The Semantics of Change as Revealed Through an Examination of Financial Accounting Standards Amendments

Financial accounting standards (FASs) are frequently changed by formal amendment. Frequent amendment imposes costs upon the users of these standards. Research into amendments and the amendment process is strikingly absent. A better understanding of what is amended and how often such amendments occur may facilitate a reduction in the frequency of such amendments enabling a corresponding reduction in the associated costs. This study examines a sample of 567 amendments to FASs. Classical information retrieval techniques are applied to establish whether meaningfully related clusters of amendments can be identified. Identified clusters are compared to a descriptive taxonomy of change (FAS amendments) proposed by Fisher (2004) to determine whether the clusters fit within the framework of the taxonomy, thereby lending it support. The results demonstrate the ability of classical information retrieval techniques to identify significant clusters of amendments and extend the applicability of the techniques into the FAS domain. In addition, the identified clusters provide preliminary support for the proposed taxonomy of change. We propose future research using amendments inserted into the original pronouncements in order to provide the context for the amendments. The additional context should enhance the ability of latent semantic indexing to identify meaningful clusters and may provide stronger support for the taxonomy.