This thesis performs portfolio optimization using three allocation methods, Certainty Equivalence Tangency (CET), Global Minimum Variance (GMV) and Minimum Conditional Value-at-Risk (MinCVaR). We estimate expected returns and covariance matrices based on 7 stock market indices with a DCC-GARCH model including an ARMA (1.1) process and an external regressor of an implied volatility index (VIX). We then simulate returns using a rolling window of 500 daily observations and construct portfolios based on the allocation methods. The results suggest that the model can sufficiently estimate expected returns and covariance matrices and we can outperform benchmarks in form of equally weighted and historical portfolios in terms of higher returns and lower risk. Over the whole out-of-sample period the CET portfolio yields the highest mean returns and GMV and MinCVaR can significantly lower the variance. The inclusion of VIX has marginal effects on the forecasting accuracy and it seems to impair the estimation of risk.

This study investigates the return generating process underlying the South African stock market. The investigation of the return generating process is framed within the Arbitrage Pricing Theory (APT) framework with the APT reinterpreted so as to provide a conceptual framework within which the return generating process can be investigated. In modelling the return generating process, the properties of South African stock returns are taken into consideration and an appropriate econometric framework in the form of Autoregressive Conditional Heteroscedastic (ARCH) and Generalized Autoregressive Conditional Heteroscedastic (GARCH) models is applied. Results indicate that the return generating process of South African stock returns is described by innovations in multiple risk factors representative of several risk categories. The multifactor model of the return generating process explains a substantial amount of variation in South African stock returns and the ARCH/GARCH methodology is an appropriate econometric framework for the estimation of models of the return generating process. The APT framework is successfully applied to model and investigate the return generating process of South African stock returns.

This thesis investigates if Environmental, Social and Governance (ESG) investments can be considered as an independent asset class. As ESG and responsible investing has increased substantially in recent years, responsible investments have entered the portfolios with other asset classes too. Therefore, there is a need in studying ESG investment properties with other financial asset classes. By collecting daily price data from October 2007 to December 2018, we research the directionalities between ESG, ethical, conventional, commodities and currency. Initially, we employed a MODWT, multiscale investment horizon wavelet analysis transformation of the data. The decomposed wavelet data is then applied in pairwise linear and non-linear Granger causality estimations to study the directionality relationships dependent on investment horizon. Additionally, econometric filtering processes have been employed to study the effects of volatility on directionality relationships. The results mainly suggest significant directionality relationships between ESG and the other asset classes. On the medium-term investment horizon, almost all estimations indicate strict bidirectionality. Thus, on the medium-term, ESG can be said to be integrated with the other asset classes. For the long-term horizon, most relationships are still predominantly bidirectional between ESG and all other asset classes. The biggest differences are found on the short-term horizon, with no directionality found between ESG and commodities that cannot be explained by volatility. Furthermore, most directionality relationships also disappear when controlling for the volatility transmission between ESG and currency on the short-term horizon. Thus, our findings suggest significantly more integration between ESG and ethical and conventional as bidirectionality overwhelmingly prevails regardless of investment horizon. As previous research has found similarities between ethical and conventional as well as ESG having similar characteristics to commodities as conventional and ethical, we suggest that ESG should be considered as being integrated and having strong similarities with other equities. Thus, it should be treated as being part of the conventional equity asset class. Deviations from bidirectionality could be caused by ESG variable specific heterogeneity. However, despite our rejection of ESG as an independent asset class, it still carries significant potential as it excludes firms with climate-harming practices, thereby helping in combating climate-related as well as social and governance issues the world is facing.

