The Uncertainty of Machine Learning Predictions in Asset Pricing
Yuan Liao, Xinjie Ma, Andreas Neuhierl, Linda Schilling · 2025
Read The Uncertainty of Machine Learning Predictions in Asset Pricing by Yuan Liao, Xinjie Ma, Andreas Neuhierl, Linda Schilling (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-00549.txt.
Description
Yuan Liao, Xinjie Ma, Andreas Neuhierl, and Linda Schilling ask, in a 2025 paper, how sure we should be of a neural network's stock return forecast. Machine learning in asset pricing usually hands over a single number, with no error bar. Investors then treat shaky and solid forecasts alike.
The team shows that network forecasts share the same large-sample behavior as classic nonparametric estimates. That link yields a closed-form standard error for each forecast, plus a workable bootstrap when the formula is awkward. Feeding these error bars into a cautious portfolio rule gives a principled reason to shrink bold bets. In tests, the approach improved out-of-sample results. Honest uncertainty, the authors argue, is part of the forecast, not a footnote.
Catalog record
| CREATOR | Yuan Liao, Xinjie Ma, Andreas Neuhierl, Linda Schilling |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 144 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-00549