Big Data In Finance Research Articles

Big Data in Finance: Highlights from the Big Data in Finance Conference Hosted at the University of Michigan October 27-28, 2016

How can financial data be made more accessible and more secure, as well as more useful to regulators, market participants, and the public? As new data sets are created, opportunities emerge. Vast quantities of financial data may help identify emerging risks, enable market participants and regulators to see and better understand financial networks and interconnections, enhance financial stability, bolster consumer protection, and increase access to the underserved. Data can also increase transparency in the financial system for market participants, regulators and the public. These data sets, however, can raise significant questions about security and privacy; ensuring data quality; protecting against discrimination or privacy intrusions; managing, synthesizing, presenting, and analyzing data in usable form; and sharing data among regulators, researchers, and the public. Moreover, any conflicts among regulators and financial firms over such data could create opportunities for regulatory arbitrage and gaps in understanding risk in the financial system. The Big Data in Finance Conference, co-sponsored by the federal Office of Financial Research and the University of Michigan Center on Finance, Law, and Policy, and held at the University of Michigan Law School on October 27-28, 2016, covered a number of important and timely topics in the worlds of Big Data and finance. This paper highlights several key issues and conference takeaways as originally presented by the contributors and panelists who took part.

FinQL: A Query Language for Big Data in Finance

This paper presents the design rationale and specifications for FinQL, a new, domain specific, query language that addresses the complexities of time series data, research idea articulation, and backtesting in finance. The language empowers end users by enabling them to write queries in a syntax similar to the natural language used in the domain. We present the underlying domain model, define the structure and syntax of FinQL, and show how it enables both longitudinal and cross-sectional analysis as well as some automation of the research process. We demonstrate the expressive power of FinQL through examples and extend it to activity-based and range-based series.

Using Big Data in Finance: Example of Sentiment-Extraction from News Articles

There is much discussion and research in finance on using "big data" to understand market "sentiment."