This thesis investigates the predictability of trading strategies in the European and American stock market from 2001 to 2013. More specific, we examine the indices CAC40, DAX, FTSE100, NASDAQ and S&P500 first with the simple moving averages, then with trading rules based on the forecasts of time – series models and finally with the combination of the technical trading rules and time –series models. The significance of the examined trading rules tested with standard t – tests. The standard tests results show that technical trading rules are the most profitable strategy, second follows the combined and then the time – series rules as the least profitable trading strategy related to buy – and – hold strategy. / Σκοπός της παρούσας εργασίας είναι η διερεύνηση της προβλεπτικής δυνατότητας στρατηγικών επενδύσεων που εφαρμόζονται στην Ευρωπαϊκή και Αμερικάνικη χρηματιστηριακή αγορά, για τη χρονική περίοδο 2001-2013. Πιο συγκεκριμένα θα εξετασθούν οι δείκτες CAC 40, DAX, FTSE 100, NASDAQ και S&P 500, με κανόνες κινητών μέσων όρων, με κανόνες που βασίζονται σε μοντέλα πρόβλεψης χρονολογικών σειρών και με κανόνες συνδυαστικών των δύο ανωτέρω. Οι παραπάνω στρατηγικές θα συγκριθούν με την στρατηγική διακράτησης (Buy-and-Hold), που έχει ορισθεί ως benchmark στρατηγική και η σημαντικότητα των αποτελεσμάτων θα εξετασθεί με στατιστικούς ελέγχους t-statistics.

The present study aims to develop an analysis package for combining forecasts that are obtained from different forecast methods. The package is composed of three modules, namely, the data input module, the data analysis module and the combination module. In data input module, the data is entered and saved as an Excel file with the given name. In data analysis module, the program computes the forecasts of the selected methods and displays the forecast results, the mean absolute errors, the mean square errors and the mean absolute percentage errors of these methods. In combination module, the forecast results, computed in the data analysis module, are combined according to the selected combination methods. All the detailed calculations of the forecasts and the values assigned by the program to minimize the mean absolute deviations, the mean square errors and the mean absolute percentage errors are displayed under the columns of the related method on the Excel spreadsheet of the file.