Panteion University of Social and Political Sciences
Volatility modelling of financial markets
Abstract
dc:descriptionForecasting oil price volatility is considered of major importance for numerous stakeholders, including, policy makers, industries and investors. The first study (Chapter 2) examines and evaluates the main factors that oil price volatility forecasters should consider before constructing their forecasting models. Such factors are related to: i) direct vs iterated forecasts, ii) the incorporation of continuous and jump components, iii) the importance of semi variance volatility measures, and iv) OLS vs time-varying parameter (TVP) estimation procedures. We evaluate the performance of these factors for both the realized and implied volatility measures of the WTI crude oil price, based on statistical loss functions, as well as, their economic use. The results show that depending on whether end-users are interested in forecasting the realized or the implied volatility, the factors influencing the accuracy of forecasts are different.In the third Chapter, we enhance the modelling framework by incorporating exogenous information in the proposed models. Nowadays, it is noteworthy the fact that the global uncertainty plays a major role to the economic outlook and more specifically the financial and energy markets. In this study (Chapter 3), we focus on the impact of the various uncertainty factors on the oil price volatility, which is considered crucial not only for the global economy but also for the financial markets because of its financialization. However, uncertainty can be captured by different factors, which provide dissimilar information to oil price volatility. We categorize those factors to the following classes: Implied volatility indices, financial stress indices and other indicators related to the uncertainty environment, such as economic policy uncertainty, geopolitical risk and business conditions. Our main findings provide strong evidence that taking uncertainty indicators into account enhances the predictive accuracy of oil price volatility at all forecasting horizons. Moreover, the results indicate that the Dynamic Model Averaging (DMA) is considered significant for forecasting oil price volatility by combining the different indicators of the three classes and giving the corresponding weight to the model.However, crude oil investors would be interested not only in maximizing their profits but also in minimizing the risk of their portfolios, which could be managed by hedging the crude oil portfolio in an efficient way. In this regard, the existing literature has studied the interrelations between crude oil and other asset classes, including stock, foreign exchange markets and market reflecting macroeconomic conditions focusing mainly on their returns. In this study (Chapter 4), we concentrate on the time-varying correlation of the volatility measures of crude oil and three asset classes using a dynamic conditional correlation (DCC) model. The main objective of this study is to examine the optimal portfolio weights, constructed by the variance-covariance matrix, for portfolios comprised of the aforementioned volatility measures and to identify whether the investors could benefit from the interactions of the WTI crude oil and the three assets. The aim of this study is to focus on volatility and not on returns, since investors and academics concentrate their attention to the volatility of crude oil recent years. The results of the correlations indicate a time-varying behavior, which gives a signal to investors that they have to re-balance their portfolios regularly in order to minimize their portfolios' risk. Finally, the findings show that the asset that offers higher opportunities for hedging the WTI crude oil volatility is that of the U.S. T-bills, which represents the market related to macroeconomic conditions.Finally, Chapter 5 aims to investigate the predictive information of the daily crude oil realized volatility on the U.S. economy. More specifically, oil price volatility has attracted the attention of the academic community because of its crucial impact on the economic outlook. However, there is a gap in literature with regard to the impact of crude oil price volatility on special aggregates of industrial production. In this study (Chapter 5), we propose MIDAS models including different daily crude oil realized volatility measures in order to investigate their impact in an out-of-sample analysis. For comparison reasons, we study the effect not only of crude oil but also of other assets' realized volatility, such as S&P500 index, U.S. dollar index and U.S. T-bills. Moreover, we set a group of monthly macroeconomic, oil-related and uncertainty-related variables as predictors of the industrial production aggregates in order to evaluate whether the impact of the daily realized volatility measures is significant or the monthly predictors themselves offer adequate information. The results show that crude oil realized positive semivariance can definitely provide higher forecasting performance in the models used for producing energy-related industrial production measures, which is not the case for the non energy-related ones.
Degree
thesis:*- Grantor dc:publisher
- Panteion University of Social and Political Sciences
- Year dc:date
- 2022
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Delis, Panagiotis
- Δελής, Παναγιώτης
Subjects
dc:subject × 12Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/52479