Please tell me a little bit about your career and Kpler.
After completing my education in India, I moved to Singapore in 2011 to join Olam; it was my first job.
I moved to Dubai in 2017, where I ran my research firm, working with grain trading desks worldwide. I joined Kpler in Dubai at the end of 2023, looking after agricultural commodities. One of my responsibilities is to add a layer of opinion-based research to Kpler’s suite of data products. I work closely with Kpler teams to enhance our coverage across agriculture, grains & oilseeds, veg oils and biofuels.
Kpler was founded in Belgium in 2014. The two co-founders, François Cazor and Jean Maynier, who still control the company today, initially identified a gap in the LNG market.
From there, they expanded to oil and tankers and later dry bulks. In 2023, Kpler acquired the leading ship tracking and maritime analytics providers, MarineTraffic and FleetMon, becoming the number one in vessel tracking with more than 1 billion AIS pings daily.
Kpler is the go-to source for global trade intelligence and the best cargo-tracking platform on the market. Kpler can use its data to aggregate shipments and do all sorts of analyses, including weekly, daily, and monthly volumes between countries by commodities, diversions, commodities on water, ton-miles changes and so on.
Knowing the quantities of the commodities on the water right now gives us significant insights into the market. We call it Kpler Insight. We add to the data product by providing expert analysis and opinion, including news, research reports, price analysis and ad hoc research.
As an analyst, I found that we could track the vessels, but finding out what was on board the ship wasn’t always possible. It was easy from the main ports but challenging for the smaller ones and coastal traffic. How do you get around that problem?
It is challenging in dry bulk, but we can track all vessels over 5,000 t. We use over 250 hard sources, including line-ups, port reports, arrivals and fixtures.
Our process uses rigorous checks for data quality. We have a robust data quality team whose job is to ensure we get the cargo and quantity right. Rigorous back-testing proves our high accuracy in tracking maritime cargo.
Kpler has perfected this process in gas and oil, and we are now applying it to dry shipments. The team has significantly scaled our coverage across ags, major and minor bulks, and commodities such as steel and fertiliser.
It comes with challenges. Recently, we have seen many vessels reporting incorrect AIS positions. Some people don’t want the world to know what they’re up to, particularly in the Black Sea.
They’re turning off their tracking, aren’t they?
For a long time, many vessels showed their position at Moscow’s airport. Now, they’re showing it in more believable places, but a third of all Russian coal exports are loaded with the transponder switched off.
It is similar for wheat and barley in the Black Sea ports and corn out of Caspian ports going to Iran, especially on the southern side, closer to Iran.
Kpler’s risk & compliance product evaluates coverage, projects vessel positions during the AIS gap and qualitatively identifies AIS coverage as low-risk or high-risk.
While AIS is the primary source, we also use other sources to bring depth and colour to our data.
How do you add value to the data Kpler provides you?
I’m an agricultural commodities analyst. My job is to help clients understand and trade agricultural markets. Everything we build and develop is towards this end.
Most companies analyse markets from the supply side, and the big companies spend millions on crop forecasting. We bring something different to the market by focusing on demand and trade.
There are numerous importing countries, and getting accurate and reliable import data is challenging. Kpler’s cargo tracking data gives us an understanding of destination markets. We use it to do bottom-up country-level demand analysis and predict monthly trade flows forward. A good grasp of future flows helps us build quality balance sheets that can be reliably used to trade.
Do you use artificial intelligence?
We have several applications in which AI is bringing efficiency to processes, like automatically assessing anomalies in our data. We have applications of machine learning for processing earth observation data for oil inventories, pattern detection for commodity trading, and forecasting future trades.
Do you make price predictions?
We discuss price direction and share our views on prices. We maintain a view. You can’t have a predictive trade analysis without a view of prices.
Which market do you find the most challenging?
Russia is the largest wheat exporter but has no established forward market. The global wheat market has become a spot market with little ability to buy down the curve. It has made the market short-term. As a result, you cannot trade an annual balance, only next month’s supply.
