Bold Stock Recommendations: Informative or Worthless?
Bibliographic record
Abstract
ABSTRACT We select a small set of recommendations that lie in the upper and lower tail of the empirical distribution of divergences between a recommendation, and the consensus over the window (−30, −1) days prior to that recommendation. We classify these extremely divergent recommendations as bold, and then subdivide them into informative bold recommendations that lead other analysts (leading‐bold) and those that are ignored by other analysts (contra‐bold) based on the consensus change in the 30 days after the announcement. We focus on the information conveyed to the market by these bold, leading‐bold, and contra‐bold recommendations through their effects on cumulative abnormal returns (CAR). We find that bold recommendations are not anticipated by market participants (CARs are negative before a bold buy and positive before a bold sell). The next finding is that the market responds strongly to both leading and contra‐bold recommendations over the (0, +4)‐day window and that these reactions are stronger than that to nonbold recommendations. In contrast, over the longer (0, +30)‐day window, leading‐bold recommendations earn additional returns whereas contra‐bold ones reverse significantly due to lack of confirmation. The overall pattern is one of rational market reaction both in the short and long windows. We support the rationality of the market reaction by showing that the percentage of leading‐bold recommendations exceeds that of contra‐bold recommendations, and that these two types of recommendations cannot be separated using observable analyst characteristics such as experience or brokerage size.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.005 | 0.053 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".