Submitted by Caroline Xavier (caroline.xavier@pucrs.br) on 2017-11-03T12:04:14Z No. of bitstreams: 1 DIS_ANDRE_LOCATELLI_COMPLETO.pdf: 771690 bytes, checksum: fe8fbda3561c48ca1cee72f699327cba (MD5) / Approved for entry into archive by Caroline Xavier (caroline.xavier@pucrs.br) on 2017-11-03T12:04:27Z (GMT) No. of bitstreams: 1 DIS_ANDRE_LOCATELLI_COMPLETO.pdf: 771690 bytes, checksum: fe8fbda3561c48ca1cee72f699327cba (MD5) / Made available in DSpace on 2017-11-03T12:04:36Z (GMT). No. of bitstreams: 1 DIS_ANDRE_LOCATELLI_COMPLETO.pdf: 771690 bytes, checksum: fe8fbda3561c48ca1cee72f699327cba (MD5) Previous issue date: 2017-08-24 / The present dissertation aims, through the theories of economic political cycles, to investigate if they influence the returns and volatility of the Ibovespa, index of the S?o Paulo Stock Exchange. The four main theories dealing with the theme, Traditional Party Theory, Traditional Opportunist Theory, Rational Party Theory and Opportunistic Rational Theory will be addressed. The data used will be the Ibovespa daily returns and the daily returns of the S & P 500, one of the main indices of the North American stock market and that will serve to capture the changes of the external stock market. In order to calculate the influence of economic policy cycles on the returns and volatility of the Ibovespa, the ARCH and GARCH econometric models have been used, which have been widely used in such works and have been shown to be consistent in the estimation of time series. The ARCH model had better results for the estimated model. Four different Dummy variables, each representing a different time period, were tested to determine whether economic policy cycles influenced Ibovespa returns and volatility in those periods. At the 5% significance level, abnormal returns in the periods included in the Dummy variables were not found nor was there statistically significant change in variance in the same periods. At a significance level of 10%, the influence of economic policy cycles on the volatility of the Ibovespa in the period of 180 days, ranging from 12 months to 6 months before the presidential elections, was found. / A presente disserta??o tem como objetivo, atrav?s das teorias de ciclos pol?ticos econ?micos, investigar se os mesmos influenciam nos retornos e na volatilidade do Ibovespa, ?ndice da bolsa de S?o Paulo. Ser?o abordadas as quatro principais teorias que tratam sobre o tema, Teoria Partid?ria Tradicional, Teoria Oportunista Tradicional, Teoria Partid?ria Racional e Teoria Oportunista Racional. Os dados utilizados ser?o os retornos di?rios do Ibovespa e os retornos di?rios do S&P 500, um dos principais ?ndices do mercado acion?rio norte americano e que servir? para captar as mudan?as do mercado acion?rio externo. Para calcular a influ?ncia dos ciclos pol?ticos econ?micos sobre os retornos e a volatilidade do Ibovespa foram utilizados os modelos econom?tricos ARCH e GARCH, que t?m sido amplamente utilizados em trabalhos dessa natureza, e que t?m se demonstrado consistentes na estima??o de s?ries temporais. O modelo ARCH teve melhores resultados para o modelo estimado. Foram testadas quatro diferentes vari?veis Dummy, cada uma representando um per?odo de tempo diferente, para calcular se os ciclos pol?ticos econ?micos influenciavam os retornos e a volatilidade do Ibovespa naqueles per?odos. N?o foram encontrados, ao n?vel de signific?ncia de 5%, retornos anormais nos per?odos englobados pelas vari?veis Dummy nem se observou altera??o da vari?ncia de forma estatisticamente significativa nos mesmos per?odos. A um n?vel de signific?ncia de 10% foi encontrado a influ?ncia dos ciclos pol?ticos econ?micos na volatilidade do Ibovespa no per?odo de 180 dias que compreende entre 12 meses e 6 meses antes das elei??es presidenciais.

