Event-Based Exchange Rate Modelling

Specialeforsvar ved Tobias Søberg Nierhoff

Titel:Event-Based Exchange Rate Modelling

Abstract: It is commonly known that exchange rates change continually, and there are numerous of economic factors that determine the movements. Political events may also have an effect on the exchange rate. An example is the EU referendum in the United Kingdom in 2016, where UK voted in favour of leave. The days after the referendum the British Pound was weakened significantly against almost any currency. This thesis wants to investigate how the Britsh pound was effected by the EU referendum. Based on option data, we investigate the GBPUSD behaviour with a few different models, in which the output are risk-neutral densities that describe the GBPUSD return on the day after the EU referendum. In the first model, we derive the pricing formula for a European call option for exchange rates, and find the relationship between options and the risk-neutral densities. The model shows high uncertainty, but is not sufficient to describe the GBPUSD behaviour. Thus, we construct two different mixture models, in which we are able to obtain densities on the exchange rate conditional on the outcome of the EU referendum. The differences in the models are the weights for each conditional density. One model uses betting quotes as risk-neutral event probabilities, and the other model try to estimate the weights within the model. Both models work very well, and the exchange rates conditional on leave are very close to the actual GBPUSD on the day after the event. In order to investigate whether the betting market is correlated with the FX market, the weights for each model are compared. There is no significant correlation between the markets, as the weights are not similar. In the last part of the thesis we investigate the behaviour of GBPUSD on March 29, 2019, where UK was supposed to leave EU. To do this, we expand the mixture models to obtain more than two outcomes. Conditional on a Brexit with no deal, the models predict a negative effect on the British pound. However, the small probability of no deal Brexit in mind, the models do not show significant effect on GBPUSD, which was also the result on the day after March 29, 2019. It seems the world was expecting an extension of Brexit and so was the FX market.

Vejleder: Rolf Poulsen
Censor:   David Sloth Pedersen, Danske Bank