Work

Bond yield modelling and its application in the European Union

Year: 2013
Type: dissertation
Abstract: Forecasting crises has always been an interesting and important topic for econometricians \nor statisticians. Literature suggests that government bond yields can be a valid \nleading indicator for this purpose. This thesis uses government bond yields and applies \nvarious models to forecast the crisis which happened recently. \nChapter 2 investigates a model utilising the term structure of interest rates to predict \noutput growth and recession in the UK. In contrast to previous literature, information \nretrieved from the whole yield curve is used rather than just the yield spread. Using \ndi�erent methods, our models are found to outperform the yield spread models both in \nin-sample and out-of-sample forecasting. Notably, the B-spline fitting model is able to \nforecast the 2009-2010 recession. Moreover, the model with lag of growth shows great \nforecasting ability in out-of-sample output growth forecasting. In most cases, models \nbased on B-spline perform better than the ones based on the Diebold-Li framework. \nChapter 3 examines the existence of time series non-linearities in the real output \ngrowth / recession-term spread relationship. Vector Autoregression (VAR), Threshold \nVAR (TVAR), Structural break VAR (SBVAR), Structural break threshold VAR (SBTVAR) \nare applied in the analysis. The in-sample results indicate there are non-linear \ncomponents in this relationship. And this non-linearity tends to be caused by structural \nbreaks. The best in-sample model also shows its robustness on arrival of new information \nin the out-of-sample tests. Evidence shows the model with only structural break \nnon-linearity outperforms linear models in 1-quarter, 3-quarter and 4-quarter ahead forecasting. \nThe European sovereign debt crisis has become a very popular topic since late 2009. \nIn Chapter 4, the sovereign debt crisis is investigated by calculating the probabilities \nof the potential future crisis of 11 countries in the European Union. We use sovereign \n2 \n3 \nspreads of the European countries against Germany as targets and apply the GARCH \nbased vine copula simulation technique. The methodology solves the di�culties of calculating \nthe probabilities of rarely happening events and takes sovereign debt movement \ndependence, especially tail dependence, into consideration. Results indicate that Italy \nand Spain are the most likely next victims of the sovereign debt crisis, followed by Ireland, \nFrance and Belgium. The UK, Sweden and Denmark, which are outside EMU, are \nthe most financially stable countries in the sample.
Author: Dalu Zhang
Language: English
Open Access status: green
OpenAlex ID: W2067229071
Cites: 100
Cited by:
Related to: 9