The Impact of IFRS Adoption on Financial Reporting Quality and Consistency

Abstract

This research presents a novel, cross-disciplinary investigation into the impact of International Financial Reporting Standards (IFRS) adoption by applying computational linguistics and network analysis techniques to financial disclosures, a methodological departure
from traditional accounting research. We move beyond conventional econometric analyses
of accounting metrics to examine the textual and structural properties of financial reports
themselves as direct indicators of reporting quality and consistency. By constructing a
unique corpus of 15,000 annual reports from 2,500 publicly listed firms across 15 jurisdictions over a 12-year period (2010-2022), we quantify reporting quality through measures
of textual complexity, semantic coherence, and disclosure network density. We measure
consistency through the temporal stability of these textual features and the convergence of
reporting structures across firms within adopting countries. Our findings reveal a significant, non-linear relationship between IFRS adoption and textual reporting quality: an initial
period of increased syntactic complexity and reduced readability (a ’disruption phase’) is
followed by a subsequent convergence towards more standardized, coherent, and less obfuscated narrative structures. Furthermore, we identify that the consistency of financial
reporting, measured as the reduced variance in textual feature vectors across peer firms,
improves markedly post-adoption, but this effect is strongly moderated by the strength of a
country’s legal enforcement regime. The most original contribution lies in our application of
a bio-inspired optimization algorithm—an Ant Colony Optimization variant—to model the
path of reporting convergence, treating disclosure elements as nodes in a graph that ’ants’
(representing reporting firms) traverse to find optimal, standardized pathways. This model
successfully predicts the time-to-convergence for different jurisdictions with high accuracy.
Our results challenge the binary view of IFRS impact, proposing instead a phased, evolutionary model of reporting quality transformation driven by complex adaptive learning within
financial reporting ecosystems. This research establishes a new paradigm for assessing accounting standards by leveraging computational text analysis and complex systems theory,
offering regulators and auditors a suite of novel, real-time diagnostic tools for monitoring
reporting quality beyond numerical compliance.