The compatibility of market returns and income numbers observed after adjustments to retained earnings

Policy makers have long been concerned that adjustments to retained earnings might give rise to unwarranted differences across income numbers of adjusting and nonadjusting firms. Researchers of financial markets, meanwhile, have generated a body of evidence showing that markets set security prices using an information set broader than, and superior to, accounting measurements of net income. This evidence suggests that markets are capable of pricing securities efficiently even if adjustments to retained earnings have given rise to unwarranted differences across income numbers. This dissertation compared cumulative market returns observed in the two years after adjustments to retained earnings with contemporary changes in net income that were not attributable to economy and industry effects. Independent comparisons were performed with 1,026 observations associated with seven types of adjustments. On the assumption that markets do price securities efficiently, the comparisons may provide a basis for judging, normatively, the appropriateness of post-adjustment income numbers. These comparisons provided evidence of (1) full compatibility following adjustments for compensated absences and for SFAS No. 52 foreign currency translation, (2) partial compatibility following adjustments for oil and gas exploration and events of prior period, and (3) incompatibility following adjustments for lease commitments, poolings of interests, and SFAS No. 8 foreign currency translation. The results of these comparisons appeared to support three major conclusions: First, the security returns generated by efficient financial markets may be useful as a basis for evaluating the relevance of alternative measurements of net income. Second, compatibility between market returns and income numbers probably cannot be achieved until accounting policy makers adopt the concept of comprehensive income. Third, the poolings of interests method of accounting for certain business combinations may create a conflict between interests of managers and shareholders, and that conflict may be sufficiently serious as to justify the method\u27s termination