English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Modern decision-making in fixed income asset management benefits fromintelligent systems, which involve the use of state-of-the-art machine learningmodels and appropriate methodologies. We conduct the first study of bond yieldforecasting using long short-term memory (LSTM) networks, validating itspotential and identifying its memory advantage. Specifically, we model the10-year bond yield using univariate LSTMs with three input sequences and fiveforecasting horizons. We compare those with multilayer perceptrons (MLP),univariate and with the most relevant features. To demystify the notion ofblack box associated with LSTMs, we conduct the first internal study of themodel. To this end, we calculate the LSTM signals through time, at selectedlocations in the memory cell, using sequence-to-sequence architectures, uni andmultivariate. We then proceed to explain the states' signals using exogenousinformation, for what we develop the LSTM-LagLasso methodology. The resultsshow that the univariate LSTM model with additional memory is capable ofachieving similar results as the multivariate MLP using macroeconomic andmarket information. Furthermore, shorter forecasting horizons require smallerinput sequences and vice-versa. The most remarkable property found consistentlyin the LSTM signals, is the activation/deactivation of units through time, andthe specialisation of units by yield range or feature. Those signals arecomplex but can be explained by exogenous variables. Additionally, some of therelevant features identified via LSTM-LagLasso are not commonly used inforecasting models. In conclusion, our work validates the potential of LSTMsand methodologies for bonds, providing additional tools for financialpractitioners.

Long short-term memory networks and laglasso for bond yield forecasting: Peeping inside the black box