On Friday 16 March Kristofer Månsson successfully defended his doctoral thesis in Statistics at Jönköping International Business School (JIBS).
Kristofer Månsson has conducted his research at the Department of Economics, Finance and Statistics at JIBS and Sparbankernas Forskningsstiftelse has financed his dissertation work.
The doctoral thesis is entitled “Issues of multicollinearity and conditional heteroscedasticy in time series econometrics”. In its later parts the thesis shows the development of different shrinkage estimators for count data models, which may be used when the explanatory variables are highly inter-correlated, which is illustrated by a real life example that analyses the effect of different types of vehicles on the number of accidents in different Swedish counties. In this example the usage of different types of vehicles are highly inter-correlated.
The thesis deals with the field of time series econometrics, the development of robust methods when testing for Granger causality in the presence of generalized autoregressive conditional heteroscedasticy (GARCH) and causality-in-variance (i.e. spill over) effects.
GARCH errors mean that the variance of the error term in a statistical model in a certain time period depends on the error in the previous time period. Hence, large (small) errors tend to be followed by large (small) errors. Spillover effects mean that a large (small) error in one time series also may lead to a large (small) error in another time series. The presence of GARCH and spill over effects may lead to an over-rejection of true null hypothesis. Kristofer Månsson develops some methods that are robust to GARCH and spillover effects.
Faculty opponent at the defence was Professor Sneh Gulati from Florida International University, USA. The Examining committee consisted of Associate Professor Astrid Hilbert, Linnaeus University, Associate Professor Abdullah Almasri, Karlstad University and Professor Per Olof Bjuggren, JIBS. Chairperson at the defence was Main tutor professor Ghazi Shukur, JIBS.
036-10 17 64