Fraud detection is a key issue for investors and financial authorities. The Ponzi schemeorganized by Bernard Madoff is a magnified example of a financial fraud, always possiblewhen well-orchestrated. Traditional methods to detect fraud require costly and lengthyinvestigations that involve complex financial and legal knowledge, as well as highly skilledanalysts. Based on the motto “too good to be true” that should be adopted by any rationalinvestor, we propose herein the use of a robust performance measure (named GUN*) toconstruct an Index for detection of anomalies (called IDeA). This index is based on thebasic intuition that it is not possible to properly evaluate a fund as “good” regardless thecharacteristics and risk aversion of investors. After defining the intuition behind such anindex and its economic theoretical background, we illustrate our innovative operationsresearch methodology for fraud detection and demonstrate its usefulness studying theemblematic case of the fraud by Madoff.