Commodities research house of the year, Asia: Marex Spectron

Energy Risk Asia Awards 2020: Broker’s unique approach to data collection results in highly praised modelling and analysis

Georgi Slavov

Commodity markets have made a big leap forward in terms of the availability and quality of data in the past year or two – and traders in Asia have become more demanding about the information they pay for. The upshot is that research providers now face a greater challenge in analysing and packaging data in the most effective way.

With this in mind, Marex Spectron has revamped the way it handles, processes and extracts value from fundamental data, says Georgi Slavov, global head of fundamental research at the London-based interdealer commodity broker. The resulting analysis and models have drawn high praise from clients and won Marex Spectron Energy Risk Asia’s 2020 commodities research house of the year.

One Singapore-based coal trader, for example, referred to Marex Spectron’s “highly empirical” analysis, “well-thought out” conclusions and “unique data points” on Asia.  

The recent explosion of fundamental data on commodities has made it more difficult to handle, process and distil data into useful information, notes Slavov. Marex Spectron has developed a systematic methodology to collect and process information on a daily, hourly or sometimes even minute-by-minute basis. It combines traditional import/export figures with more esoteric statistics, such as satellite, weather and shipping data.  

“We’re heavy users of probability and multilinear regression modelling, which are the building blocks of machine learning, quantitative finance and modern macroeconomic research these days,” Slavov says. 

Marex Spectron extracts signals from the data to create intra- and inter-market signalling processes as the markets in question are sometimes indirectly connected. “Information that has been priced in by one market often has not been priced in by another,” he notes. It can therefore be used for cross-market arbitrage. 

The team writes algorithms that assess the viability, quality and directional trends of hundreds of data sources, which are then grouped into clusters that create the models. 

It is only in the past two or three years that this sort of information has become available in certain commodity markets, and it has been a game changer for analytics on crude oil, oil products, natural gas and ferrous metals, Slavov adds. 

An important development over the past 18 months has been the expansion of the team’s liquefied natural gas (LNG) research, which has happened in line with the firm’s LNG broking activities. Marex launched a financial LNG desk in Singapore in September last year, then added financial freight LNG in London and physical freight LNG in both locations earlier this year. These complemented the financial LNG desk the firm set up in London three years ago.

This gives the firm comprehensive international pricing insight to feed into its research. “That was the missing link for us to connect the global energy markets,” Slavov says.  

“It is a unique proposition for our clients to have access to research as well as pricing and execution on financial LNG, physical LNG and LNG freight markets in the same place in a sell-side business,” he adds.

Another service Marex Spectron has developed since 2018 is its China pollution model.

In China, local authorities control whether major emitters, such as power plants or steel mills, shut down or continue to operate according to pollution thresholds in their area. The location and levels of pollution will be affected above all by wind direction and strength, which are governed by atmospheric pressure, Slavov explains. 

The China model forecasts where pollution will exceed certain thresholds and how it will consequently affect commodity markets. Financial and physical traders of iron ore, steel and scrap metal, and coking coal have made use of the service, Slavov says. 

One copper concentrate trader said the China pollution model is one of the many reports put out by Marex Spectron that stands out from the crowd. “It has been a good indicator for when there will be environmental curbs enforced and how long they will last,” he says. 

Alongside these new developments, Marex Spectron has maintained in-depth analysis of Asian thermal coal markets, where some rivals have not. This area is neglected by most other research houses, said Michael Everston, a Singapore-based thermal coal trader at German energy group RWE.