Submitted by Sara Ribeiro (sara.ribeiro@ucb.br) on 2018-11-09T18:21:24Z No. of bitstreams: 1 JoseHenriquedeOliveiraVarandaDissertacao2018.pdf: 3107527 bytes, checksum: ea06abcabf1c014758cc880bcf0b0726 (MD5) / Approved for entry into archive by Sara Ribeiro (sara.ribeiro@ucb.br) on 2018-11-13T16:00:57Z (GMT) No. of bitstreams: 1 JoseHenriquedeOliveiraVarandaDissertacao2018.pdf: 3107527 bytes, checksum: ea06abcabf1c014758cc880bcf0b0726 (MD5) / Made available in DSpace on 2018-11-13T16:00:57Z (GMT). No. of bitstreams: 1 JoseHenriquedeOliveiraVarandaDissertacao2018.pdf: 3107527 bytes, checksum: ea06abcabf1c014758cc880bcf0b0726 (MD5) Previous issue date: 2018-07-02 / This study evaluates which modifications can restore the theoretical performance of dynamic asset allocation strategies that uses insurance and leverage, specifically those known as Constant Proportion Portfolio Insurance (CPPI), when confronted with realistic premises and scenarios. Simulations using GARCH models are applied to assess the effects of path dependency and volatility on those strategies and to evaluate how selected modifications mitigates those effects. These modifications are tested using the Farinelli- Tibilleti ratio and derivations, like de Upside Potential Ratio. As main finding, the modifications that mitigates path dependency can restore the theoretical performance of portfolio insurance with high significance, making those preferred strategies in relation to Buy-and-Hold (BH) or Constant-Mix (CM) for most investors in several scenarios. This work also presents a novel modification, adapted for the risk-free market in Brazil, that resulted in the best performing portfolio insurance strategy with great significance. / Este trabalho avalia quais modificações reestabelecem o desempenho teórico das estratégias dinâmicas de alocação de ativos com seguro e alavancagem, denominadas Constant Proportion Portfolio Insurance (CPPI), quando confrontadas com premissas e cenários realistas. São realizadas simulações de modelos da família GARCH, com parâmetros estimados do mercado, para exercitar os efeitos da dependência do caminho e da volatilidade nestas estratégias e avaliar como as modificações selecionadas ajudam a combate-los. A significância das modificações é testada pela medida Farinelli-Tibiletti, sobre tudo a combinação que resulta na razão Upside Potential, onde conclui-se que existem modificações significantes que são capazes de resgatar o desempenho teórico da estratégia CPPI, inclusive tornando-a preferível às estratégias clássicas Buy-and-Hold (BH) e Constant-Mix (CM) em certos cenários. Por fim, o trabalho apresenta uma modificação inovadora, derivada do ajuste à realidade do mercado brasileiro, que acabou por apresentar o maior nível de desempenho relativo do método CPPI, com elevada significância.

Title: Actuarial and Exposure-based Models for Hail Peril Author: Bc. Matúš Drobuliak Department: Department of Probability and Mathematical Statistics Supervisor: RNDr. Michal Pešta, Ph.D., Department of Probability and Mathe- matical Statistics Abstract: This thesis covers an introduction to catastrophe modelling and focuses on statistical methods for extreme events. This includes methods of estimating parameters of claim distribution with a focus on probability weighted moments estimation technique. Furthermore, times series modelling, skew t-distribution, and two model clustering techniques are examined as well. This is later utilised in the practical application part of this thesis, which uses real data provided by an insurance company operating in the Czech Republic. Probability distribution fitting of large claims caused by hailstorms and Monte Carlo simulation of future losses are demonstrated later. Keywords: Catastrophe modelling, Hail peril, Probability weighted moments, Extreme events, ARMA-GARCH, Monte Carlo simulation iii

本篇論文的目的，希望以台灣為出發點來看，台灣與國際股市間的互動性，所著重的指標仍以共變數與相關係數為主，但同時也考慮了相關係數具有序列相關的影響，希望藉由這種模型的設計能夠找出台灣與國際股市之間的互動性，進而加以應用。 本文探討台灣與美國 NYSE，台灣與日本 Nikkie 225，台灣與香港，台灣與韓國的股票市場間日相關係數與共變異數，採用兩變數 GARCH 模型，實證結果發現： 1，其實台灣與國際市場間的共變異數矩陣，不是為一常數，是具有時間數列的關係存在的，因而具有可預測性。 2，台灣與韓國股市之間的關連性很小，因為他們之間永久共變數與暫時共變數都不顯著異於零；而台灣與美國 NYSE、台灣與日本 Nikkie225、或台灣與香港恆生指數等，他們之間永久共變數與暫時共變數都顯著異於零，這表示台灣與這些國家門存在有相當的關係。 3，台灣與國際市場的條件日相關係數是具有正向的時間趨勢的。 4，此一兩變數的 GARCH 模型於樣本外的共變數矩陣之預估能力，相較於傳統的 CAPM 模型，較具有準確性，因而可以在國際投資方面，將資金更有效的國際市場上，獲得更好的投資績效。 / we examine the co-movements of equity returns in five major international markets by characterizing the time-varying cross-country covariances and correltions. Using a generalized positive definite multivariate GARCH model, we find that Taiwan and Korea stock markets have zero permanent and transitory covariance.The other pairs of markets examined display significant permanent and transitory covariance. We also find that ,while conditional correlations betweem returns are gernerally small, they change considerably over time. An event analysis suggests that basing diversification strategies on these conditional correlations is potentially beneficial.

