| Accounting valuation models have been widely studied by researchers and commonly used by practitioners. Almost all the accounting valuation models require earnings as one of the inputs (or the only input). However, one question has not been addressed: whether earnings of a longer interval or earnings of a shorter interval should be used in the valuation models. The fact that earnings can be aggregated over time and this intertemporal aggregated earnings contains fewer measurement errors is intrinsic to accounting.; Although earnings aggregation is intrinsic to accounting, it has received little attention by researchers when using accounting valuation models to estimate variables of interest. Also, although there are few studies that examine the effect of earnings aggregation, these studies focus mainly on the contemporaneous explanatory power of earnings for returns. Therefore, the effect of earnings aggregation on inferring prices via accounting valuation models remains unclear, and simply aggregating earnings over a longer interval may potentially improve the estimates from the valuation models.; Despite the fact that the results for the cross-sectional sample may not be encouraging, the results for the sub-samples support the expectation that earnings aggregation improves the ability of the valuation model to infer prices for some types of firms. For firms with negative earnings and for small firms, using aggregated earnings of a longer interval in the valuation model generally generates smaller errors in inferring prices than using annual earnings, and the differences between the errors can be significant.; These results contribute to the understanding of the fundamental accounting attribute of earnings aggregation. More specifically, this study contributes to the valuation research insofar as the results show that it is beneficial to aggregate earnings over a longer interval when applying accounting valuation models for some specific types of firms. |
