Examinando por Materia "Google Trends"

Mostrando 1 - 2 de 2
  • No hay miniatura disponible
    Publicación
    Impacto de Google trends y Twitter en la capitalización de mercado de bitcoin
    (Universidad EAFIT, 2022) Miranda Argel, Juan Fernando; Ballesteros Egurrola, Enrique Luis; Gaitan Riaño, Sandra Constanza
    Bitcoin is a decentralized currency supported by a blockchain system behind its issuance, so it transmits security regarding its value, since it cannot be simply duplicated or falsified. However, given that the volumes of transactions are high and quite variable, it is necessary to study their speculative nature, since studying the influence of some media factors allows us to observe how this type of less regulated asset can be the target of speculation and manipulation. through media such as Twitter and Google. In this study, we use the VAR model to analyze Bitcoin time series, its popularity, and some features. With the purpose of identifying if there is any correlation between fame in social networks and interest in Bitcoin within the Google search engine and the capitalization of this asset in the market. The results show that the influence of Google search interest and Twitter popularity does not have a large effect on Bitcoin's market capitalization.
  • No hay miniatura disponible
    Publicación
    Nowcasting de la economía colombiana a través de indicadores de expectativas y opiniones de los agentes económicos
    (Universidad EAFIT, 2021) Gómez Montoya, Sara; Botero García, Jesús Alonso
    This research project was intended to estimate the performance of the Colombian economy reflected through the ISE (Indicator of Monitoring of the Economy) from conventional and unconventional indicators of expectations of economic agents through the Nowcasting model, which uses a set of high-frequency, standardized and stationary indicators to forecast the monthly ISE in Colombia and the quarterly GDP with the average of the 3 estimated ISE values. This was achieved through the machine learning approach and dynamic factor models. The main purpose of this project was to answer the question of "How explanatory and useful are the confidence indicators and expectations of economic agents for predicting economic growth?" This question was answered through dynamic correlations, which showed procyclical, advanced, contemporary and lagged movements of the variables.