Arthur Charpentier presented at the UQ'aM Economics Seminar on insurance segmentation and mutualization. He discussed how insurance allows for risk sharing by pooling together individuals with different risk levels. Insurers can differentiate prices through segmentation based on observable risk factors, but this leads to a "spiral of segmentation" where each group is charged their own risk-based price. Charpentier presented models to transfer risk with and without segmentation using perfect and imperfect information. He proposed using Lorenz curves and pseudo-Lorenz curves to visualize the performance of actuarial pricing models and their ability to correctly order risks. Machine learning techniques can also be applied to approximating insurance premiums based on risk characteristics.