Please use this identifier to cite or link to this item: http://hdl.handle.net/1946/3612
In this paper we investigate the predictability of asset returns, utilizing daily data for an Icelandic Government bond, covering a five year period from May 2004 to August 2009. In the first part we examine whether some simple forms of technical analysis can predict price movements. By comparing the unconditional empirical distribution of daily returns to the conditional distribution, conditioned on specific technical indicators, we find some support for the profitability of these technical trading rules. Nevertheless, reported gains may not be high enough to translate into profits after transaction costs are considered. Rules combining technical analysis and popular time series models are also examined.
In the second part, weekly observations of a few key macroeconomic variables are used to predict bond price changes. Using data up to the global financial crisis for model estimation to use in the post-crisis test period proved to be unsuccessful. On the other hand, combining the macroeconomic variables with technical analysis improved some of the result from the first part and might be of some practical value.