Each model requires different depth of the model validation. Additional to that each payoff requires different numerical methods. It is essential to choose the most appropriate model among the available list based on the customer requirement and capabilities. Thus, a model acceptable in one financial institution may not be suitable in another.
Our job is to select the model based on customer risk appetite, internal risk framework and model governance. In line with the regulatory requirement we validate the model by building an independent model or models and benchmarking the objective model against them.
Stress testing of the original model under extreme scenarios to determine the domain of the market data and model parameters for which the model maintains its accuracy and stability. to analyse the model performance in explaining the daily P&L changes, to identify the market data or model parameter uncertainty.
InteractivePricing specializes in high-performance engines for the pricing of derivatives. Our area of expertise is parallel and high-performance computing (grid, GPA and FPGA), highly optimized numerical methods.