Resumen

GARCIA MEDINA, Andrés. The use of Twitter in financial analysis: An approach from the econophysics. Inter disciplina [online]. 2017, vol.5, n.12, pp.23-39.  Epub 09-Dic-2024. ISSN 2448-5705.  https://doi.org/10.22201/ceiich.24485705e.2017.12.61469.

Mathematical techniques from statistical physics, especially from random matrix theory (RMT), have been used to analyze textual data from Twitter in the context of global financial markets. For this, we have analyzed a period of time of 7 months along 2014, considering the returns of 20 global financial indices in order to compare the results. Textual information was extracted by assembling different programming languages, constructing time series of polarity through sentiment analysis. RMT revealed that there are true correlations in financial indices and polarities. In addition, a good concordance was found between the temporal behavior of the extreme eigenvalues of returns and polarities, with similar results for the inverse participation ratio, which gives us information about the emergence of common factors in global financial information, regardless if we are using polarities or returns as data source. Our results suggest that using polarity as a new financial indicator provide of useful information about collective and even individual behavior of global financial indices. This builds a strong and novel evidence against the efficient market hypothesis, and supporting the school of behavioral finance where the market prices are affected by the irrational decisions of investors, which are influenced by trending news and social networks.

Palabras llave : random matrix theory; Twitter; sentiment analysis; behavioral finance.

        · resumen en Español     · texto en Español     · Español ( pdf )