Abstract:
In this thesis, we find optimal entry and exit prices and optimal lot size to trade the EUR/USD currency in the 15-minute period after the releases of US unemployment claims numbers. We use 10-second exchange rate data and the news announcement data from January 2010 to January 2014 in this study. We fit the dynamic of the EUR/USD prices to a 2-Regime mean-reverting jump-diffusion model using the Expectation-Maximization algorithm. The out-of-sample tests show that it is better to avoid some trades and choosing an appropriate lot size can reduce losses from force closure and improve returns. The results also suggest that using a combination of long and short strategies for each type of news may improve the trading performance.
