Bitcoin Investment Strategies Based on Google Trends and AI Models

Autor: Raúl Gómez Martínez, Juan Gabriel Martínez Navalón, Camilo Prado Román
Rok vydání: 2022
Předmět:
Zdroj: Tripodos. :129-141
ISSN: 2340-5007
1138-3305
DOI: 10.51698/tripodos.2022.52p129-141
Popis: The evolution of the price of bitcoin has captured the attention of analysts in recent years. But how can a cryptocurrency be valued? Given that the price is linked to expectations, we propose, in this paper, to predict the trend of bitcoin using Google Trends as an explanatory variable. To do so, we develop two alternative algorithmic trading systems that buy or sell bitcoin depending on whether the searches for this term in Google increase or decrease. The approach is powered using artificial intelligence. The results of these trading systems are positive and show that trading strategies can be implemented based on investors’ mood about an asset, in this case measured through Google Trends. The use of artificial intelligence in trading is new and this is an example of its potential.
Databáze: OpenAIRE