Project CEMAPRE internal
| Title | Pricing automobile insurance and the effects of technology advancement estimation (continuation) |
| Participants | Lourdes B. Afonso, Renata Alcoforado, Agnieszka Bergel, Rui M.R. Cardoso, Corina Constantinescu, Alfredo Egídio dos Reis (Principal Investigator), Gracinda R. Guerreiro, Veronique Maume-Deschamps, Weihong Ni, Eugenio Rodriguez Martinez |
| Summary | As a follow up, we refer to our 2019 pricing automobile insurance Cemapre internal project. Although having already done most work on the 2019 project subject we still have to do some "tiding up" and further unfinished parts. Also we've been working on effects on ruin probabilities around classical and open bonus-malus methods based on claim counts. With a focus in the context of automobile insurance, this project would like to look at the current status of insurance industry in terms of new uncertainties and dependence between claim frequency and size. By new uncertainties we mean identified issues brought by recent technology developments. On the one hand, we try to start from a neutral perspective against the uncertainties embedded with unfamiliar/innovative policies. By making dynamic adjustment of premiums based upon claim records stemming from these policies thanks to the features in a Bonus-Malus System (BMS), we believe the model not only promotes an initial competitive market position for an insurer but also ensures the control of the risks. On the other hand, there are several identified issues related to recent technology advancement that have affected people's driving behaviours, e.g. the use of mobile phones according to the U.S. Department of Transportation. To quantitatively evaluate risks associated with a certain human habit is believed to facilitate an insurer's risk management strategies as well as pricing an insurance product and that will be the second concentration in this proposed project. We use a Bayesian perspective to explain the introduction of unforeseeable risks and develop it in order to allow dependence to be captured "naturally". |