Financial Math Seminar

Maximum likelihood estimation of Levy processes in finance

We study the maximum likelihood estimation of Levy process models commonly used in finance. The probability density is computed from its characteristic function using the Hilbert transform approach, where the error converges exponentially in terms of the computational cost. Simulation studies show that the method is fast and accurate for monthly, weekly and daily data. Case studies illustrates the effectiveness of various Levy processes for modelling real financial data.