Levels of voluntary disclosure in IPO prospectuses: an empirical analysis

Abstract

Purpose - This paper aims to focus onhowforecasts information is disclosed in IPO prospectuses. In France, managers report either detailed forecasts or only a brief summary. Design/methodology/approach - The authors investigate the determinants and consequences of the varying levels of details provided in these forecasts. The research is based on a sample of 82 IPOs on the Euronext Paris market (2000‐2002). Findings - The paper shows that only two variables are associated with highly detailed forecast disclosures: forecast horizon and firm age. It is also found that the forecast error decreases as the level of detail in the forecast disclosures increases. This finding is robust to a reverse causality test (Heckman two‐stage self‐selection procedure) and suggests that the level of detail in forecast disclosures enhances the reliability of earnings forecasts. Research limitations/implications - The paper suffers from at least two potential flaws. First, omitted variables, such as the possession of good news or proprietary costs. can influence both forecast errors and the level of detail of forecasts. Second, the negative association between the level of detail in forecast information and forecast errors may either show that detailed information leads to less forecast error or reflect a self‐selection bias. Practical implications - This research could have implications for stock market regulators as it suggests that mandatory disclosure of highly detailed forecasts would improve the effeciency of the markets by reducing forecast error. Originality/value - This paper contributes to be literature by presenting evidence tha the way forecast information is disclosed in IPO prospectuses is of importance and by documenting a negative association between forecast error and the level of detail in forecast disclosures.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 53 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.