Date of Award

Spring 4-2-2024

Document Type

Dissertation

Degree Name

Doctor of Business Administration (DBA)

First Advisor

Dr. Linda McCann, DBA, CMA, CPA (inactive)

Second Advisor

Dr. Susan Miserak, PhD

Third Advisor

Mr. David Vetta, CFP, MBA, BS

Abstract

Abstract

This dissertation was developed to aid nonprofessional investors in recognizing the possibility of fraud occurring in a company they are holding or considering purchasing in their investment portfolios. Although there are published anti-fraud models and strategies available to use in evaluating the likelihood of fraud in a holding, those investment tools are more oriented to larger institutions and do not consider the limitations of nonprofessional investors in terms of (1) the time it takes to do the research, (2) the ability to access relevant data, and (3) the lack of advanced investment knowledge required to perform an in-depth analysis of a company’s financial fraud status.

Given the limitations of this growing stock market section, the question was whether there is an investment tool that nonprofessional retail investors could use to evaluate the risk of financial statement fraud. To answer this question an anti-fraud prediction model (the Retail Investor Risk Model (RIRM)) and red flag checklist were developed that takes one hour or less and uses readily available public information to evaluate the risk of financial statement fraud.

A comprehensive logistic regression model was constructed by analyzing fifty companies charged with committing financial statement fraud according to SEC AAER investigations and an equal number of competitors with no fraud reported over the same periods. The model integrated variables using a three-prong approach encompassing corporate governance, financial health, and considered ‘too-good-to-be-true’ measures, resulting in a correct fraud predictive capability exceeding 60% accuracy. Key variables driving this predictive power include the composition of the board (the presence of female and independent members), financial metrics like changes in debt-to-equity ratio and gross margin index, and market behavior indicators like stock price volatility. The model satisfied stringent tests for statistical significance and goodness of fit, strengthening confidence in its reliability and applicability. These metrics align closely with established theories like the fraud triangle theory and findings from prior research.

Using easy and quick-to-compute measures, the new model gives the nonprofessional investor a tool to evaluate the likelihood of fraud occurring in an investment holding. A Red Flag checklist expanded the capabilities of the model with additional variables and data that were also deemed appropriate for the retail investor to consider when evaluating the possibility of fraud occurring – industry (e.g., technology or healthcare), complexity with revenue recognition (e.g., contracts), and the sales growth rate.

This research contributes to existing literature by introducing a new approach to fraud detection. Utilizing an easy-to-use statistical model alongside understandable red flags, the study offers a unique methodology tailored for retail investors. Importantly, the model and red flags are designed to be accessible and understandable to all investors, particularly those with a basic two-year education in business.

Download

Share

COinS