The world's largest commodities traders are investing heavily in data processing and evaluation within the race to realize a technological edge over competitors.
Groups like Vitol and Trafigura, which traditionally relied on political connections, handshakes and logistical skills to move natural resources from distant locations to willing buyers, are increasingly focused on applying artificial intelligence to probably the most physical industries.
“In some ways it’s an arms race,” Russell Hardy, chief executive of Vitol, the world’s largest oil trader, said on the FT Commodities Global Summit in Lausanne this month. Trading houses are attempting to make use of AI in two fundamental ways, he said: to enhance business efficiency and to develop a trading advantage by having more analytical capabilities than competitors.
“I believe we’re all attempting to get to the moon first,” Hardy said. “But I’d say it contributes more to business efficiency today than cracking the code of what the market goes to do that afternoon.”
Privately held Vitol, which employs about 1,800 people, posted record net profits of $15.1 billion in 2022 and about $13 billion in 2023, making it probably the most profitable corporations on the earth on a per capita basis.
The push to make use of the newest technological tools is partly a response to competition from hedge funds and other data-driven trading teams that move fewer physical commodities but have built lucrative businesses in commodity-linked securities and other financial products.
The most advanced data-driven trading operation within the industry is arguably at Miami-headquartered hedge fund Citadel, which hired Macquarie commodities trader Sebastian Barrack in 2017 to steer a significant push into energy and commodities.
In search of an information edge, certainly one of Barack's first steps was to rent a 20-person team of weather forecasters. The broader commodities trading team has since grown to greater than 300 people, including analysts and engineers.
Oil and refined products is an area where available data on supply levels, demand patterns and logistical variables have increased sharply in recent times, Barrack told the Financial Times. “The tremendous growth in data available to us is literally making us more informed investors.”
For data-driven trading strategies, the energy transition guarantees to be a boon as it is going to increase complexity and require more sophisticated tools to model markets, especially when a brand new area lacks historical information, he added. “The more detailed, the more complex, the more there may be an absence of retrospective data, the higher for us.”
According to a study by LCH Investments, Citadel generated record revenue of $16 billion in 2022, displacing Bridgewater as probably the most successful hedge fund of all time. About half of that got here from commodities as the corporate, like other traders, benefited from extreme volatility in energy markets following Russia's invasion of Ukraine, the Financial Times reported. Citadel declined to reveal its 2023 performance.
The ability to process large amounts of information is especially vital within the rapidly growing area of ​​electricity trading, where large amounts of data are generated as a consequence of the regulated nature of electricity markets.
Advisor McKinsey Data-driven trading corporations are estimated to have generated 1 / 4 of gas and electricity trading profits globally in 2022, up from lower than 5 percent in 2021.
This competition has forced traditional commodities traders like Trafigura, which had record sales of $7.4 billion in 2023, to take a position to maintain up. An electricity trading department was founded three years ago.
Richard Holtum, head of gas, power and renewable energy at Trafigura, said his team “uploads several billion discrete bits of information to the cloud on daily basis”. “The challenge is to make use of AI to question this data higher and more efficiently, thereby improving the trading decisions we then make,” he said. “I believe we’re currently on the tip of the iceberg of what AI can do.”
Founded in 2004 by Marco Dunand and Daniel Jaeggi, Switzerland-based Mercuria was primarily an oil trader but strengthened its power trading operations in 2014 by acquiring a part of JPMorgan's physical commodities business.
Dunand told the Financial Times that the data advantage of data-driven players like Citadel helps them take larger positions out there, but that AI may help Mercuria bridge that gap.
“If you wanted to make use of Citadel for instance of collecting data, I believe it could take a number of money and time. . . That’s why we invest a number of effort and time into developing our own AI machines to one way or the other fill these gaps,” he said. Mercuria earned about $2.7 billion in 2023, he added, barely lower than the record $3 billion the previous yr.
However, physical retailers won't stop getting their hands dirty.
“You can actually be a market participant without physically trading, but that’s not for us,” Dunand said. “I believe ultimately the world needs energy and we’re energy traders. So for those who don’t move these things, you realize the world doesn’t work.”
