This paper proposes a general form of mixture diffusion process to model asset price dynamics, using a mixture of infinite number of parametric diffusions. We show that the underlying asset price dynamics of the risk-neutral probability density distributions can be modeled precisely by the said mixture diffusions. %Furthermore, the diffusion process with stochastic drift and volatility terms derived from mixture diffusion also matches the entire implied volatility surface.
↧