US banks are losing tens of billions of dollars annually to first-party fraud, which is estimated to be as much as one-quarter or more of total consumer credit chargeoffs in the United States. The magnitude of these losses is due to two factors: first-party fraud is difficult to detect because fraudsters behave similarly to legitimate customers until they "bust out" and disappear; secondly, the exponential nature of the relationship between the number of participants in a fraud ring and the overall dollar value controlled by the operation makes it particularly susceptible to graph-based methods of fraud detection. Graph databases can help detect and mitigate three types of fraud: first-party bank fraud, insurance fraud, and e-commerce fraud. Traditional methods of fraud detection are geared towards discrete data and not connections, leading to false positives and undesired side effects in customer satisfaction and lost revenue opportunity. Entity link analysis using a graph database can uncover collusions of the type described above with a high probability of accuracy, especially when running checks during key stages in the customer and account lifecycle such as at account creation, investigation, credit balance threshold hits, or check bounces.