The stochastic volatility of daily foreign exchange (FX) derivatives poses a number of risks for the international banking community. Settlement risk, liquidity risk and capital adequacy are just a few immediate concerns that arise from such volatility. This thesis examines the impact of close-out netting on minimising the stochastic volatility of inter-bank FX derivatives. The problem with close-out netting is that although it is a simple formula of taking the differences between two banks at one point in time, it is the stochastic and volatile nature of FX rates that makes measuring the full impact of netting difficult. The objective of this thesis is to establish a realistic international banking framework or modelling environment in which close-out netting can be scientifically applied and examined. Five international daily FX rates will be used as sufficient approximations for five international banks. A generalised autoregressive conditionally heteroschedastic (GARCH) modelling approach is adopted as a robust and rich FX volatility paradigm. Then through Monte Carlo simulation of the resulting fitted GARCH models, we generate the distributions -with and without close-out netting. The findings of this thesis are interesting, showing that close-out netting is far more than just a simple mathematical process. Netting surely does reduce each bank's exposure to FX volatility, however, its multivariate nature reveals some important results for banking risk research and bank analysts.

L'objectif de cette thèse est d'étudier le problème de la modélisation des changements de régime dans les modèles à corrélations conditionnelles dynamiques en nous intéressant plus particulièrement à l'approche Markov-switching. A la différence de l'approche standard basée sur le modèle à chaîne de Markov caché (HMM) de base, nous utilisons des extensions du modèle HMM provenant des modèles graphiques probabilistes. Cette discipline a en effet proposé de nombreuses dérivations du modèle de base permettant de modéliser des structures complexes. Cette thèse se situe donc à l'interface de deux disciplines: l'économétrie financière et les modèles graphiques probabilistes. Le premier essai présente un modèle construit à partir d'une structure hiérarchique cachée markovienne qui permet de définir différents niveaux de granularité pour les régimes. Il peut être vu comme un cas particulier du modèle RSDC (Regime Switching for Dynamic Correlations). Basé sur le HMM hiérarchique, notre modèle permet de capter des nuances de régimes qui sont ignorées par l'approche Markov-Switching classique. La seconde contribution propose une version Markov-switching du modèle DCC construite à partir du modèle HMM factorisé. Alors que l'approche Markov-switching classique suppose que les tous les éléments de la matrice de corrélation suivent la même dynamique, notre modèle permet à tous les éléments de la matrice de corrélation d'avoir leur propre dynamique de saut. Dans la dernière contribution, nous proposons un modèle DCC construit à partir d'un arbre de décision. L'objectif de cet arbre est de relier le niveau des volatilités individuelles avec le niveau des corrélations. Pour cela, nous utilisons un arbre de décision Markovien caché, qui est une extension de HMM.

<p>We discuss the Normal inverse Gaussian (NIG) distribution in modeling volatility in the financial markets. Refining the work of Barndorff-Nielsen (1997) and Andersson (2001), we introduce a new parameterization of the NIG distribution to build the GARCH(p,q)-NIG model. This new parameterization allows the model to be a strong GARCH in the sense of Drost and Nijman (1993). It also allows us to standardized the observed returns to be i.i.d., so that we can use standard inference methods when we evaluate the fit of the model.</p><p>We use the realized volatility (RV), calculated from intraday data, to standardize the returns of the ECU/USD foreign exchange rate. We show that normality cannot be rejected for the RV-standardized returns, i.e., the Mixture-of-Distributions Hypothesis (MDH) of Clark (1973) holds. {We build a link between the conditional RV and the conditional variance. This link allows us to use the conditional RV as a proxy for the conditional variance. We give an empirical justification of the GARCH-NIG model using this approximation.</p><p>In addition, we introduce a new General GARCH(p,q)-NIG model. This model has as special cases the Threshold-GARCH(p,q)-NIG model to model the leverage effect, the Absolute Value GARCH(p,q)-NIG model, to model conditional standard deviation, and the Threshold Absolute Value GARCH(p,q)-NIG model to model asymmetry in the conditional standard deviation. The properties of the maximum likelihood estimates of the parameters of the models are investigated in a simulation study.</p>