Anmerkungen:
Nach Informationen von SSRN wurde die ursprüngliche Fassung des Dokuments October 5, 2018 erstellt
Beschreibung:
Financial markets are becoming increasingly event driven. These can take the form of economic releases, speeches and elections. The arrival of this new information often causes the underlying asset to jump. Consequently, the expectation of these known announcements can produce significant distortions in option volatility surfaces. This presents a challenge, or opportunity, for pricing and risk management. We therefore undertake an examination of large events. Firstly, we perform an empirical analysis of the UK referendum in 2016. This is in order to give market practitioners intuition for these extreme scenarios and to guide us towards a quantitative approach. As a result, we introduce a new bimodal jump model. By varying the two jump parameters, we show how to parsimoniously produce the aforementioned smile distortions – specifically: skews, frowns and W-shapes. Furthermore, we demonstrate the model's calibration to the observed volatility surface leading up to the referendum, and the French election in 2017. If jump parameters are frozen with time, the model can be used in the temporal volatility interpolation. Finally, we discuss how the jump probabilities are becoming increasingly observable via alternative data, such as betting markets, election polls and research reports