Neural Forecasting of the Italian Sovereign Bond Market with Economic News

TL;DR

This paper introduces a neural network approach utilizing economic news to improve forecasting of the Italian 10-year interest rate spread, demonstrating superior performance over traditional methods.

Contribution

It presents a novel integration of economic news data with deep learning models, specifically LSTM-based networks, for bond market forecasting.

Findings

LSTM-based neural networks outperform classical machine learning models.

Inclusion of economic news improves forecast accuracy.

DeepAR model enhances interest rate spread predictions.

Abstract

In this paper we employ economic news within a neural network framework to forecast the Italian 10-year interest rate spread. We use a big, open-source, database known as Global Database of Events, Language and Tone to extract topical and emotional news content linked to bond markets dynamics. We deploy such information within a probabilistic forecasting framework with autoregressive recurrent networks (DeepAR). Our findings suggest that a deep learning network based on Long-Short Term Memory cells outperforms classical machine learning techniques and provides a forecasting performance that is over and above that obtained by using conventional determinants of interest rates alone.