Prior research has established that idiosyncratic volatility of the securities prices exhibits a positive trend. This trend and other factors have made the merits of investment diversification and portfolio construction more compelling. A new optimization technique, a greedy algorithm, is proposed to optimize the weights of assets in a portfolio. The main benefits of using this algorithm are to: a) increase the efficiency of the portfolio optimization process, b) implement large-scale optimizations, and c) improve the resulting optimal weights. In addition, the technique utilizes a novel approach in the construction of a time-varying covariance matrix. This involves the application of a modified integrated dynamic conditional correlation GARCH (IDCC - GARCH) model to account for the dynamics of the conditional covariance matrices that are employed. The stochastic aspects of the expected return of the securities are integrated into the technique through Monte Carlo simulations. Instead of representing the expected returns as deterministic values, they are assigned simulated values based on their historical measures. The time-series of the securities are fitted into a probability distribution that matches the time-series characteristics using the Anderson-Darling goodness-of-fit criterion. Simulated and actual data sets are used to further generalize the results. Employing the S&P500 securities as the base, 2000 simulated data sets are created using Monte Carlo simulation. In addition, the Russell 1000 securities are used to generate 50 sample data sets. The results indicate an increase in risk-return performance. Choosing the Value-at-Risk (VaR) as the criterion and the Crystal Ball portfolio optimizer, a commercial product currently available on the market, as the comparison for benchmarking, the new greedy technique clearly outperforms others using a sample of the S&P500 and the Russell 1000 securities. The resulting improvements in performance are consistent among five securities selection methods (maximum, minimum, random, absolute minimum, and absolute maximum) and three covariance structures (unconditional, orthogonal GARCH, and integrated dynamic conditional GARCH).

This thesis will investigate if adding an exogenous variable (implied volatility) to the variance equation will increase the performance for the GARCH(1,1) and EGARCH(1,1) models based on the OMXS30 index. These models are also compared with the implied volatility itself as a forecasting/modeling method. To evaluate the models the realized variance will be used as an unbiased estimator of the conditional variance. The findings suggest that adding implied volatility to the variance equation increase the overall performance.

In this paper, we explore the relationship between a Global Mood Time Series, provided by Wall Street Birds, and the Dow Jones Industrial Average (DJIA) from April 2011 to December 2011. My econometric results show that there is no long run equilibrium relationship between the level of global mood and the level of the DJIA. These results apply to the whole period, as well as in the six-month subperiods. Furthermore, daily changes in global mood do not Granger cause DJIA returns. However, changes in global mood do appear to be useful in forecasting the volatility of the DJIA, and my results suggest that GARCH models of volatility of large-cap indexes, and potentially the market as a whole, could be strengthened by including online sentiment measures of Big Data. Measuring global mood, and quantifying its impacts, can potentially lead to superior portfolio construction as forecasting volatility is an important input in portfolio optimization. The results, as a whole, suggest that Big Data can have important implications for investment decision-making.

