O mercado de ações e o Bitcoin: uma análise via regressão quantílica

Abstract

While conventional currencies are totally tied to the economy of a nation and the need for mediation in their negotiations, Bitcoin does not have this limitation and operates more freely and independently, online, on the Internet, which makes it totally decentralized and not tied to the health of a national economy or economic block of countries. In this context, the general objective of the research was to verify the existence of the relationship between the value of the quotation of the IBOVESPA, Shanghai, S&P 500, S&P/TSX, NYSE and Hangsenghsi markets with the value of Bitcoin, during the COVID-19 pandemic. The data collected corresponded to the period from January 2019 to August 2022. The study, relied on the application of the quantile regression technique, for the analysis of the association between the response variable (stock variable) with the dependent variable (Bitcoin variable) in the various quantiles of the conditional distribution. The results indicate that there is a positive impact of the IBOVESPA, Shanghai, S&P 500, SP/TSX and NYSE markets on the value of Bitcoin, and the models can be estimated by both quantile regression and MQO, but, in contrast, for the Hangsenghsi market, the most suitable model for estimating the data is the quantile regression, where quantiles lower than 30% the impact is negative, and, above that, no quantile is significant. It is concluded that the application of the quantile regression model becomes an alternative to the MQO by not requiring normality, in isolated case, may present different interpretations for the coefficients estimated in the quantiles.