The Mercuria team had a requirement for a specialist tool to commodity markets globally in order to assist their trading decisions. It required a large amount of data aggregation of the commodity market data with some specialist calculations and had to be highly flexible to service the changing demands of the trading desk. Using Excel and manually manipulating the data through spreadsheets was not an option, due to the sheer quantity of data involved and the complex nature of the underlying calculations. The team spent several months looking at various options including some traditional database style solutions with some batch-based calculations.
Mercuria invited CubeLogic to present their technology solutions in the Trading Office. It quickly became apparent that they had found a potential partner that could deliver the requirements. The use of a real-time cube seemed like an obvious technology fit. A cube application allows large scale real-time aggregation with unrivalled flexibility in slicing the data, something that was a key requirement. âWhen we first saw what CubeLogic could do, it was immediately clear that this could be just the solution we had been searching for. We spent two days with the CubeLogic team and they were quickly able to demonstrate an understanding of our problem and show us some tangible outputs of how it could be solvedâ, said a Dry Commodity Analyst at Mercuria.
The Mercuria team needed a first rate flexible solution, but also speed of delivery was extremely important. Report data was required by the Trading desk in a matter of weeks, not months. There followed some intense analysis of the business requirements, with Mercuria and CubeLogic working side-by-side to ensure all the bases were covered and every detail was fully understood. The most important element in getting the right end result was cube design; with the right design, a cube will deliver performance and flexibility to the end user. âWe needed to completely understand the business problem before the cube design could fully emerge. Once that was clear, we carried out some complex dimension modelling, which is the process by which an OLAP cube is initially designed. It flushes out all the possible calculations and outputs and ensures optimum performance of the end resultâ, says Lee Campbell, CTO of CubeLogic.
Another key factor in success was the team effort to achieve the goal, with Mercuria and CubeLogic working hand-in-hand as the Cube deliverables emerged. There were a large amount of differing requirements. The cube had to handle numerous assumption matrices and calculation points in order for the end user to obtain the very best view of the market. Also important was the ability to âtime-sliceâ the data back using all the historical movements. In this way correlations of commodities, seasonality and other trends can be analyzed to make informed decisions going forward.
Within one month of commencing the project the first cube was running, delivering business benefit to Mercuria. The Commodity cube consists of a robust data interface layer which allows the users to quickly and efficiently load up new market information as it becomes available. It can support many different sources of the data. All the key matrix and decision information is also maintained through the interface. Changes in assumptions or rules can be easily uploaded in minutes.
The Cube itself allows amazing real-time flexibility for the user. Reports can be generated to view commodity data using various views time-slices, regions, countries or markets. Itâs possible to view the different market data through complex hierarchies. Hierarchies are also available to break down commodity data from Regions, Countries and other categorizations. âWeâve gotten everything we wanted and total flexibility to construct numerous views and complex analysis of the underlying data. The performance is really impressiveâ, states Mercuria.
The Cube itself handles millions data points in real time. It does complex calculations on the data in real time to compute averages and summations, allocations, quantity breakdowns, unit of measure conversions, drilldowns and historical trend analysis.
Mercuria are planning a second phase for the Trade Flow Analysis Cube, building on the success of the first delivery. In parallel they are also implementing Risk Cube, another CubeLogic product, to handle the management and reporting for their Credit Risk team. Risk Cube handles all the complex netting, exposure aggregation and collateral management, again in a flexible, real time cube.