Η ακριβής πρόβλεψη της μελλοντικής μεταβλητότητας αποδεικνύεται ιδιαίτερα χρήσιμη για την τιμολόγηση παραγώγων προϊόντων και την αντιστάθμιση κινδύνων στη διαχείριση χαρτοφυλακίων. H τεκμαρτή μεταβλητότητα, όπως αυτή αντανακλάται στις τιμές των δικαιωμάτων προαίρεσης, αποτελεί την εκτίμηση της αγοράς για τη μελλοντική πραγματοποιηθείσα μεταβλητότητα και έχει αποδειχθεί ότι είναι πιο αποτελεσματική από την αντίστοιχη πρόβλεψη που προκύπτει από την ανάλυση ιστορικών χρονοσειρών. Η παρούσα διατριβή πραγματεύεται τη δημιουργία ενός δείκτη τεκμαρτής μεταβλητότητας για την Ελληνική χρηματιστηριακή αγορά, χρησιμοποιώντας έναν τρόπο υπολογισμού, ο οποίος είναι ανεξάρτητος από κάθε υπόδειγμα τιμολόγησης δικαιωμάτων προαίρεσης και βασίζεται σε ένα σταθμισμένο άθροισμα τιμών δικαιωμάτων. Η μεθοδολογία αυτή εφαρμόζεται για πρώτη φορά σε μια περιφερειακή, αναπτυσσόμενη αγορά, όπως το Χρηματιστήριο Αθηνών. Ο εν λόγω δείκτης τεκμαρτής μεταβλητότητας έχει τις προοπτικές να γίνει δείκτης αναφοράς των προσδοκιών για τη μελλοντική μεταβλητότητα στην Ελληνική μετοχική αγορά, καθώς αποδεικνύεται ότι υπερισχύει στατιστικά της ιστορικής μεταβλητότητας. Επίσης, οι επενδυτές του Χρηματιστηρίου Αθηνών μπορούν να χρησιμοποιούν το επίπεδό του και τις ημερήσιες μεταβολές του για να λάβουν επενδυτικές αποφάσεις, καθώς τα αποτελέσματα της οικονομετρικής ανάλυσης αποδεικνύουν ότι υπάρχει αρνητική και ασύμμετρη σχέση μεταξύ των μεταβολών του δείκτη τεκμαρτής μεταβλητότητας και των αποδόσεων του υποκείμενου μετοχικού δείκτη FTSE/Χ.Α.-20. Τέλος, η εμπειρική έρευνα καταγράφει την επιρροή της τεκμαρτής μεταβλητότητας των κυριοτέρων χρηματιστηρίων του εξωτερικού στην εγχώρια τεκμαρτή μεταβλητότητα, ενώ επιπλέον προσπαθεί να αναπτύξει ένα υπόδειγμα για την πρόβλεψη της τεκμαρτής μεταβλητότητας αυτής καθαυτής. / In this thesis a new measure of Greek stock market volatility based on the implied volatility of FTSE/ATHEX-20 index options is proposed. Greek Implied Volatility Index is calculated using the model-free methodology that involves option prices summations and is independent from the Black and Scholes pricing formula. The specific method is applied for the first time in a peripheral and illiquid market as the Athens Exchange. The empirical findings suggest that implied volatility includes information about future volatility beyond that contained in past realized volatility and in addition, prove that there is a statistically significant negative and asymmetric contemporaneous relationship between implied volatility changes and the underlying equity index returns. Finally, the volatility transmission effects on the Greek stock exchange from the major global exchanges are tested and documented. The basis of the international integration analysis, instead of the commonly used realized returns or variances, is the implied volatilities, as proxied by the corresponding implied volatility indices.

In this work, neural networks are used to forecast daily Realized Volatility of the EUR/USD, GBP/USD and USD/CHF currency pairs time series. Their performan-ce is benchmarked against nowadays popular Hetero-genous Autoregressive model of Realized Volatility (HAR) and traditional ARIMA models. As a by-product of our research, we introduce a simple yet effective enhancement to HAR model, naming the new model HARD extension. Forecasting performance tests of HARD model are conducted as well, promoting it to become a reference benchmark for neural networks and ARIMA.

Sparčiai kintant finansinių institucijų veiklos sąlygoms, didėjant finansinių rinkų nepastovumui bei veiklos mastams, atsirandant vis naujoms finansinėms priemonėms, o kartu su jomis ir naujoms finansinių institucijų rizikos rūšims, ypač išaugo prekybinio portfelio rizikos valdymo poreikis. Tą įrodo ir tai, kad po eilės solidžių ir pakankamai konservatyvių finansinių institucijų, tokių kaip Baring, Daiwa, Sumitomo, Metallgesellschaft, Orange County, Long Term Capital Management, bankrotų ar milžiniškų nuostolių patyrimo praeito amžiaus paskutiniame dešimtmetyje didžiosios pasaulio finansinės institucijos bei šalių centriniai bankai ėmė iš esmės griežtinti rinkos rizikos valdymo procedūras. Kaip parodė vėlesnė minėtų kompanijų istorijos analiz�����, pagrindinė jų nuostolių ar bankrotų priežastis buvo nesugebėjimas tinkamai valdyti prekybinį portfelį įtakojančios rinkos rizikos. Prekybinio portfelio rizikos valdymas tampa vis aktualesnis ir Lietuvos finansinėms institucijoms, ypač investicinių bei pensijų fondų valdytojams, investuojantiems į užsienio vertybinius popierius, denominuotus kitomis valiutomis nei litas ar euras. / The scientific problem of the dissertation is search of adequacy of the trading portfolio risk management methods and models to the current economic, technological, and informational circumstances of financial institutions. The main features of science novelty characteristic to this research are the following: (i)the comparative study on Value at Risk estimation methods allowed to make important theoretic conclusion that selection of Value at Risk estimation methods depends mostly on characteristics of the portfolio under investigation; theoretic recommendations regarding selection of Value at Risk estimation methods were suggested as well; (ii)the comparative study on performance of volatility forecasting models allowed to make important theoretic conclusion that selection of Value at Risk estimation methods depends on characteristics of the data under investigation and selected criteria for assessment of forecasting accuracy; in the context of risk management, the priority was given to operational rather than statistical accuracy assessment techniques in the context of risk management; (iii)the comparative study on risk adjustment measures allowed making important theoretic conclusion that selection of risk adjustment measures depends mostly on characteristics of the portfolio under investigation; theoretic recommendations regarding selection of risk adjustment measures were suggested as well.

