Abstract
The PSE noted incredible increase in both trading volumes and prices of traded stocks during last five years. The PX index (former PX-50) reached the level of 1600 points at the end of 2006, which is almost four times higher than in 2001.
Cointegration analysis can show us if the growth has been driven by some hidden common factor(-s), either optimistic perception of the market or common fundamentals, or if the main forces have been in case of each stock individual and specific and the fact, that the increase was similar among many of stocks is only due to coincidence. We have found that the results differ substantially upon the choice of frequencies of the data.
The interrelations are very small when using daily data, on the other hand weekly data lead to opposite result. Furthermore the analysis of daily data implies that the relations became closer in the long term and that they are almost negligible during the period of high growth (2005-2006), but again the weekly data showed the opposite.
As far as results of the VECM and VAR estimates concern, they were surprisingly good: using weekly data we were able to explain up to 80% of variance in stock returns comparing to 20% with daily data. This difference can be explained partly as a consequence of high noise in daily data and excessive volatility on emerging markets.