Fourier method techniques at the interface between local and stochastic volatility

Practitioner Lecture Series

In asset classes such as FX it is important to construct a model that permits smile dynamics to be tuned between sticky-delta and sticky strike modes of behaviour. Sticky delta is naturally described by purely stochastic volatility, as the variance process is invariant with respect to changes in the level of the tradeable asset. Sticky strike, however, is described by a local volatility function that depends on the level of the tradeable asset. As a result, LSV models which combine features of local and stochastic volatility models have become ubiquitous in industry. They are, however, difficult to calibrate and heavily reliant on numerical solution (requiring 2D PDEs with correlation terms). In this talk, I attempt to step back and look at hybrid type models and recent modelling advances which have the potential to simplify these challenges, by taking advantage of those local and stochastic volatility models which permit some measure of analytical tractability at the initial calibration stage.




Dr Iain J Clark (IainJClark)

Wednesday, November 20, 2013 - 18:00
to 19:00