Strategic Narratives, Institutional Positioning, and Text Signals

Summary

This study asks when market text should be followed and when it should be faded. It separates institutional statements (Say), media repetition (Echo), and revealed institutional positioning (Do). A linear-quadratic model describes an informed institution that can speak and trade before a partly credulous crowd. The paper also develops tools to identify echoed articles using timing and embedding similarity, align text embeddings with subsequent returns, and infer who moved first from price paths.

The model gives conditions under which an institution optimally talks an asset down while buying it, and a distribution-free result links negative covariance between statements and positioning to negative predictive content in the words. In controlled simulations with known ground truth, echo sentiment predicts returns negatively in all 29 markets; a rolling Say-Do correlation detects false-alarm events with AUC 0.90. The authors also describe cases where the tools fail. These results come from simulated markets and do not establish performance on live market text or real trading data.

Key ideas

• The framework distinguishes institutional statements, media echoes, and revealed positioning.
• A negative covariance between statements and positioning implies that words have negative predictive content under the paper's identity.
• The model specifies conditions in which an institution benefits from speaking negatively while buying.
• The measurement methods combine article timing and similarity, return-aligned text embeddings, and price-path signatures.
• In 29 controlled simulated markets, echo sentiment predicts returns negatively, while a rolling Say-Do correlation flags false alarms with AUC 0.90.

Tags

# Say, Echo, Do: Strategic Narratives and Revealed Positioning in Financial Markets


# Say, Echo, Do: Strategic Narratives and Revealed Positioning in Financial Markets









Machine-learning signals built from financial text treat what institutions say, and what the media repeat, as evidence about value. But whoever shapes a narrative may be trading against it. We study markets with three observable voices: institutional statements (Say), media repetition (Echo) and revealed positioning (Do). We ask when words should be followed and when they should be faded. In a linear-quadratic model of an informed institution that speaks and trades before a partly credulous crowd, talking an asset down while buying it is optimal exactly when $\varphi^2<2λk<\varphi$. A distribution-free identity then shows that when the observable Say-Do covariance is negative, words carry negative predictive content and should be faded. For measurement, we derive (i) an exact factorised posterior over which articles are echoes, combining arrival times with embedding similarity; (ii) a return-aligned contrastive objective that attains its bound exactly when squared embedding distances are an increasing affine function of squared outcome distances, with the tightest loss-based certificate of which neighbour rankings survive imperfect training; and (iii) a path-signature statistic for who moved first. In a controlled market with known ground truth, echo sentiment predicts returns with a significantly negative sign in all 29 simulated markets, the rolling Say-Do correlation flags false-alarm events with an AUC of 0.90, and return-aligned embeddings organise headlines by consequence rather than topic. We also report where the tools fail.

Shown in full with attribution under the source's licence. Licence: abstract CC0

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.