Weak Supervision and Black-Litterman for Automated ESG Portfolio Construction
Bibliographic record
Abstract
The authors propose an approach that combines modern machine learning techniques in natural language processing with portfolio optimization to incorporate views of companies’ environment, social, and governance (ESG) performance. This is automatically done through curating and subsequently converting large-scale news data into portfolio management decisions. They train a machine learning news data classifier to automatically identify several key ESG issues in news data over time. They then aggregate these issues over time to generate a views vector under the Black–Litterman portfolio framework and finally compare the performance of an ESG-tilted portfolio against a standard Black–Litterman portfolio. They also show how this can be achieved at scale, in a fully automated manner, and with consistency over large periods of time. Their methodology thus demonstrates a reasonable and agile method for asset managers to incorporate ESG considerations into their portfolios free of any exclusionary frameworks and without sacrificing performance. <b>TOPICS:</b>ESG investing, quantitative methods, statistical methods, big data/machine learning, portfolio construction <b>Key Findings</b> ▪ The authors describe a theoretical framework for using an automated NLP system to incorporate ESG criteria into portfolio optimization decisions. ▪ The authors demonstrate the technical implementation details for incorporating ESG signals into augmented portfolio weights. ▪ The authors demonstrate the competitive performance of such a portfolio through a long-term historical backtest with the S&P 500 Index.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.000 | 0.002 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 teacher head, 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".