Table 3. There is evidence for three sub-periods that investors get reward for skewness as predicted by
Kraus-Litzenberger theory. On other hand cokurtosis seems to affect expected return in limited way.
The pattern of covariance-risk premium remains strange and insignificantly different from zero. The
empirical tests made for Brazilian market by Attayde and Flores Jr (2000) and for Athens stock market
by Messis et al (2007) come up with same conclusion as shown by our results, that the skewness play
the most important role, while the gain of adding kurtosis is negligible. Hung and Xu (2003) on the
contrary find limited evidence for the existence of higher order pricing factor for UK.