Replication in Financial Economics

Abstract

The existing replication policies at top finance journals are far weaker than the policies at top economics journals. This paper explores both the costs and benefits of having a stronger replication policy in the context of my failed 2010 initiative to develop a unified policy across all top finance journals. For example, the most obvious cost of a replication policy is the additional burden it imposes on authors in answering questions about both the code and data. Indeed, this cost is disproportionately placed on our most productive researchers – potentially leading to less innovation. On the other hand, having a strong policy would likely reduce research misconduct – in particular, soft misconduct such as p-hacking. I present a framework to mitigate the costs associated with replication and maximize the benefits.

Keywords: Replication, Research Culture, Hard misconduct, Soft misconduct, Proprietary data, Robustness, P-hacking, Reverse p-hacking, Falsification, Fabrication, Plagiarism

JEL Classification: G00, G01, G10, G11, G12, G13, G14, G15, G17, G18, G19, G21, G22, G23, G24, G28, G29, G31, G32, G33