Source: Kim & Nikolaev (2024) — “Context-Based Interpretation of Financial Information”
I like finance papers that accidentally become papers about human behaviour. This one does that. It looks like a study about data, markets, or disclosure, but underneath it is really about how people decide what to believe.
Financial statements contain two kinds of information: numbers and narrative. The numbers — earnings, assets, liabilities, cash flows — are the formal accounting record. The narrative — MD&A sections, footnotes, risk factors — explains what the numbers mean, what drove them, and what they imply about the future. Most financial research treats these two channels separately. This paper asks a different question: what happens at their interaction? Specifically, when the narrative context around a number changes, does the informativeness of that number change? Using deep learning to capture the interactions between numeric and narrative disclosure, the researchers find that contextualisation substantially increases the informativeness of accounting numbers. The value of the interaction — the degree to which narrative context changes what a number implies about a firm’s future — is not just additive. The informational value of interactions dominates the direct informational value of either the numbers or the narrative alone. This result challenges a simplification that runs through a lot of accounting research: that text and numbers can be studied independently and their effects summed. If the interaction between them is where most of the information actually lives, then methods that study them separately are systematically underestimating the significance of both.
My takeaway: AI and data do not remove judgment from finance. They move judgment to the part where people decide what the signal means, whether it is trustworthy, and whether they are brave enough to act on it.