Source: Cao Y., Liu M. & Liu Z. (2025) — “When Charts Clarify and Mislead: Visual Disclosure in Earnings Call Presentations”
This is one of those risks that sounds abstract until you imagine explaining it after the fact. Then it suddenly becomes very concrete, very expensive, and very difficult to hide behind a slide deck.
Earnings call presentations have evolved significantly over the past two decades. What used to be text-heavy documents are now slide decks full of charts, graphics, and visual summaries. The question this paper asks is simple but important: do those visuals help investors or mislead them? The researchers study 432,003 slides and 175,231 charts from over 17,000 earnings call decks between 2003 and 2023 using transformer-based computer vision models. The scale of this analysis would have been impossible without AI — this is precisely the kind of task where automated processing enables research that couldn’t otherwise happen. The findings split along informative versus misleading dimensions. On the informative side: firms that use more charts with genuine informational content — clear data representations without visual tricks — show lower implied volatility around earnings calls. Visual information, when done well, actually reduces investor uncertainty. The effect is strongest for firms that are inherently harder to value: more business segments, more intangible assets, higher pre-call uncertainty. Visuals seem to help most where the underlying fundamentals are most complex. On the misleading side: the researchers identify five chart features associated with distortion — missing or truncated Y-axes, dual Y-axes, 3D effects, and omitted data labels.
My takeaway: the danger is rarely the dramatic thing in the headline. It is the quiet gap between knowing a risk exists and assigning someone to do something about it. Very unglamorous. Very important.