Saved in:
Bibliographic Details
Main Authors: Cortesi, Federico Vittorio, Iannone, Giuseppe, Crippa, Giulia, Poggio, Tomaso, Beneventano, Pierfrancesco
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2603.02620
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908861745594368
author Cortesi, Federico Vittorio
Iannone, Giuseppe
Crippa, Giulia
Poggio, Tomaso
Beneventano, Pierfrancesco
author_facet Cortesi, Federico Vittorio
Iannone, Giuseppe
Crippa, Giulia
Poggio, Tomaso
Beneventano, Pierfrancesco
contents Neural networks applied to financial time series operate in a regime of underspecification, where model predictors achieve indistinguishable out-of-sample error. Using large-scale volatility forecasting for S$\&$P 500 stocks, we show that different model-training-pipeline pairs with identical test loss learn qualitatively different functions. Across architectures, predictive accuracy remains unchanged, yet optimizer choice reshapes non-linear response profiles and temporal dependence differently. These divergences have material consequences for decisions: volatility-ranked portfolios trace a near-vertical Sharpe-turnover frontier, with nearly $3\times$ turnover dispersion at comparable Sharpe ratios. We conclude that in underspecified settings, optimization acts as a consequential source of inductive bias, thus model evaluation should extend beyond scalar loss to encompass functional and decision-level implications.
format Preprint
id arxiv_https___arxiv_org_abs_2603_02620
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Same Error, Different Function: The Optimizer as an Implicit Prior in Financial Time Series
Cortesi, Federico Vittorio
Iannone, Giuseppe
Crippa, Giulia
Poggio, Tomaso
Beneventano, Pierfrancesco
Machine Learning
Computational Finance
Neural networks applied to financial time series operate in a regime of underspecification, where model predictors achieve indistinguishable out-of-sample error. Using large-scale volatility forecasting for S$\&$P 500 stocks, we show that different model-training-pipeline pairs with identical test loss learn qualitatively different functions. Across architectures, predictive accuracy remains unchanged, yet optimizer choice reshapes non-linear response profiles and temporal dependence differently. These divergences have material consequences for decisions: volatility-ranked portfolios trace a near-vertical Sharpe-turnover frontier, with nearly $3\times$ turnover dispersion at comparable Sharpe ratios. We conclude that in underspecified settings, optimization acts as a consequential source of inductive bias, thus model evaluation should extend beyond scalar loss to encompass functional and decision-level implications.
title Same Error, Different Function: The Optimizer as an Implicit Prior in Financial Time Series
topic Machine Learning
Computational Finance
url https://arxiv.org/abs/2603.02620