TY - GEN
T1 - Predicting Daily Trends in the Lima Stock Exchange General Index Using Economic Indicators and Financial News Sentiments
AU - Ulloa, Adrian
AU - Espezua, Soledad
AU - Villavicencio, Julio
AU - Miranda, Oscar
AU - Villanueva, Edwin
N1 - Publisher Copyright:
© 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.
PY - 2022
Y1 - 2022
N2 - Predicting the future trend of the Lima Stock Exchange market is challenging because of its high volatility, transaction costs, and illiquidity. In this work, we investigate machine learning models able to use technical indicators, economic variables, and financial news sentiments to forecast the daily return trend of the S&P/BVL Peru General Index. To the best of our knowledge, no other published S&P/BVL predicting tool considered these joint sources of information as relevant input features. To do so, fifteen economic indicators relevant to the local market and sentiment-tagged financial news headlines were used as extra input features for multiple machine learning classification models and feature selection methods. In addition, the performance of the static learning approach (the only one used for this particular problem so far) was compared against an online learning approach, which could dynamically better adapt to such a volatile, emergent market. The results showed an increase in performance when using the economic variables and news sentiment in comparison to existing predicting tools of the local market. When comparing both learning approaches, online learning yielded better predictive accuracy than its static counterpart. To the best of our knowledge, this is the first effort to include all these novel features for predicting trends in the Lima Stock Exchange.
AB - Predicting the future trend of the Lima Stock Exchange market is challenging because of its high volatility, transaction costs, and illiquidity. In this work, we investigate machine learning models able to use technical indicators, economic variables, and financial news sentiments to forecast the daily return trend of the S&P/BVL Peru General Index. To the best of our knowledge, no other published S&P/BVL predicting tool considered these joint sources of information as relevant input features. To do so, fifteen economic indicators relevant to the local market and sentiment-tagged financial news headlines were used as extra input features for multiple machine learning classification models and feature selection methods. In addition, the performance of the static learning approach (the only one used for this particular problem so far) was compared against an online learning approach, which could dynamically better adapt to such a volatile, emergent market. The results showed an increase in performance when using the economic variables and news sentiment in comparison to existing predicting tools of the local market. When comparing both learning approaches, online learning yielded better predictive accuracy than its static counterpart. To the best of our knowledge, this is the first effort to include all these novel features for predicting trends in the Lima Stock Exchange.
KW - Machine learning
KW - News sentiment analysis
KW - S&P/BVL
KW - Stock market trend prediction
UR - http://www.scopus.com/inward/record.url?scp=85128974147&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-04447-2_3
DO - 10.1007/978-3-031-04447-2_3
M3 - Conference contribution
AN - SCOPUS:85128974147
SN - 9783031044465
T3 - Communications in Computer and Information Science
SP - 34
EP - 49
BT - Information Management and Big Data - 8th Annual International Conference, SIMBig 2021, Proceedings
A2 - Lossio-Ventura, Juan Antonio
A2 - Valverde-Rebaza, Jorge
A2 - Díaz, Eduardo
A2 - Muñante, Denisse
A2 - Gavidia-Calderon, Carlos
A2 - Valejo, Alan Demétrius
A2 - Alatrista-Salas, Hugo
PB - Springer Science and Business Media Deutschland GmbH
T2 - 8th Annual International Conference on Information Management and Big Data, SIMBig 2021
Y2 - 1 December 2021 through 3 December 2021
ER -