Valuations provider Pricing Partners told delegates to this month’s MathFinance Conference in Germany about its recent work towards achieving smart expansion and fast calibration for jump diffusion models. The vendor promoted its recent work on approximation methods using its smart expansion method to calibrate jump diffusion models to attendees to the quantitative finance event last week.
The work was presented by Eric Benhamou, CEO of Pricing Partners, who said: “Our experience is that sophisticated models have been more resilient during the crisis. This new work allows a better consideration of risks for the smile and it is an important step forward.”
Jump diffusion models are most frequently used in option pricing and they are a form of mixture model, in which the jump process and the diffusion process are combined. A jump process is a type of stochastic process that has large discrete movements (jumps), rather than small continuous movements. A diffusion process, on the other hand, is based on continuous-time Markov process with continuous sample paths. The hybrid model of jump diffusion therefore states that prices have large jumps followed by small continuous movements.