The scientific problem of the dissertation is search of adequacy of the trading portfolio risk management methods and models to the current economic, technological, and informational circumstances of financial institutions. The main features of science novelty characteristic to this research are the following: (i)the comparative study on Value at Risk estimation methods allowed to make important theoretic conclusion that selection of Value at Risk estimation methods depends mostly on characteristics of the portfolio under investigation; theoretic recommendations regarding selection of Value at Risk estimation methods were suggested as well; (ii) the comparative study on performance of volatility forecasting models allowed to make important theoretic conclusion that selection of Value at Risk estimation methods depends on characteristics of the data under investigation and selected criteria for assessment of forecasting accuracy; in the context of risk management, the priority was given to operational rather than statistical accuracy assessment techniques in the context of risk management; (iii)the comparative study on risk adjustment measures allowed making important theoretic conclusion that selection of risk adjustment measures depends mostly on characteristics of the portfolio under investigation; theoretic recommendations regarding selection of risk adjustment measures were suggested as well. / Sparčiai kintant finansinių institucijų veiklos sąlygoms, didėjant finansinių rinkų nepastovumui bei apimčiai, atsirandant vis naujoms finansinėms priemonėms, o kartu su jomis ir naujoms finansinių institucijų rizikos rūšims, ypač išaugo prekybinio portfelio rizikos valdymo poreikis. Prekybinio portfelio rizikos valdymas tampa vis aktualesnis ir Lietuvos finansinėms institucijoms, ypač investicinių bei pensijų fondų valdytojams, investuojantiems į užsienio vertybinius popierius, denominuotus kitomis valiutomis nei litai ar eurai. Prekybinio portfelio rizikos valdymas yra labai aktuali tema ir atskirų šalių centriniams bankams bei tarptautinėms finansų sistemos priežiūros institucijoms.

In this masterthesis I have rewied basic approaches to volatility estimating. These approaches are based on classical and Bayesian statistics. I have applied the volatility models for the purpose of volatility forecasting of a different foreign exchange (EURUSD, GBPUSD and CZKEUR) in the different period (from a second period to a day period). I formulate the models EWMA, GARCH, EGARCH, IGARCH, GJRGARCH, jump diffuison with constant volatility and jump diffusion model with stochastic volatility. I also proposed an MCMC algorithm in order to estimate the Bayesian models. All the models we estimated as univariate models. I compared the models according to Mincer Zarnowitz regression. The most successfull model is the jump diffusion model with a stochastic volatility. On the second place they were the GJR- GARCH model and the jump diffusion model with a constant volatility. But the jump diffusion model with a constat volatilit provided much more overvalued results.The rest of the models were even worse. From the rest the IGARCH model is the best but provided undervalued results. All these findings correspond with R squared coefficient.

The thesis focuses on modelling and forecasting the exchange rate time series volatility. The basic approach used for the conditional variance modelling are class (G)ARCH models and their variations. Modelling of the conditional mean is based on the use of AR autoregressive models. Due to the breach of one of the basic assumption of the models (normality assumption), an important part of the work is a detailed analysis of unconditional distribution of returns enabling the selection of a suitable distributional assumption of error terms of (G)ARCH models. The use of leptokurtic distribution assumption leads to a major improvement of volatility forecasting compared to normal distribution. In regard to this fact, the often applied GED and the Student's t distributions represent the key-stones of this work. In addition, the less known distributions are applied in the work, e.g. the Johnson's SU and the normal Inverse Gaussian Distribution. To model volatility, a great number of linear and non-linear models have been tested. Linear models are represented by ARCH, GARCH, GARCH in mean, integrated GARCH, fractionally integrated GARCH and HYGARCH. In the event of the presence of the leverage effect, non-linear EGARCH, GJR-GARCH, APARCH and FIEGARCH models are applied. Using suitable models according to the selected criteria, volatility forecasts are made with different long-term and short-term forecasting horizons. Outcomes of traditional approaches using parametric models (G)ARCH are compared with semi-parametric neural networks based concepts that are widely applicable in clustering and also in time series prediction problems. In conclusion, a description is given of the coincident and different properties of the analyzed exchange rate time series. The author further summarized the models that provide the best forecasts of volatility behaviour of the selected time series, including recommendations for their modelling. Such models can be further used to measure market risk rate by the Value at Risk method or in future price estimating where future volatility is inevitable prerequisite for the interval forecasts.

