ANALYSIS OF THE AVERAGE SHARE PRICE OF COMPANIES LISTED
ON BSE DEPENDING ON THE PROFIT AND EXCHANGE SEGMENT.
DIFFERENT TECHNIQUES OF GENERAL LEAST SQUARE AND
COMPUTING COEFFICIENT COVARIANCE FOR MEAN PRICE
EQUATION ESTIMATION
Teodor Hada, “Bogdan Vodă” University of Cluj Napoca
Emil Olteanu, 1 Decembrie 1918” University of Alba Iulia
Iulian Bogdan Dobra, 1 Decembrie 1918” University of Alba Iulia

ABSTRACT: The main purpose in our scientific approach is the analysis of the evolution of
the share price of Romanian companies listed on the Bucharest Stock Exchange in 2007-2013, and
how this reacts to the registered profit. The forecast evaluation details for adjusted sample 20082013 indicated that the Bias and Variance Proportions are small, which implies that the error of
prediction is concentrated in covariance proportion and shows that the forecast is quite
“acceptable”. The Theil Inequality Coefficient gives an acceptable indicator in measuring the “fit”
of the model than the Mean Absolute Percentage Error. Finally the comparison between real data
and the forecast outlined the fact that the PriceF reacts abnormally. Therefore analysis shows that if
Profit increases, the stock price falls (i.e. Criterion 4 - PriceF&Profit) in almost 48% cases,
compared with the real situation of about 24% cases (i.e. Criterion 4 - Price&Profit). In conclusion,
even if we managed to translate the low prices of financial-economic crisis period into “normal
capital market time”, using the model, one may notice that in 2009-2013 the cases number of shares
that recorded higher prices grow up to 113 instead of 77 (i.e. Appendix C). Yet if we look at the
evolution of Price&Profit / PriceF&Profit tandem we note that the profit recorded by the company
is not a strength factor for the dynamic average share price of the Romanian listed companies.
Keywords: share price, profit, GLS Weights, robust coefficient covariances, ANCOVA
JEL code: G11, G12, G14