| This research seeks to identify the fraudulent financial reporting characteristics of the computer industry, the most likely industry for fraudulent reporting. Using a holistic strategic-systems lens approach supported by logistic and neural network analyses, this research identifies strategic, core and supporting business process, and financial performance variables found within the business and financial reporting models of fifty-two-computer companies accused of fraudulent financial reporting practices by the Accounting and Enforcement Division of the Securities and Exchange Commission between September 20, 1995 and April 5, 2002. These firms are then compared to similar ones not charged during the same time period.; This research finds some financial characteristics similar to and some distinct from non-computer manufacturing, merchandising, and other industries. Similarly, the computer industry has increasing profit margins and return on total assets and decreasing cash flows in the presence of deteriorating accounts-receivable and inventory turnover ratios.{09}Dissimilar are measures of increasing sales revenue, receivables, ending inventories, ratio of inventory to total assets, all considered significant fraudulent characteristics of non-computer industries. The research also finds, consistent with the industry's strategic and business process risks, three ratios (for R&D investment, for sales and marketing expenses, and for free cash flows), that explain by their weakness the SEC-accused firm's apparent fraudulent reporting behavior. The research finds a fourth defining variable, the ratio of tax benefits from the exercising of stock options to operating cash flows, typically available in many industries and missing from extant research. Finally, the research finds logistic regression analysis more reliable than the neural network analysis in predicting suspected fraudulent out-of-sample firms, and finds both analyses generally complementary. |
