Proofs and Cross-Validations: Three Lessons for Financial Data Science

Abstract

Like any new research program, financial data science must successfully demonstrate its utility to researchers who are accustomed to working with more established analytical frameworks and tools. This is especially important in the early stages of financial data science, where much of the methodological groundwork of the field will be laid. Given this, in this article we draw on the history of mathematics, an exemplar of a successful scientific endeavor, to provide three lessons for researchers in financial data science that we hope will assist them in aligning their research priorities more closely with those of mainstream finance. We close the article with some additional guidance on the related topic of effectively writing and presenting financial data science research.

TOPICS: Statistical methods, quantitative methods, big data/machine learning

Key Findings

• • Financial Data Science must be in epistemic dialogue with traditional finance.

• • Financial Data Science must aim for epistemic transparency.

• • Financial Data Science must aim for epistemic connectivity.