An options-pricing approach to election prediction

John Fry, Matt Burke

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

The link between finance and politics (especially opinion polling) is interesting in both theoretical and empirical terms. Inter alia the election date corresponds to the effective price of an underlying at a known future date. This renders a derivative pricing approach appropriate and, ultimately, to a simplification of the approach suggested by Taleb (2018). Thus, we use an options-pricing approach to predict vote share. Rather than systematic bias in polls forecasting errors appear chiefly due to the mode of extracting election outcomes from the share of the vote. In the 2016 US election polling results put the Republicans ahead in the electoral college from July 2016 onwards. In the 2017 UK general election, though set to be the largest party, a Conservative majority was far from certain.
Original languageEnglish
Pages (from-to)1583-1589
Number of pages7
JournalQuantitative Finance
Volume20
Issue number10
Early online date12 Jun 2020
DOIs
Publication statusPublished - 2 Oct 2020

Cite this