Contribution to financial econometrics and sensitivity analysis.

Summary Since the end of 2008, my research work has gradually turned towards problems related to financial econometrics. More precisely, the aim was to incorporate in classical GARCH models probability laws capable of taking into account the phenomena of asymmetry and leptokurticity that can be observed in most financial series. The inclusion of these two stylized facts in GARCH models is fundamental both for improving the predictive power of the models (forecasting) and for the valuation of derivatives (pricing). The abandonment of the Gaussian hypothesis is however not without consequences.