A fundamental task in financial engineering is to develop empirically realistic as well as tractable derivative models. For tractability reasons many standard models are assumed to have time-homogeneous local characteristics (i.e. drift, diffusion coefficient, jump measure), which however, are undesirable from the empirical standpoint in many applications, as they cannot capture time dependent behavior such as seasonal spikes observed in electricity spot prices, and cannot achieve satisfactory results in calibrating the term structure of interests (e.g. implied volatilities). The aim of this project is to study the theory and applications of a new technique called additive subordination for modeling time dependency in financial engineering

Department of Systems Engineering and Engineering Management, CUHK