Trading Commodities: Algorithms

In this post I’d like to discuss commodity market participants at NYSE Liffe, as there is often a bit of confusion around who is active in our markets and where the potential increase in volume might come from.

It’s no secret that 60 to 70 percent of our markets are occupied by the end-user or commercial market participants (e.g. those having the physical commodity to support their futures/options position). The remainder is split between hedge funds, banks, proprietary trading shops and algorithmic trading companies. “Lack of competition” makes commodity markets particularly interesting for the latter to enter. One can argue that comparatively low liquidity makes it difficult to apply algorithms to the markets, but I recently attended the TradeTech conference in Moscow and understood that attempting to analyze and pair the right markets could bring substantially favorable results.

Non-traditional market participants could be broadly split into two categories in terms of their trading strategies: position trading (banks and funds) and HFT (high-frequency trading). The former would be searching for correlations and correlation relationship, the latter would be keen to find a trend. The latter would also often trade arbitrage (either classical or statistical).

It seems that a majority of NYSE Liffe contracts have similar contracts on competing exchanges (ICE, CME, etc). For example, there is strong correlation between NYSE Liffe Cocoa futures and ICE Cocoa futures. The same applies to the correlation between NYSE Liffe Milling Wheat futures and CME wheat futures (75%) and NYSE Liffe Rapeseed futures and ICE Canola (72%). This type of arbitrage is usually referred to as dimensional (location-wise).

Apart from the above, there are less traditional ways of arbitrage trading – basis vs futures, index vs futures, and calendar arbitrage (based on various expiry months). Index vs futures arbitrage might gain the traction as NYSE Euronext launched single commodity indices in July 2011.

In any case, speaking to my Russian colleagues it appeared clear to me that with more institutional money flowing into commodities, managers are looking for new and innovative ways of seeking alpha. As such, algorithmic trading is being applied to the most highly volatile and active commodity subsets like Cocoa, Robusta Coffee or Milling Wheat.