Supply is plentiful for this month and the next because the northern hemisphere has just harvested, and Russian and Ukrainian farmers are keen to sell. They have no incentive to carry as they have no certainty over future prices.
This phenomenon has also led to people not investing in storage, making the physical cost of carrying relatively high.
If Egypt and Algeria come in and buy, which happened a couple of weeks ago, the market will rally because they just bought ten days’ worth of production. Once those ten days pass, the exporters start discounting again to get sales. There is no long-term strategic thinking.
Can traders use the Matif to hedge forward?
It is far from a perfect hedge for Russian wheat because of convergence, delivery and quality differences.
What is happening in soybeans, and how will the US election impact the market?
The US elections and US / China trade relations are the number one factors in the market today. Renewable biodiesel would be second, but that is more for the long term.
The US crushing sector has been running at maximum capacity for several years. It is the bottleneck. They are increasing that capacity, but as far as the soybean balance sheet goes, we assume it will continue to run at maximum capacity. That way, we may not worry about the downstream renewable side, at least for the soybean balance sheet.
China is the elephant in the soybean market. Secrecy has increased, and the market has gone from somewhat transparent to completely opaque, making it harder to predict what China will do.
Using Kpler data, we realised that China will import a record quantity of soybeans in July. Monthly arrivals will be over 11 million tonnes—about 10 per cent from Argentina and the rest from Brazil. This is not because of better margins or a sudden increase in demand. It looks like China is concerned about possible trade issues.
China traditionally imports US beans in October, November, and December and then reverts to Brazilian origin by the end of January.
I believe China wants to avoid being dependent on US beans. Imports from the US will not fall to zero, but China prefers Brazilian origin to US origin. Brazilian beans will go to China, and US beans will have to fight it out to other destinations.
The US is currently growing a considerable crop of soybeans. They are in the ground, and there are no red flags regarding production. We expect the US to have relatively high stocks in 2025 and very high stocks in 2026. This will collapse the structure of the soybean market and increase carries.
Argentina is a factor in soybeans. The country’s farmers are sitting on over 10 million tonnes of stock because it’s a currency for them. It is a dollar-denominated commodity.
Could Argentina change its policy of favouring meal over soybean exports?
The President, Javier Milei, is a libertarian and, ideally, would want to remove all the taxes and export duties, but he just can’t afford to do it. Even if he does, the differential export taxes will remain. The crushing industry is significant and politically influential. It is also a major source of employment and government income.
I’d like to return to something you mentioned about the difficulty of getting reliable information.
A lot of Chinese data around domestic crops and stocks is kept private in view of their food security challenges. This information can be quite valuable to traders. Recently we have started seeing mismatches in customs data as well, which further increases the difficulty of understanding and trading Chinese demand.
I fear that the same thing is starting to happen in Russia. The government tries to maintain an informal price floor on wheat exports and penalises companies that fail to play along. Foreign companies have left, and there has been a consolidation at the export level. So far, there hasn’t been much attempt to control the information flow, but I’m afraid it could be the next step.
After the breakup of the USSR, governments largely left the agricultural commodity trade, but by what you say, they are getting back in.
They are getting involved in the markets in the name of food security. China, Russia, India, Pakistan, and, to some extent, Bangladesh all try to control agricultural markets. Meanwhile, the Egyptian and Saudi governments are the world’s largest grain buyers. Algeria is a big importer; again, it is the government.
Out of the commodities you cover, which one do you prefer?
Although wheat is the most challenging, it is the one I prefer. It has a variety of end-user applications but doesn’t generally go to animal feed or industrial use. It ends up on people’s tables. When I look at pasta, bread, or biscuits, I try to guess the quality and origin of the wheat. It brings tangible meaning to what I do.
© Commodity Conversations® 2024
My new book, Commodity Professionals – The People Behind the Trade, is now available on Amazon.
