Head of International Equities, Madison Investments
Tom Tibbles and his team have focused for decades on implementing a well-tested and successful investment strategy to invest portfolios of international stocks. Over the last few years, he has led the team to embrace the Wolfram technology stack to make the process explicit in software and to enhance, accelerate and improve the quality and consistency of the workflow.
Financial data can be sliced cross-sectionally, through time or simultaneously by both curating and provisioning processed data in multidimensional matrix structures—“DataCubes.” Doing so has made it highly efficient to execute the desired types of data manipulations and visualizations in Mathematica.
The project pipeline began by writing custom APIs to extract data locked in silos; legacy procedures were then translated and separated into hundreds of “CustomMetrics” to clean and increase the information content of individual data segments. After the release of Mathematica 12, the project expanded to take advantage of the entity store data framework.
Additional projects have focused, within a Wolfram Language package, on automating the integration and enhancement of data and sequencing the workflow steps across multiple internal and external data sources and applications. Lastly, user experience was vastly improved with the custom development of a GUI to access, examine further and manipulate data while dynamically displaying the visual reports.