't wait for a statute — it built infrastructure. The Association of Banks in Singapore launched its Trade Finance Registry in June 2023, built specifically to let lenders check whether the same cargo was already financed somewhere else. By early 2025, the registry's original Duplicate Financing Check had processed more than 18,000 queries across roughly forty participating banks.

Then, on February 12, 2025, the registry went further. Partnering with the Singapore Trade Data Exchange, it launched a real-time Bill of Lading Genuineness Check — and the way it actually works is the interesting part. Instead of a bank employee calling or emailing a shipping line to ask "is this real," the registry routes the request automatically: SGTraDex acts as the data conduit, sending the query to carrier data aggregators like Portcast, which pull directly from ocean carriers — Ocean Network Express was the first to plug in — and return the answer to the bank in something close to real time. Around 70% of the registry's member banks signed on to use it.

It's a small piece of plumbing, built for the exact failure that sank Agritrade: a bill of lading, presented as valid, that nobody on the financing side could independently confirm because verifying it meant trusting a phone call. This replaces the phone call with a direct line to the one party in the chain who actually knows whether the document is real — the carrier itself. It doesn't ask any one bank to be more careful. It makes the lie physically harder to tell.

What Commodity Fraud Means for a Trading Desk

None of this is a call to distrust every counterparty, which would be both paranoid and commercially impossible — physical trading has always run on relationships, and it always will. The lesson from Mercuria, twice over, from three Singapore traders collapsing in the same season, from Trafigura's two entirely separate half-billion-dollar and fifty-five-million-dollar failures in the same decade, isn't "trust no one." It's narrower than that, and more useful: trust the relationship, verify the transaction, every single time, regardless of how many times the relationship has proven itself before.

That means treating a bill of lading or a warehouse receipt as a claim to be checked, not a fact to be filed. It means asking whether the same cargo or collateral could be financed anywhere else — the exact question no single bank could answer about Qingdao, Agritrade, or ZenRock until it was too late. It means recognizing that a valid corporate registration, even with identity verification behind it now, tells you a filing was made correctly, not that the entity behind it is trading in good faith. And it means understanding that fraud in this market doesn't have one shape. Sometimes it's a container of the wrong metal. Sometimes it's the same real metal, pledged twice. Sometimes it's a bribe paid to win the business in the first place, or a contract clause quietly written so nobody ever reports it.

The market's next real advantage won't belong to whoever moves fastest. Mercuria moved fast in 2014 and lost a warehouse full of collateral that may never have existed; it moved fast again in 2020 and lost $36 million to painted rocks. Trafigura moved fast for years and lost half a billion dollars from one direction and fifty-five million from a completely unrelated one, in the same decade. The advantage will belong to whoever can prove, before the money moves, that the cargo, the collateral, the counterparty, and the people who won the business are all still describing the same real transaction — not four separate stories that happen, on paper, to agree.

📁 Desk Toolkit: The Physical Commodity Starter SuiteBuild documentary, collateral, and counterparty verification into every transaction from the start. Our toolkit includes duplicate-financing red flags, counterparty due diligence checklists, and the complete 25-Point POP Audit Checklist. Instant Access • Professional Word/PDF Templates

🎓 Accredited Desk Training: Compliance & Financial CrimeUnderstand the regulatory landscape now governing physical trade — from UK corporate fraud liability to anti-corruption compliance. ICC Academy Incoterms® 2020 Certificate · ACAMS CTMA (Trade-Based Financial Crime) [Explore Accredited Programs on the Knowledge Hub ↗]

The Commodity Room is an independent media, educational, and document risk assessment platform. Content published on this site does not constitute formal legal counsel, investment advice, trade financing, banking guarantees, or physical marine inspection services. All parties must execute independent due diligence through licensed legal counsel and certified inspection agencies.

Primary Citations & Reference Records

• Trafigura Pte Ltd & Anor v. Prateek Gupta & Ors [2026] EWHC 159 (Comm), judgment January 30, 2026; damages later increased to approximately $700 million.

• Mercuria Energy Trading Pte Ltd & Anor v. Citibank NA & Anor [2015] EWHC 1481 (Comm), judgment May 22, 2015, arising from the Qingdao/Penglai warehouse metal fraud involving Decheng Mining.

• Singapore Police Force: charges against Lim Beng Kim ("Lulu Lim"), former CFO of Agritrade International Pte Ltd, relating to $19.9 million disbursed against invalid bills of lading (originally charged 2021); charges against founder Ng Say Pek, arrested and charged December 2, 2025 following over five years as a fugitive, relating to $8 million obtained via fictitious finance invoices in 2019.

• Singapore State Courts / Singapore Police Force: sentencing of Lim Oon Kuin, founder of Hin Leong Trading (Pte) Ltd, November 18, 2024, for a $111.7 million HSBC financing fraud; sentence reduced on appeal, March 4, 2026.

• Banque de Commerce et de Placements SA, DIFC Branch and another v. China Aviation Oil (Singapore) Corp Ltd and others [2024] SGHC 145; Crédit Agricole Corporate & Investment Bank v. PPT Energy Trading Co Ltd, Singapore Court of Appeal — arising from the 2020 collapse of ZenRock Commodities Trading Pte Ltd.

• Serious Fraud Office v. Glencore Energy UK Ltd, Sentencing Remarks of Mr Justice Fraser, Southwark Crown Court, November 3, 2022 (£280 million penalty on seven Bribery Act counts).

• U.S. Department of Justice, Office of Public Affairs: "Justice Department's Investigation into International Commodities Trading Companies' Foreign Bribery Schemes Results in Six Corporate Resolutions and 20 Individuals Convicted," reporting more than $1.7 billion across Sargeant Marine, Vitol, Glencore, Freepoint, Gunvor, and Trafigura.

• U.S. Commodity Futures Trading Commission, Release Number 8834-23 (Freepoint, December 14, 2023) and Release Number 8921-24 (Trafigura, June 17, 2024).

• Asian Development Bank: Global Trade Finance Gap Survey, 9th edition, reporting a $2.5 trillion global trade finance gap for 2025, released January 15, 2026.

• UK Electronic Trade Documents Act 2023.

• UK Home Office / Economic Crime and Corporate Transparency Act 2023, Section 199 ("failure to prevent fraud" offense), in force September 1, 2025.

• UK Companies House: mandatory identity verification for directors and persons with significant control, in force from November 18, 2025.

• Association of Banks in Singapore / Singapore Trade Data Exchange: Trade Finance Registry (launched June 2023) and Bill of Lading Genuineness Check (launched February 12, 2025).