http://onlinelibrary.wiley.com.libproxy.smu.edu.sg/doi/10.1111/1911-3846.12063/epdf

The Ties that Bind: The Decision to Co-Offend in Fraud^†

Clinton Free¹ and Pamela R. Murphy²

Contemporary Accounting Research

Volume 32, Issue 1, pages 18–54, March 2015

It is frequently observed that fraud has a greater economic impact on society than any other category of crime. Arguing that both research and practitioner frameworks in auditing and forensic accounting have tended to adopt an individualizing perspective predicated primarily on solo offending, this article adopts an inductive approach to consider why individuals co-offend in fraud. It reports the results of a set of interviews with 37 individuals convicted of a range of frauds including financial statement fraud, insider trading, credit card fraud, money laundering, and asset misappropriation. In each instance, the fraud was perpetrated by a group of two or more co-offenders. Based on inductive, exploratory case coding, we find that reasons for co-offending vary according to the type of bond that exists between co-offenders. Two dimensions of fraudulent co-offending are identified—the primary beneficiary of the fraud and the nature of group attachment—to derive three distinct archetypes of bonds between co-offenders: (1) individual-serving functional bonds, (2) organization-serving functional bonds, and (3) affective bonds. Key elements of each archetype as well as their impact on the decision to co-offend are examined. Our findings suggest that the social nature of fraud is not merely an incidental feature of the crime but is instead a potential key to understanding its etiology and some of its distinctive features. They also support the need for diagnostic tools to move beyond individualistic analyses of fraud toward a broader, group-sensitive assessment of fraud risk.

http://onlinelibrary.wiley.com.libproxy.smu.edu.sg/doi/10.1111/1911-3846.12090/epdf

Trading Behavior Prior to Public Release of Analyst Reports: Evidence from Korea

Bobae Choi¹, Kooyul Jung² and Doowon Lee¹

Contemporary Accounting Research

Volume 32, Issue 1, pages 105–138, March 2015

Abstract

This paper investigates information leakage from analyst reports prior to their public release. Previous studies document abnormal trading by institutions or short selling before announcement of recommendation changes. Such prerelease abnormal trading is interpreted as evidence of information leakage from analyst reports. However, if sophisticated investors obtain information similar to what analysts have from other sources, abnormal prerelease trading patterns would be observed even if there were no information leakage from analyst reports. This paper, using a unique data set from Korea, aims to determine whether a direct causal link between recommendation changes and prerelease trading exists, by comparing trading behavior of client investors with non-client investors. We find that abnormal prerelease trading by client investors, especially client institutions, is earlier in timing and greater in magnitude than that of other investor groups, supporting the information leakage hypothesis. We further find that net buying by client institutions and client large individuals is positively associated with firm, analyst, and earnings forecast change variables that influence formulation of recommendation changes and their impacts.