Online forecasting for financial time series.
4 affiliations
- 1Department of Statistics, University of Oxford, St Giles' 24-29, Oxford, OX1 3LB UK.
- 2Oxford-Man Institute of Quantitative Finance, University of Oxford, Oxford, OX2 6GG UK.
- 3The Alan Turing Institute, University of Oxford, 96 Euston Rd, London, NW1 2DB UK.
- 4Mathematical Institute, University of Oxford, Andrew Wiles Building 24-29, Oxford, OX2 6GG UK.
Abstract
We introduce OFTER (online forecasting via temporal embedding regression), a time series forecasting pipeline tailored for financial problems centered on mid-sized multivariate time series. OFTER utilizes the non-parametric models of k-nearest neighbors and generalized regression neural networks, integrated with a dimensionality reduction component. To circumvent the curse of dimensionality, we employ a weighted norm based on a modified version of the maximal correlation coefficient. The pipeline we introduce is specifically designed for online tasks, has an interpretable output, and is able to outperform several state-of-the-art baselines. The computational efficacy of the algorithm, its online nature, and its ability to operate in low signal-to-noise regimes, render OFTER an ideal approach for financial multivariate time series problems, such as daily equity forecasting. Our work demonstrates that while deep learning models hold significant promise for time series forecasting, traditional methods carefully integrating mainstream tools remain very competitive alternatives with the added benefits of scalability and interpretability.
Related
- AHAT-Fin: an adaptive hybrid attention and market-state-aware framework for financial time-series forecasting
- GLoR-forecast: state-contingent global-local text routing for financial and socio-economic time-series forecasting
- Large-Scale Benchmarking of Quantum Neural Network Configurations for Financial Time Series Forecasting
- Encoding Time Series Dynamics with Sliding-Window Signatures for Online Forecasting
- Hierarchical orthogonal ESNs with adaptive signal decomposition for chaotic financial time series prediction
BibTeX
@article{michael2026online,
title = {{Online forecasting for financial time series.}},
author = {Nikolas Michael and Mihai Cucuringu and Sam Howison},
journal = {International journal of data science and analytics},
year = {2026},
month = oct,
doi = {10.1007/s41060-026-01302-7},
url = {https://europepmc.org/article/MED/42892820},
note = {CC BY},
}
RIS
TY - JOUR
TI - Online forecasting for financial time series.
AU - Nikolas Michael
AU - Mihai Cucuringu
AU - Sam Howison
PY - 2026
DA - 2026/10/16
JO - International journal of data science and analytics
DO - 10.1007/s41060-026-01302-7
UR - https://europepmc.org/article/MED/42892820
AB - We introduce OFTER (online forecasting via temporal embedding regression), a time series forecasting pipeline tailored for financial problems centered on mid-sized multivariate time series. OFTER utilizes the non-parametric models of k-nearest neighbors and generalized regression neural networks, integrated with a dimensionality reduction component. To circumvent the curse of dimensionality, we employ a weighted norm based on a modified version of the maximal correlation coefficient. The pipeline we introduce is specifically designed for online tasks, has an interpretable output, and is able to outperform several state-of-the-art baselines. The computational efficacy of the algorithm, its online nature, and its ability to operate in low signal-to-noise regimes, render OFTER an ideal approach for financial multivariate time series problems, such as daily equity forecasting. Our work demonstrates that while deep learning models hold significant promise for time series forecasting, traditional methods carefully integrating mainstream tools remain very competitive alternatives with the added benefits of scalability and interpretability.
SN - 2364-415X
SN - 2364-4168
ER -