Fusion of Sentiment and Asset Price Predictions for Portfolio\n Optimization

Résumé

The fusion of public sentiment data in the form of text with stock price\nprediction is a topic of increasing interest within the financial community.\nHowever, the research literature seldom explores the application of investor\nsentiment in the Portfolio Selection problem. This paper aims to unpack and\ndevelop an enhanced understanding of the sentiment aware portfolio selection\nproblem. To this end, the study uses a Semantic Attention Model to predict\nsentiment towards an asset. We select the optimal portfolio through a\nsentiment-aware Long Short Term Memory (LSTM) recurrent neural network for\nprice prediction and a mean-variance strategy. Our sentiment portfolio\nstrategies achieved on average a significant increase in revenue above the\nnon-sentiment aware models. However, the results show that our strategy does\nnot outperform traditional portfolio allocation strategies from a stability\nperspective. We argue that an improved fusion of sentiment prediction with a\ncombination of price prediction and portfolio optimization leads to an enhanced\nportfolio selection strategy.\n

Citer ce document

Accès au document

Ce document est hébergé sur son dépôt institutionnel d'origine.

Auteur(s)

Statistiques

Consultations : 1

Téléchargements : 0