3.2. Regression Results
After estimating the cost and profit efficiency scores based on stochastic frontier method, in
order to investigate the relationship between stock returns and efficiency, the stock returns (CASR) are
regressed against corresponding annual change in efficiency estimates while controlling for
controlling for size and risk using the annual percentage change in total assets (TA) and the annual
percentage change in equity to assets (EQAS). The regression results, derived from equation 4, are
reported in Table 5. Since an unbalanced panel data set is used to examine this relationship in these
transition countries, there are several types of panel data models including constant coefficients
model 17 , fixed effects model and random effects model. To determine the choice of appropriate
methodology, several specification tests are empoyed such as likelihood ratio, Breusche-Pagan
Lagrange Multiplier (LM) test, and Hausman’s test. In both cases, the test results indicate that the
fixed effects model should be preferred 18. Year dummy variables are included into the model to take
into account the potential time effects in the stock returns. The models are estimated also using
White’s transformation to control for cross-section heteroscedasticity with corrected degrees of
freedom.