I discovered 2 new analysis papers that discover Language Fashions in Monetary Economics.

* The primary one [Hansen et al] exhibits that LLMs like ChatGPT can decode **Fedspeak** higher than people monetary specialists and BERT-based fashions.
* The second [Lopez-Lira et al] makes use of ChatGPT to formulate buying and selling methods primarily based on sentiment evaluation.

I’m not an economics professional, however I feel each of them have some limitations. For instance, the second doesn’t think about buying and selling prices. What do you suppose?

I talk about them [here](https://medium.datadriveninvestor.com/using-chatgpt-to-decode-stock-price-movements-an-academic-survey-b1b6cf2bbf0b?sk=af3bc35f39032c0e26a908d88987916f)