Choice of GAAP: Evidence from the Group and Single Entity Accounts of UK Firms – A Machine Learning Approach
56 Pages Posted: 29 Jul 2022
Date Written: July 17, 2022
Abstract
We examine firms’ choice of the extent of IFRS adoption in the mandatory IFRS-implementation period, defined with respect to reporting standards used in financial statements. We examine this issue using UK firms’ presentation of both group and single entity statements. We classify firms using IFRS for both as full adopters, firms using IFRS for group but not for parent-level statements as partial adopters, and firms using UK GAAP as non-adopters. We implement a Least Absolute Shrinkage and Selection Operator (LASSO) approach to select proxies for category-specific models of the extent of IFRS reporting, and compare those models across the three categories of firms. Models with between six and 22 variables (out of 42 candidates) provide excellent explanatory power for the models of choice between different reporting options. We also find that these models for partial adopters are more similar to those for full adopters than to those for non-adopters. Finally, using several proxies for stock price informativeness in a difference in differences framework, we find an increase in synchronicity for sample firms from pre- to post-2005, consistent with increased global market synchronicity after IFRS adoption. We find that non-adopters had larger increases in information quality than did either full or partial adopters, and that full adopters had larger increases in information quality than did partial adopters. These findings are consistent with firms choosing the extent of commitment to IFRS to satisfy investors’ information demands rather than managerial opportunism.
Keywords: IFRS adoption, Reporting Incentives, Label Adoption, Parent Company Reporting, Group Reporting, Financial Companies, Accounting in the UK
JEL Classification: M41, M48, B41
Suggested Citation: Suggested Citation