The Cost of Fraud Prediction Errors

We compare seven fraud prediction models with a cost-based measure that nets the benefits of correctly anticipating instances of fraud, against the costs borne by incorrectly flagging non-fraud firms. We find that even the best models trade off false to true positives at rates exceeding 100:1. Indeed, the high number of false positives makes all seven models considered too costly for auditors to implement, even in subsamples where misreporting is more likely. For investors, M-Score and at higher cut-offs the F-Score, are the only models providing a net benefit. For regulators, several models are economically viable as false positive costs are limited by the number of investigations regulators can initiate, and by the relatively low market value loss a "falsely accused" firm would bear in denials of requests under the Freedom of Information Act (FOIA). Our results are similar whether we consider fraud or two alternative restatement samples.

MeDML: Med-Dynamic Meta Learning - A multi-layered representation to identify provider fraud in healthcare

Every year, health insurance fraud costs taxpayers billions of dollars and puts patient’s health and welfare at risk. Existing solutions to detect fraudulent providers (hospitals, physicians, etc.) aim to find unusual pattern at claim level features but fail to harness provider-provider and provider-patient interaction information. We propose a novel framework, Med-Dynamic meta learning (MeDML), that extends the capability of traditional fraud detection by learning patterns from 1) patient-provider interaction using temporal and geo-spatial characteristics 2) provider's treatment using encounter data (e.g. medical codes, mix of attended patients) and 3) referral using underlying provider-provider relationships based on common patient visits within 30 days. To the best of our knowledge, MeDML is first framework that can model fraud using multi-aspect representation of provider.MeDML also encapsulates provider's phantom billing index, which identifies excessive and unnecessary services provided to patients, by segmenting frequently co-occurring diagnosis and procedures in non-fraudulent provider's claims. It uses a novel framework to aggregate the learned representations capturing their task-specific relative importance via attention mechanism. We test the dynamically generated meta embedding using various downstream models and show that it outperforms all baseline algorithms for provider fraud prediction task.

Measuring Accounting Fraud and Irregularities Using Public and Private Enforcement

Most accounting studies use only public enforcement actions (SEC cases) to measure accounting fraud. However, private cases (securities class actions) also play an important enforcement role. We discuss the legal standards and processes for both public and private enforcement regimes, emphasize the importance of screening cases for credible fraud allegations, and show both yield credible fraud measures. Further, we demonstrate these research design choices affect inferences from prior research and a hypothetical research setting. Finally, we show common measures of accounting irregularities using Audit Analytics to proxy for fraud result in significant false positives and negatives and develop a fraud prediction model for use in future research. We recommend using both public and private enforcement with appropriate screening when examining accounting fraud to reduce Type I and II errors, or reporting the sensitivity of findings across regimes. This is particularly important given the reduction in accounting-related enforcement after 2005.