Selecting Computational Models for Asset Management: Financial Econometrics versus Machine Learning—Is There a Conflict?

Abstract

Successful forecasting requires integrating financial theory, market behavior, exploding sources of data, and computational innovation. Building accurate computational models can be achieved by assembling the most comprehensive toolbox. Both financial econometrics and machine learning approaches help to achieve this objective. Machine learning tools provide the ability to make more accurate predictions by accommodating nonlinearities in data, understanding complex interaction among variables, and allowing the use of large, unstructured datasets. The tools of financial econometrics remain critical in answering questions related to inference among the variables describing economic relationships in finance; when properly applied, their role has not diminished with the introduction of machine learning.

TOPICS: Big data/machine learning, simulations, statistical methods

Key Findings

• • Successful forecasting requires integration of financial theory, market behavior, exploding sources of data, and computational innovation.

• • The well-known documented stylized statistical factors associated with financial market variables and their importance in choosing and applying the proper statistical techniques and machine learning algorithms are discussed.

• • Guidelines are provided for integrating the characteristics of data, inference features of financial econometrics, and prediction capabilities from machine learning.