The results of this study are important for audit firms to consider when designing their group discussions, required by US and International auditing standards, regarding the susceptibility of their clients’ financial statements to fraud and ways these potential frauds could be enacted.  The study’s findings suggest that interacting groups with brainstorming guidelines and pre-mortem groups (i.e., where participants engage in a backward-thinking process, in which they envision the event of an undiscovered fraud emerging after the financial statement audit period) generate the highest quantity and quality of fraud ideas.  The study’s results are important for audit firms to consider, since the quantity and quality of ideas generated in the fraud group discussions are key characteristics of the output of such discussions.  Specifically, the quantity of ideas is important because the “mere mention” of ideas during the meeting is likely to elevate auditors’ overall skepticism and alert them to a variety of potential frauds they could encounter. The quality of ideas is an inherently important output of the discussion, and the rarity/uniqueness of ideas (a dimension of quality in the study) is important because audit team members are likely to already be aware of obvious potential frauds, and so the group discussion could be best used to construct more “out of the box” frauds that are less likely to be known.

Trotman, K.T., Simnett, R., and A. Khalifa. 2009. Impact of the Type of Audit Team Discussions on Auditors’ Generation of Material Frauds. Contemporary Accounting Research. 26(4): 1115 – 1142.

The PCAOB has emphasized the importance of detecting fraud as an objective of the financial statement audit and has prompted auditors to ensure that they fulfill the requirements of SAS No. 99, which instructs auditors to engage in a fraud brainstorming session for each audit engagement. The goal of the brainstorming session is to share ideas among members of the audit team about ways in which management could intentionally misstate the financial statements, conceal intentional misstatements, and engage in asset misappropriation.  In a similar fashion, ISA 315 and ISA 240 require the audit team to engage in a group discussion about the vulnerability of a client’s financial statements to intentional or unintentional misstatements, but these international standards do not specifically refer to the term “brainstorming”. 

The authors observe that there are a variety of group discussion formats that could satisfy the requirements of these standards. This motivates the authors’ investigation regarding which types of discussions are optimal, in the sense of generating fraud ideas (quantity and quality), estimating the likelihood of the frauds occurring, and engaging in mental simulations of potential fraudulent misstatements (including envisioning how the fraud may be enacted).  The authors examine the output of the following types of group discussions: 

• An interacting group with brainstorming guidelines (i.e., the group is told that “criticism is ruled out; the more creative the idea, the better; combination and improvement are sought; and most important: quantity is wanted)
• An interacting group without the brainstorming guidelines stated above.
•  An interacting group with pre-mortem instructions (i.e., auditors are told to envision a scenario where it is months after an audit is completed and no material fraud was uncovered, but the media has now revealed that there was a material financial reporting fraud for the audit client).

The research evidence is collected prior to 2007 at a senior associate auditing training in Sydney and Melbourne for a Big 4 audit firm.  Prior to the experimental session, the senior audit associate participants read an introduction to the experiment case (e.g., company background, industry information, financial performance, instances of unethical behavior) and participate in a discussion of the audit strategy and engagement scope.  During the experimental session, the participants are placed into groups of three, given all of the case materials about the audit client that they read prior to the experimental session, and are instructed to discuss and record a list of potential misstatements due to fraud.  They are also asked to record the significant account/disclosures impacted, significance of the fraud, and likelihood of occurrence. The participants were then asked to assess the likelihood that a material misstatement due to fraud occurred at the client, select and describe one such potential misstatement, how the fraud could have been perpetrated, and how the perpetrator could have covered up the fraud.  

• Both interacting groups with brainstorming guidelines and pre-mortem groups generate a higher quantity of fraud ideas and higher quality fraud ideas than interacting groups without brainstorming guidelines.
• Though not statistically significant, the authors do find that the pre-mortem groups and interacting groups with brainstorming guidelines derive more rare/unique fraud ideas than interacting groups without brainstorming guidelines.
• There were no significant differences among group types with respect to participants’ assessments of the likelihood of fraud occurring.
• The authors found a significant positive relationship between the completeness of participants’ mental simulations and their fraud likelihood assessments for the interacting groups with brainstorming guidelines and pre-mortem groups, but not for interacting groups without brainstorming guidelines.