Building Probabilistic Causal Models using Collective Intelligence

The Journal of Financial Data Science Volume 4, Issue 2 in May 2022

Posted: 8 Feb 2022

See all articles by Olav Laudy

Olav Laudy

Causality Link

Alexander Denev

University of Oxford

Allen Ginsberg

Causality Link

Date Written: March 19, 2021

Abstract

The purpose of this paper is to show a novel approach to automatically generating Probabilistic Causal Models (Bayesian Networks (BN)) by applying Natural Language Processing (NLP) techniques to a corpus of millions of digitally published news articles in which views by different authors are expressed on the future states of economic and financial variables, and geopolitical events. The BNs that we will show how to derive will represent the wisdom-of-the-crowds forward-looking point-in-time views on various variables of interest and their dependencies. These Bayesian Networks are likely to be of interest to asset managers and to economists who want to gain a better understanding of the current drivers of an economy based upon a rigorous probabilistic methodology. Additionally, in an asset allocation context, the BNs we derive can be fed to an optimization engine to construct a forward-looking optimal portfolio given the constraints of the asset manager (e.g., budget, short constraints etc.). We demonstrate various automatically derived Bayesian Networks in a financial context

Keywords: NLP, causal graphs, Bayesian Networks, macro-finance, expert elicitation, knowledge mining.

Suggested Citation

Laudy, Olav and Denev, Alexander and Ginsberg, Allen, Building Probabilistic Causal Models using Collective Intelligence (March 19, 2021). The Journal of Financial Data Science Volume 4, Issue 2 in May 2022, Available at SSRN: https://ssrn.com/abstract=3808233 or http://dx.doi.org/10.2139/ssrn.3808233

Olav Laudy (Contact Author)

Causality Link ( email )

UT
United States

Alexander Denev

University of Oxford ( email )

Mansfield Road
Oxford, Oxfordshire OX1 4AU
United Kingdom

Allen Ginsberg

Causality Link ( email )

UT
United States

Do you want regular updates from SSRN on Twitter?

Paper statistics

Abstract Views
4,256
PlumX Metrics