Essays on the return, volatility and correlation features of emerging market bonds and related models

In this thesis, Emerging Market Local Bonds are examined from the portfolio risk perspective and certain methodologies on which the literature is limited are dealt with. MFDFA (Multifractal Detrended Fluctuation Analysis) and MFDXA (Multifractal Detrended Cross-Correlation Analysis) methods are applied to all portfolio components; the return, volatility and cross-correlation data of short-term and long-term bonds of Brazil, Indonesia, India, South Africa, Mexico, Turkey. The findings of this study have significant implications for the asset management industry as it paves the way for the improvement of forecast and risk measurement techniques by considering fractal-based models. In the second part, we focus on Turkish local fixed income as-sets by stylizing tools for the risk analysis. To investigate and model, volatility clustering, asymmetry-leverage impacts, and especially the volatility persistence several fractional models: FIGARCH, HYGARCH, FIEGARCH, and FIAPARCH are ap-plied. It is revealed that, under two different market conditions, in-sample analysis and the forecasting performances of fractional models deliver superior outcomes. In the third part of the thesis, Emerging Market Bond portfolios are evaluated with the fractional models. For different versions of Emerging Market Bond portfolios, Dynamic Conditional Correlation (DCC) GARCH and FIGARCH models are applied. Performances are demonstrated with Value at Risk (VaR) analysis. Portfolio analysis covers dynamic portfolios that consider portfolio optimizations. This thesis as a whole, provide specific takeaways by suggesting the applications of multifractal analysis and fractional models to Emerging Market Bond portfolios.