Ploutos: Towards interpretable stock movement prediction with financial large language model
Ploutos: A financial LLM framework for interpretable stock movement prediction using multimodal data and expert insights.
What it examines
The paper introduces Ploutos, a novel financial large language model framework designed to predict stock movements by integrating textual and numerical data, addressing challenges in interpretability and adaptability in financial models.
What it concludes
The research suggests that Ploutos can significantly improve stock movement prediction accuracy and interpretability. Potential applications include financial investment strategies and market analysis. Future research could explore optimizing computational costs and incorporating additional data types.
Evidence objects
The research suggests that Ploutos can significantly improve stock movement prediction accuracy and interpretability. Potential applications include financial investment strategies and market analysis. Future research could explore optimizing computational costs and incorporating additional data types.
key_findings bullet 1 · key_findings · validation V0
Raw abstract and provenance
Abstract: Recent advancements in large language models (LLMs) have opened new pathways for many domains. However, the full potential of LLMs in financial investments remains largely untapped. There are two main challenges for typical deep learning-based methods for quantitative finance. First, they struggle to fuse textual and numerical information flexibly for stock movement prediction. Second, traditional… ▽ More Recent advancements in large language models (LLMs) have opened new pathways for many domains. However, the full potential of LLMs in financial investments remains largely untapped. There are two main challenges for typical deep learning-based methods for quantitative finance. First, they struggle to fuse textual and numerical information flexibly for stock movement prediction. Second, traditional methods lack clarity and interpretability, which impedes their application in scenarios where the justification for predictions is essential. To solve the above challenges, we propose Ploutos, a novel financial LLM framework that consists of PloutosGen and PloutosGPT. The PloutosGen contains multiple primary experts that can analyze different modal data, such as text and numbers, and provide quantitative strategies from different perspectives. Then PloutosGPT combines their insights and predictions and generates interpretable rationales. To generate accurate and faithful rationales, the training strategy of PloutosGPT leverage rearview-mirror prompting mechanism to guide GPT-4 to generate rationales, and a dynamic token weighting mechanism to finetune LLM by increasing key tokens weight. Extensive experiments show our framework outperforms the state-of-the-art methods on both prediction accuracy and interpretability. △ Less
Source row: 1540 · abstract type: unknown