There was a piece in the Financial Times last week entitled “New crop data providers cash in on US shutdown” (paywall). It detailed how the US government shutdown impacted many areas of businesses from the National Transportation Safety Administration to the USDA. The USDA crop reports were interrupted for an extended period of time, leaving many agricultural traders in the dark on government reported stocks of soybeans, corn, wheat. Third party alternative data providers have stepped in to fill the gap, using their own process of data collection and predictive analysis to produce their own set of balances.
The aggregation of data in all trading strategies has exploded. In commodity trading this seems to be lagging that of equities. The idea of scraping Twitter for sentiment analysis on a particular stock to predict price movements or counting empty car parking spaces at Home Depot to predict future earnings shortfalls, has been around for some time. Hedge funds have been known to track private airplane tail data to predict M&A activity.
What is less clear is how alternative data is being used in the commodity sectors. Services like Clipper use AIS (Automatic Identification Systems) signals to predict oil and oil products flows around the globe. Position and speed are reported along with ship draft to predict the volumes of liquid on board and how the global flows are affecting the supply-demand balances. Fundamental managers use this information to build out a robust supply picture to inform their market views. No longer can a manager simply rely on a tip from a physical player exporting gasoline from Rotterdam to formulate a view. At times the price moves may be a macro driven, other times driven by OPEC messaging, or the latest US production numbers. More likely than not it is the combination of all of these and a hundred or more other factors simultaneously that the manager must aggregate, store, clean the data and figure out the best expression of the view.
At Bridge Alternatives, we have a deep interest in alternative data usage, particularly in the commodity sectors. We look forward to hearing about new methods and data sets that create alpha generation advantages. The use of machine learning, natural language processing and artificial intelligence and its applications to commodity trading will no doubt be a topic of conversation at the year’s Time Summit.