Database Challenges in Financial Misconduct Research
Jonathan M. Karpoff
University of Washington - Michael G. Foster School of Business
D. Scott Lee
University of Nevada, Las Vegas - Lee Business School
Gerald S. Martin
American University - Kogod School of Business
February 26, 2013
Georgetown McDonough School of Business Research Paper No. 2012-15
More than 100 research papers that examine financial misconduct draw from one of four primary databases. We document and explain five features in each of these four databases that pose challenges for researchers in many applications. First, the dates provided by these databases lag the initial revelation of the financial misconduct by an average of 150 to 1,017 calendar days, depending on the database. Second, the events in these databases capture, on average, only 6% to 36% of the value-relevant announcements associated with the cases of misconduct they identify. Third, these databases omit a large fraction – up to 62% – of the events they seek to capture. Fourth, between 9% and 62% of the events in these databases are duplicates in the sense that multiple events identify the same underlying cases of misconduct. And fifth, between 46% and 98% of the events in the databases do not involve financial fraud and for many applications should be culled from the sample. We show that these five database features lead to economically meaningful biases in event study applications.
Number of Pages in PDF File: 66
Keywords: Financial misconduct, restatement, class action lawsuit, Securities and Exchange Commission
JEL Classification: G38; K22; K42; M41working papers series
Date posted: July 19, 2012 ; Last revised: September 4, 2013
© 2013 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollo7 in 0.344 seconds