This research paper aims to seek trading opportunities amongst EUR/USD pair.
Although the data was relatively complete, 0.6% of data were still missing ranging from one single minute to several hours. Therefore the data had to be consolidated using “forking method” of Mai H.M. and Tnani S. Our methodology for this research is focused mainly on two mathematical methods: coefficient of autocorrelation and cointegration. We used Person’s product moment coefficient to calculate the autocorrelation on 2010 and 2011. We have calculated the coefficient of two different periods with the same granularity (60min) with an offset of 1h starting from Monday until Friday. The graph of autocorrelation showed interesting outcome throughout different months and different weekdays as well. Some weekdays such as Thursday showed a relatively high volatility and we were able to match with the amount of financial news. The cointegration is the second method which studied for each day for a period of time of 3 years, 1 year and 3 months. Unlike autocorrelation, only specific time of the day was studied and we looked for days which were the most cointegrated. Only series for a period of 3 months were cointegrated and especially at midnight.