Deep learning for bond yield forecasting: the LSTM-LagLasso
Deep learning for bond yield forecasting: the LSTM-LagLasso
We present long short-term memory (LSTM)-LagLasso, a novel explainable deep learning approach applied to bond yield forecasting. Our method involves feature selection from a large universe of potential features and forecasts bond yields using dynamic LSTM networks. It examines the internal gating signals of a trained LSTM and explains their dynamics through exogenous variables that may influence bond price formation. By considering these variables at various lags and using the Lasso technique for feature selection, we demonstrate how different hidden units within the LSTM dynamically adjust to make predictions across different temporal regimes and how their evolution is shaped by various external factors. In an empirical study on government bond yield forecasting, we demonstrate the statistical accuracy of LSTM-LagLasso compared to a multilayer perceptron (MLP) and highlight its explainability.
Asset pricing, Bond yield forecasting, Explainable AI (XAI), LagLasso, Long short-term memory network, long short-term memory network, bond yield forecasting, eXplainable AI, asset pricing
Nunes, Manuel
af597793-a85a-463c-9d12-0ae4be7e0a69
Gerding, Enrico
d9e92ee5-1a8c-4467-a689-8363e7743362
McGroarty, Frank
693a5396-8e01-4d68-8973-d74184c03072
Niranjan, Mahesan
5cbaeea8-7288-4b55-a89c-c43d212ddd4f
Sermpinis, Georgios
d8497649-1c4d-4b93-af81-57560c118690
23 January 2025
Nunes, Manuel
af597793-a85a-463c-9d12-0ae4be7e0a69
Gerding, Enrico
d9e92ee5-1a8c-4467-a689-8363e7743362
McGroarty, Frank
693a5396-8e01-4d68-8973-d74184c03072
Niranjan, Mahesan
5cbaeea8-7288-4b55-a89c-c43d212ddd4f
Sermpinis, Georgios
d8497649-1c4d-4b93-af81-57560c118690
Nunes, Manuel, Gerding, Enrico, McGroarty, Frank, Niranjan, Mahesan and Sermpinis, Georgios
(2025)
Deep learning for bond yield forecasting: the LSTM-LagLasso.
International Journal of Finance & Economics.
(doi:10.1002/ijfe.3116).
Abstract
We present long short-term memory (LSTM)-LagLasso, a novel explainable deep learning approach applied to bond yield forecasting. Our method involves feature selection from a large universe of potential features and forecasts bond yields using dynamic LSTM networks. It examines the internal gating signals of a trained LSTM and explains their dynamics through exogenous variables that may influence bond price formation. By considering these variables at various lags and using the Lasso technique for feature selection, we demonstrate how different hidden units within the LSTM dynamically adjust to make predictions across different temporal regimes and how their evolution is shaped by various external factors. In an empirical study on government bond yield forecasting, we demonstrate the statistical accuracy of LSTM-LagLasso compared to a multilayer perceptron (MLP) and highlight its explainability.
Text
Int J Fin Econ - 2025 - Nunes - Deep Learning for Bond Yield Forecasting The LSTM‐LagLasso (1)
- Version of Record
More information
Accepted/In Press date: 29 December 2024
Published date: 23 January 2025
Keywords:
Asset pricing, Bond yield forecasting, Explainable AI (XAI), LagLasso, Long short-term memory network, long short-term memory network, bond yield forecasting, eXplainable AI, asset pricing
Identifiers
Local EPrints ID: 497185
URI: http://eprints.soton.ac.uk/id/eprint/497185
ISSN: 1076-9307
PURE UUID: e47d17fb-dad0-4890-858b-3d86d878fcb9
Catalogue record
Date deposited: 15 Jan 2025 18:02
Last modified: 30 Aug 2025 02:11
Export record
Altmetrics
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics
