Ιδρυματικό Αποθετήριο ΟΠΑ
Μόνιμο URI για αυτήν την κοινότηταhttps://beta-pyxida.aueb.gr/handle/123456789/1
Νέα
Αυτή είναι η κοινότητα από το παλιό σύστημα με ID:cid:1
Περιηγούμαι
Πλοήγηση Ιδρυματικό Αποθετήριο ΟΠΑ ανά Συγγραφέα "Agoraki, Maria-Eleni"
Τώρα δείχνει 1 - 4 από 4
- Αποτελέσματα ανά σελίδα
- Επιλογές ταξινόμησης
Τεκμήριο Analysis of determinants of stock performance: the case of shipping firms(30-12-2022) Ζουμπουρλής, Σπυρίδων; Αλβέρτης, Φραγκίσκος; Zoumpourlis, Spyridon; Alvertis, Fragkiskos; Athens University of Economics and Business, Department of Accounting and Finance; Kavussanos, Emmanuel G.; Androutsopoulos, Ion; Agoraki, Maria-EleniΣκοπός της παρούσας διπλωματικής εργασίας είναι να εξετάσει τους κύριους μοχλούς της απόδοσης των τιμών μετοχής εισηγμένων ναυτιλιακών εταιρειών. Αυτοί οι παράγοντες αναλύονται σε μεταβλητές που αφορούν συγκεκριμένους δείκτες επιχειρήσεων, μακροοικονομικές μεταβλητές και παράγοντες που αφορούν τον συγκεκριμένο κλάδο. Η μελέτη χρησιμοποιεί ένα δείγμα 75 εταιρειών που είναι εισηγμένες στις Ηνωμένες Πολιτείες για την περίοδο 1990 έως 2020. Ενώ οι εταιρείες αυτές είναι εισηγμένες στις ΗΠΑ, έχουν την έδρα τους σε 11 διαφορετικές χώρες, γεγονός που επιτρέπει διατομεακές συγκρίσεις. Η έρευνα αυτή χρησιμοποιεί τις μεθόδους της πολλαπλής γραμμικής ανάλυσης παλινδρόμησης, ιδιαίτερα της ανάλυσης παλινδρόμησης πάνελ για την επίτευξη του σκοπού. Τα ευρήματα δείχνουν ότι οι κύριοι παράγοντες πρόβλεψης των αποδόσεων των μετοχών των ναυτιλιακών εταιρειών είναι το ασφάλιστρο κινδύνου αγοράς, η απόδοση των μετοχών αξίας, η απόδοση των μετοχών μικρής κεφαλαιοποίησης, η οικονομική ανάπτυξη και η υλικότητα των περιουσιακών στοιχείων. Μεταξύ αυτών των σημαντικών προγνωστικών παραγόντων, μόνο η υλική αξία περιουσιακών στοιχείων έχει αρνητική επίδραση στις αποδόσεις των μετοχών, υπονοώντας ότι οι επενδυτές προτιμούν ναυτιλιακές εταιρείες μικρότερης έντασης κεφαλαίουΤεκμήριο The capacitated vehicle routing problem. A comparison between integer linear programming and metaheuristic search methods(2021) Chiotis, Minas A.; Χιώτης, Μηνάς; Athens University of Economics and Business, Department of Business Administration; Kouretas, Georgios; Agoraki, Maria-Eleni; Nikolopoulou, AmaliaThis thesis presents a comparison between two ways of solving a Capacitated Vehicle Routing Problem (CVRP) - a special variation of the classic Vehicle Routing Problem (VRP) - with Linear Integer Programming versus metaheuristic methods from Operating Research Google developers (OR-Tools). This problem is traditionally as one of the most important issues for companies delivering goods and services and in general in the sector of Logistics and Supply chain. The scope of this thesis is to deliver the best solution between two different methodologies and propose the best use of them at any case they face.In the first part we can find the literature about the Supply Chain Management and the VRP problem in general and the variations of it. There are mentions about the creators and researchers of the sector and highlighting the milestones of their life which was dedicated in their work.We analyze that in the CVRP case we examine the scope of the problem is to serve as many as possible customers and to minimize the total distance of each truck aiming to minimize the total cost of transportation and maximizing the company’s profitability and competitive advantage against other providers. The constraint is the capacity of each vehicle which is not tolerated to be exceeded. The comparison between them focuses on which of the two methods is the most effective in a manner of time, precision, and efficiency.After that we analyze the methodologies we are going to use and based on we are going to implement the experiment at each case. Here we use the linear integer programming to solve the problem and on the other hand a bunch of metaheuristics which come from OR Tools. Details about the integration is listed in the proportionate chapter.Then implementation and conclusion take place and are starting to represent the whole mindset of the problem, the issues, the measurements and finally the results based on how we make our comparison. All in all, the outcome is different and at any case the user (manager) is called to make a decision that has to choose between these two methodologies.