Hawkes processes and financial applications
Hawkes processes are a family of stochastic models to arrival of events in which the occurrence of any event would increase the intensity of further events occurring soon after.
This has been applied into areas including ecology, crime prediction, neuroscience, social networks and terrorist acts and recently finance.
With substantial changes in complexity of the modern financial market and system and rise of high frequency trading, existing modelling frameworks are insufficient in accommodating features such as non-Gaussian characteristics including long or fat tails, extreme events and clustering etc. It also can’t respond well to big questions on financial instability shaken by more permanent damages from financial crises and/or temporary but disastrous losses from mini-flash crash (eg latest flash crash in sterling in early October, 2016).
This project, therefore, aims to answer novel but cutting edge questions in this field such as convergence of multivariate Hawkes processes, forecasting ability of incorporating Hawkes processes into various risk modelling, calibration of them etc.
Research in this area requires good foundation in financial mathematics, programming skills using R or MatLab and inter-disciplinary research skills, particularly solid understanding of financial market structure, design and policy impact.
We are interested in pursuing this project and welcome applications if you are self-funded or have funding from other sources, including government sponsorships or your employer.
Please contact the supervisor when you want to pursue this project, citing the project title in your email, or find out more about our PhD programme in Mathematics.