Bibliographic record
Abstract
Financial analysts play an intermediary role in financial markets, resulting in two steps for information to be fully absorbed into the stock price: analysts’ reaction to information, and investors’ reaction to analysts’ recommendations. Thus any observed inefficiency in stock pricing could result from two possibilities: analysts failed to fully incorporate the market information into their stock analysis, or the information released in the analysts’ report is not fully believed by investors. The documented optimism of financial analysts may suggest the possibility of the later case. To test the accuracy of analysts from another perspective, we follow a market microstructure model and use intraday market data to estimate the probability of an information event, the probability of good or bad news, and the rates that different traders arrive at the market. By comparing those estimates based on days with and without recommendation changes, we find inconsistent results with regard to a difference in the probability of an information event. For some stocks, we do observe an increase in the likelihood of news on days when analysts change their recommendations, but this is not the case for most stocks. However, even though they are inaccurate most of the time, uninformed investors usually believe financial analysts. Furthermore, it seems that uninformed investors disbelieve analyst recommendation changes at those instances when analysts are most accurate. Because of this, we hypothesise that market makers might suspect that orders in the opposite direction of an analyst’s recommendation change are more likely to come from informed traders. This is consistent with the intuition that most traders are uninformed and will simply follow the advice of a perceived expert, and therefore those that don’t follow that advice may be more likely to have special information of their own. We check whether there are any differences in the probability of information-based trading (PIN) and for the conditional probability of information-based trading conditioned on sell (PIN|sell) and buy (PIN|buy) between days with and without recommendation changes. We did not find any significant difference, indicating that although we may observe a higher arrival rate of informed traders on recommendation change days, the probabilities of information-based trading do not change substantially. More informed traders seem to come to the market merely because the higher arrival rate of uninformed traders on recommendation days gives them a good opportunity to camouflage their behaviour. And the specialists likely would not have to change their behaviour on those days by increasing or shifting bid-ask spreads since the increased costs from the higher volume of informed trading are balanced by increased profits from the higher volume of uninformed trading. Furthermore, regression of the probabilities of informed trading (conditional or unconditional) on firm size, trading volume, and volatility of daily return shows nothing significant, so we weren’t able to identify influential factors that affect informed trading or explain differences in informed trading between firms.
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 | 0.015 | 0.206 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.002 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.004 | 0.004 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 0.002 |
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".