Recent Trends in Empirical Finance

Abstract

Empirical finance is in crisis. The profession’s most important discovery tool is historical simulation; yet, according to the author, most backtests and time series analyses published in journals are flawed. The problem is well known to professional statistics and mathematics organizations, which have publicly criticized the misuse of mathematical tools among finance researchers. In this article, the author points to three problems and proposes four practical solutions. In an attempt to overcome the challenges posed by multiple testing and selection bias, the author emphasizes the need to move from an individual-centric to a community-driven research paradigm. Technologies that derive peer P-values can correct low retraction rates. Stronger theoretical foundations and closer ties with financial firms would help prevent false discoveries.

TOPICS: Simulations, portfolio theory, portfolio management/multi-asset allocation