Creative Commons

This dissertation was written as part of the MSc in Management at the International Hellenic University –intake 2014-2015. In this paper I am trying to find causal relationship between the sales of three Apple products, Ipod, Ipad, Iphone and the Google Trends search queries by using as keywords the name of every product. All data for the sales were downloaded from Statista.com as they were complete there, but crosschecked by the SEC Fillings posted by Apple quarterly. Apple is publishing the sales of all products in a quarterly basis only and without special clarifications about each product and model. For the data downloaded from Google Trends I needed to transform them from weekly to quarterly range. In all data I applied Seasonality Tests and Velocity Tests as it was expected from theory to have evidence of existence of both of them. After the appropriate transformations, when Seasonality and Velocity existed I continued with Unit root testing, in order to decide the level of Integration of my data. For Unit Root testing I used the Augmented Dickey-Fuller (ADF) test, the Dickey-Fuller GLS (ERS), the Phillips- Perron Test and the Kwiatkowski-Phillips-Schmidt-Shin (KPSS) Test. After conducting all Unit Root tests I ended up that all data were in zero level of integration I (0) so I could proceed with them without any further transformation. The most crucial test for my work was the Granger Linear Causality test between the Google Trends Index and the actual sales of every product. The results showed that there is linear causality relationship between the two variables. As for the conclusions, a full causal relationship was revealed. Indeed, I can suppose that there is the possibility to make forecasts in the sales of Apple products by examining the search queries in Google of the three keywords “Ipod”, “Iphone”, “Ipad”.