Academic journal article Auditing: A Journal of Practice & Theory

The Effect of Misstatements on Decisions of Financial Statement Users: An Experimental Investigation of Auditor Materiality Thresholds

Academic journal article Auditing: A Journal of Practice & Theory

The Effect of Misstatements on Decisions of Financial Statement Users: An Experimental Investigation of Auditor Materiality Thresholds

Article excerpt

INTRODUCTION

The foundation of the independent auditor's concept of materiality is that misstatements in accounting information below some level of materiality do not change users' perceptions of a company's economic condition. In any audit engagement, this premise greatly influences the extent and focus of audit effort (AU Sec. 150.04). Materiality judgments are integrally connected to notions about what is acceptable risk and the nature and extent of evidence that is required to reach an opinion (AU Sec. 312.08). While materiality bears directly on the conduct of an audit and the subsequent opinion, regulators and the investing community appear skeptical about its application. The SEC recently charged that auditors and their clients may be misusing the concept of materiality (Levitt 1998). In response to the SEC's concern, the Public Oversight Board (POB) established a panel on audit effectiveness whose final report calls for additional guidance to assist auditors in determining whether identified misstatements are material (POB 2000). Accordingly, we address our research to the question of whether misstatements that fall within conventional materiality thresholds (e.g., misstatements that are less than either 10 percent of net income or 0.5 percent of sales) affect the decisions of financial information users.

One important aspect in judging materiality is the magnitude of the misstatement. Although some misstatements may be considered material based on their nature, many (perhaps most) are material simply because they are sufficiently large. In determining materiality, the FASB, the SEC, and the AICPA all have a long tradition of taking a user perspective. For example, the FASB (1980) states that an item is material if "it is probable that the judgment of a reasonable person relying on the report would have been changed or influenced by the correction of the item." Similarly, when auditors are sued for being associated with allegedly misleading financial information, courts explicitly take into account the security price that would have resulted had no misstatement occurred. (1) However, controversy remains over defining the magnitude necessary to affect users' decisions. One extreme view is the legal theory used by some plaintiffs, known as "fraud on the markets," which argues that all misstatements are reflected in security prices (Newman et al. 1995; Steen 1994). Thus, it is assumed that no misstatement is too small to have no effect on market price. In contrast, pronouncements by authoritative bodies and the very practice of setting materiality thresholds indicate that auditors believe that some level of misstatement is acceptable. What is not known is whether conventional materiality levels employed by auditors are appropriate from a user perspective.

To investigate the appropriateness of common materiality thresholds from a user perspective, we conduct a series of laboratory asset markets. (2) In these markets, participants bought and sold shares of stock using financial information based on actual data from companies traded on the NYSE. Our primary manipulation involves seeding misstatements into this information. The validity of our approach is strongly supported by the finding that in trading periods where no misstatements are seeded, laboratory market prices correlate highly (r = 0.923) with prices observed on the NYSE for these same companies. We find that prices generated in the absence of a misstatement do not differ from those generated in the presence of misstatements that are within a materiality threshold that is defined conservatively (i.e., 5 percent of income or 0.25 percent of revenues if income is small) or liberally (i.e., 10 percent of income or 0.5 percent of revenues). For companies with large misstatements (defined as 3 and 5 times the liberally defined materiality threshold), prices do differ significantly from those generated in the absence of a misstatement. These findings support the notion that misstatements within conventionally employed materiality levels (where materiality is based solely on the magnitude of the misstatement) do not affect users' decisions. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.