Τεκμήριο Hedging an equity portfolio using volatility options(09/27/2021) Simantirakis, Michail; Σημαντηράκης, Μιχαήλ; Athens University of Economics and Business, Department of Business Administration; Spyrou, Spyros; Agoraki, Maria-Eleni; Kasimatis, KonstantinosThe purpose of this dissertation is to demonstrate how classic portfolios can benefit from implied volatility. Information about the introduction of implied volatility indices and volatility derivative instruments are presented, as well as the desirable characteristics of the implied volatility, which lead investors to create portfolio strategies including the appropriate implied volatility derivatives. The strong negative correlation between implied volatility indices returns and the returns of equities combined with the asymmetric relation between their returns during negative and positive days of returns on equities are the main characteristics that enhance the interest for diversification or hedging through mixed portfolios of equities and implied volatility. A variety of academic research around volatility strategies is also presented. An empirical application of strategies is implemented using the core European Equities Index (STOXX 50) and the corresponding European Volatility Index (VSTOXX). The application investigates whether it is possible to provide diversification or hedging on an equity portfolio during tough periods in financial markets, like the COVID-19 market crash in March 2020. The implementation of the strategies uses threshold models of implied volatility which give signals for adding the appropriate volatility options in the equity portfolios when it is considered necessary. The strategies create different portfolios which are compared in between them and with the benchmarking equity index portfolio. Characteristics and performance metrics, such as Sharpe ratio, Loss deviation, and conditional-VaR are calculated for evaluating the performance and the risk of portfolios. The results of the empirical application despite the limitations are consistent with previous studies and show that implied volatility derivatives could provide the desired diversification and hedging during a market crash, but investors must be willing to pay the cost of the premiums when the market is stable or rising.Τεκμήριο Predictability of bitcoin price using Twitter sentiment analysis(09/27/2021) Papatheodorou, Konstantinos; Παπαθεοδώρου, Κωνσταντίνος; Athens University of Economics and Business, Department of Business Administration; Spyrou, Spyros; Agoraki, Maria-Eleni; Kasimatis, KonstantinosThe purpose of the specific research we conduct is to make accurate predictions of Bitcoin (BTC) price fluctuations using an extensive social Sentiment Analysis in two forms. Firstly, by taking the raw Sentiment data and combine them with BTC historical prices and secondly by applying the Autoregressive Moving Average model (ARIMA) in the raw Sentiment data and correlate the result with the movement of the historical prices.Social Sentiment Analysis have been used to classify the opinion of the under study tweets producing three possible results in each tweet classification: positive, negative or neutral. The tweets chosen for the research have been mined from Twitter API and belongs to a financial analyst with a crucial influence in public opinion about cryptos. The final result of the comparison between raw Sentiment data and BTC historical prices for the under study time period gave a non significant correlation of 3.9% (Pearson coefficient metric) and 0.2% (Spearman coefficient metric).Making use of the ARIMA model to smoothen the raw Sentiment data curve had a significantly better outcome in the predictions. That, happened by bisecting the data in two period to train and test the model. The final correlation of the predictions and BTC historical prices produced a correlation of 64.5% (Pearson coefficient metric) and 61.4% (Spearman coefficient metric) that indicates a significant correlation between predictions and historical data.