AI and Financial Fragility: A Framework for Measuring Systemic Risk in Deployment of Generative AI for Stock Price Predictions
Bibliographic record
Abstract
In a few years, most investment firms will deploy Generative AI (GenAI) and large language models (LLMs) for reduced-cost stock trading decisions. If GenAI-run investment decisions from most firms are heavily coordinated, they could all give a “sell” signal simultaneously, triggering market crashes. Likewise, simultaneous “buy” signals from GenAI-run investment decisions could cause market bubbles with algorithmically inflated prices. In this way, coordinated actions from LLMs introduce systemic risk into the global financial system. Existing risk analysis for GenAI focuses on endogenous risk from model performance. In comparison, exogenous risk from external factors like macroeconomic changes, natural disasters, or sudden regulatory changes, is understudied. This research fills the gap by creating a framework for measuring exogenous (systemic) risk from LLMs acting in the stock trading system. This research develops a concrete, quantitative framework to understand the systemic risk brought by using GenAI in stock investment by measuring the covariance between LLM stock price predictions across three industries (technology, automobiles, and communications) produced by eight large language models developed across the United States, Europe, and China. This paper also identifies potential data-driven technical, cultural, and regulatory mechanisms for governing AI to prevent negative financial and societal consequences.
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 | 0.010 | 0.050 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.001 |
| Science and technology studies | 0.001 | 0.003 |
| Scholarly communication | 0.003 | 0.004 |
| Open science | 0.002 | 0.003 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
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".