Is attention truly all we need? An empirical study of asset pricing in pretrained RNN sparse and global attention models
arXiv:2508.19006v2 Announce Type: replace-cross Abstract: This study investigates the pre-trained RNN attention models with the mainstream attention mechanisms, such as additive attention, Luong's three attentions, global self-attention and sliding window sparse attention, for the empirical asset pricing research on the top 420 large-cap US stocks. This is the first paper on the large-scale state-of-the-art (SOTA) attention mechanisms applied in the asset pricing context. They overcome the limitations of the traditional machine learning-based asset pricing, such as mis-capturing the temporal d
The increasing sophistication of transformer and RNN models is allowing researchers to apply these advanced AI architectures to complex financial problems like asset pricing.
This study demonstrates that state-of-the-art AI attention mechanisms can significantly enhance empirical asset pricing, offering more accurate and comprehensive temporal insights than traditional methods.
Asset pricing models will move beyond traditional machine learning by integrating advanced deep learning techniques, improving predictive power and risk management.
- · Quantitative hedge funds
- · Asset management firms
- · AI/ML researchers in finance
- · Traditional econometric modeling in finance
- · Firms reliant on outdated forecasting methods
More accurate asset pricing and risk models will become commercially available.
This could lead to increased market efficiency and potentially reduce alpha opportunities for less technologically advanced players.
The widespread adoption of these models might introduce new forms of systemic risk if models share similar underlying biases or vulnerabilities.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG