University of Patras
Εφαρμογές σύγχρονων μεθόδων ανάλυσης δεδομένων στον τραπεζικό τομέα
Abstract
dc:descriptionIn this thesis, we propose the use of modern data analysis methods to model risk measures, required in the operation of the banking industry. More specifically, we focus on volatility, as predicting volatility in financial markets is crucial for investment strategies, managing risk, and making economic decisions. The forecasting methods used in this PhD are based on the use of advanced statistical models, which allow us to model volatility based on historical financial data. More specifically, we use GARCH (Generalized Autoregressive Conditional Heteroskedasticity) models as "state of the art" models for modeling the price fluctuations of bond yields in the time. In addition, we make use of statistical/economic indicators, such as the VIX index or economic uncertainty to verify the predictive value of such indicators in GARCH-type models to more accurately predict historical volatility. Finally, we make use of large-volume data analysis techniques, also known as Big Data, as the recent literature has shown that they can be extremely useful for predicting volatility in financial markets. Big Data includes more complex and denser volumes of data, which conventional data analysis techniques cannot handle effectively. Big Data enables the analysis of huge time series, including data from various sources and at various levels of time scale. This allows the identification of large-scale patterns and trends that may affect volatility. It is important to note that we conclude that no single method can provide absolute predictions, but the combined use of different approaches can improve the accuracy of predictions.
Degree
thesis:*- Grantor dc:publisher
- University of Patras
- Year dc:date
- 2024
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Mamalougkou, Vasiliki
- Μαμαλούγκου, Βασιλική
Subjects
dc:subject × 48- Μεταβλητότητα
- Πρόβλεψη
- Διακύμανση
- Απόδοση
- Αβεβαιότητα
- Δείκτης
- GARCH μοντέλο
- Μεγάλα δεδομένα
- Όγκος δεδομένων
- Χρονοσειρά
- Οικονομική αβεβαιότητα
- Ημερήσια δεδομένα
- Τραπεζικά ιδρύματα
- Ελλάδα
- Χρηματοπιστωτικές αγορές
- Τυπική απόκλιση (SD ή σ)
- Κίνδυνος
- Επενδυτές
- Όγκος συναλλαγών
- Τεχνικές μηχανικής μάθησης
- Τυχαία δάση
- Volatility
- Economic policy uncertainty
- Intraday data
- Banks
- Greece
- Big data
- Prediction
- Fluctuation
- Bond yields
- Garch models
- Huge time series
- Random forests
- Bank risk-taking
- Financial markets
- Investors
- Black Scholes model
- Box-Jenkins
- Utilizing machine learning
- Volume volatility
- Οικονομικά και Επιχειρήσεις
- Κοινωνικές Επιστήμες
- Οικονομετρία
- Στατιστικές, Πιθανότητες και Αβεβαιότητα
- Economics and Business
- Social Sciences
- Econometrics
- Statistics, Probability and Uncertainty
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/56343