A Theory of 'Prominent' Disclosure

'Prominence' plays an important role in financial reporting – an entity might assign an item to the footnotes, report it below the line, or comment on it in a press release versus the MD&A. We propose a model where quantitative disclosures are classified as more or less prominent, based on technically vague information. In our model, an entity must give sufficiently informative quantitative disclosures high prominence, sufficiently uninformative quantitative disclosures low prominence, while for a range of disclosures discretion is permitted. We show that a market of rational actors will react more strongly to negative (than positive) univariate quantitative disclosures when conditioned on either classification (low or high prominence). This effect strengthens as vagueness increases. We then examine how the reporting and bundling of multivariate items might occur in response to vagueness. We find that for unambiguously good or bad quantitative news, bundling increases with vagueness, and the market puts more weight on bad quantitative news disclosures than good quantitative news disclosures when conditioned on disclosure type. However, when multivariate quantitative news is mixed (good and bad items occur together), the probability of bundling is independent of vagueness, the market reaction to a bundle's classification is muted, and the market's reaction to unbundled items is amplified