A Measure of Firm Complexity: Data and Code

Abstract

ABSTRACT We propose that firm complexity is best measured with accounting disclosures. Accounting is the “language of business,” and accounting disclosures of most business activities are mandated. Therefore, relying on accounting disclosures is the best approach for consistently capturing a wide range of firm activities for a large cross-section of firms. Measuring firm complexity is important for many applications in research and practice. However, firm complexity is multifaceted, making it difficult to measure. We review past research on complexity and motivate the use of Accounting Reporting Complexity (ARC), proposed by R. Hoitash and U. Hoitash (2018), to measure firm complexity. In so doing, we discuss the advantages of ARC over other measures. We then review studies that use ARC and provide a detailed description and code to construct ARC (and related measures) based on publicly available data. The complete ARC dataset is also available for download at: https://www.xbrlresearch.com/. Data Availability: Data are publicly available from sources identified in the paper. ARC is based on XBRL filings downloaded directly from the Securities and Exchange Commission and is available for download at: https://www.xbrlresearch.com/. JEL Classifications: B40; D20; C10; G10; L25; M40.