Fördelning av resultat på den svenska börsen
This paper replicates a test performed by Burgstahler and Dichev (1997), which identifies a kink in the frequency distribution of earnings for companies listed on American stock exchanges. Their claim, that the kink is evidence of earnings management, has however been dismissed as a methodical consequence. To address this concern, this paper not only replicates the test in general, but also performs tests on a split sample, where founding family-owned companies are separated from the others, since earlier research has shown that founding family ownership is correlated with less earnings management. The results are in line with the findings of Burgstahler and Dichev, with a significant kink in the frequency distribution in the general test. The findings also show that the significance of the kink increases in the non-family-owned sample, while it disappears in the family-owned sample. This could indicate that the frequency distribution is a valid test of earnings management, hence this paper adds to the previous research that shows a correlation between founding family ownership and less earnings management.