Le phénomène de contagion, les hypothèses d'efficience et les transferts de volatilité sont parmi les théories économiques les plus importantes, car elles fournissent une vision globale sur la stabilité financière. Or, elles restent les moins comprises depuis les récentes crises récentes. Ainsi, cette thèse propose de fournir aux régulateurs économiques, aux investisseurs et aux acteurs du marché financier une vision actualisée du comportement dynamique des marchés mondiaux des Credit Default Swaps (CDS): efficacité informationnelle, interaction avec d'autres marchés financiers internationaux et exposition au risque systémique. La dynamique en constante mutation de ces marchés associée à l'évolution constante des politiques de réglementation a suscité un enthousiasme mondial pour l'étude comportementale des marchés des CDS, auquel nous contribuons à travers cinq essais interconnectés. Nous discutons, dans le premier essai, les faits stylisés des données des CDS souverains à travers l'estimation de 9 modèles de type GARCH. Ce chapitre compare les performances de plusieurs modèles prédictifs de volatilité linéaire et non linéaire et prenant en compte différentes caractéristiques financières des séries statistiques. L'application de ces modèles aux spreads de CDS de 38 pays révèle que le pouvoir prédictif de ces modèles dépend de leur capacité à capturer les faits stylisés des CDS souverains pendant l'estimation du processus de la variance. En effet, les modèles GARCH fractionnellement intégrés surpassent les modèles GARCH de base en termes de prévision, en raison de la flexibilité accordée au degré de persistance des chocs de variance. Ces résultats sont utilisés pour modéliser conjointement les rendements et la volatilité des spreads de CDS dans l'ensemble des prochains essais. Le deuxième essai examine également les caractéristiques financières des marchés internationaux des CDS souverains, en donnant de nouvelles preuves sur leurs degrés d'efficacité. En utilisant un nouveau cadre économétrique basé sur une estimation du modèle VECM-FIGARCH en trois étapes, nous montrons que les informations contenues dans les spreads de CDS et les rendements des obligations sous-jacentes ne sont pas toujours reflétées instantanément et correctement dans le niveau du risque souverain. Les résultats révèlent l'existence d'opportunités d'arbitrage avec un rejet partiel de l'hypothèse de marche au hasard dans plusieurs des 37 pays étudiés [etc...] / Contagion phenomenon, efficiency hypothesis and spillover effects are amongst the most important economic theories as they provide an overall vision of the financial stability, yet the least understood in the aftermath of the recent crises. This thesis proposes to provide policy makers, investors and broadly market participants with an updated outlook of the dynamic behavior of the global sovereign Credit Default Swaps (CDS) markets: informational efficiency, interaction with other international financial markets and systemic-risk exposure. The steadily changing dynamics of these markets combined with the constantly evolving regulatory policies have led to a shared worldwide enthusiasm regarding the behavioral study of CDS markets, in which we contribute through five interconnected essays. We first discuss, in the first essay, the statistical characteristics of the sovereign CDS data, through the estimation of 9 GARCH-class models. This chapter compares the predictability performances of several linear and non-linear volatility models taking into consideration different financial stylized facts. Application on CDS spreads of 38 countries reveals that the forecasting power of these models depends on their ability to capture sovereign CDS features while estimating the variance process. Yet, the fractionally-integrated models outperform the basic GARCH-class models due to the allowed flexibility regarding the persistence degree of the variance shocks. These results are used to jointly model returns and volatility of CDS spreads in the forthcoming essays.The second essay also investigates the financial characteristics of the international sovereign CDS markets, by giving new evidences on their efficiency degrees. Using a new framework based on a 3-step estimation of a VECM-FIGARCH model, we show that information contained in CDS spreads and bond yields are not always instantaneously and properly reflected in the current sovereign risk level. Results reveal the existence of arbitrage opportunities with a partial rejection of the randomness hypothesis in some of the 37 studied countries. While the previous essay used the conditional expectation of CDS spreads to study the market behavior, the next essays rather focus on the properties of the variance and covariance. The predictability of sovereign CDS volatility, based on the information contained in some country-specific and global macroeconomic factors, is investigated in the third chapter [etc...]