The role a sound process plays in achieving successful, long-term outcomes has been widely cited across many professional and personal circles. When I think about process improvement, my mind immediately takes me to romanticized examples, like Moneyball's depiction of Billy Beane and the Oakland A's, or one of the many well-chronicled Jeff Bezos' corporate principles (please, no more PowerPoint). However, for many of us, process improvement looks a bit more like implementing the KonMari method or finding the best morning routine to promote productivity and peace of mind (thanks, Tim Ferriss). No matter how you like to think about it, it's easy to see that more and more time is being spent refining the “right” approach to problem solving.
Financial circles are no different. Thorny problems necessitate having sound processes to break one potentially overwhelming task into a series of smaller, more manageable ones. Some of the largest problems I come up against time and time again have to deal with using new datasets to test investment ideas. The amount of data available to investors is staggering and many are often left wondering where to begin.
In a series of documents, I'll inspect a dataset that is new to me by defining a process, gathering the necessary inputs, and conducting exploratory analysis. In part one below, I use the 2iQ Research Global Insider Transaction dataset to test the influence insider transactions have on future asset returns. The second installment will take that analysis a step further by reviewing approaches to feature engineering and model evaluation.
For more information on the 2iQ Research Global Insider Transaction dataset, visit the Open:FactSet Marketplace.