Abstract
Despite numerous studies on financial contagion over the last few decades, the definition of financial contagion is not universally accepted. Furthermore, economic studies on cryptocurrency are limited. This research presents a pioneering study investigating cryptocurrencies contagion in the Asian financial markets. We use three methods of measuring contagions. The first method adjusts the bias in the correlation coefficient. The second method identifies exceedances (extreme return) and performs a multinomial logistic regression to explain the coexceedances. Finally, we also use a vector autoregression (VAR) system. We documented evidence that the (unadjusted) correlation coefficients between five cryptocurrencies and thirteen Asian financial markets tended to be low in both high and low variance periods. After adjusting the bias in the correlation coefficient, we found that there was no longer evidence of a significant change in the magnitude of the propagation mechanism from five cryptocurrencies to Asian foreign exchange and stock markets. Using a multinomial logit analysis, we also acknowledged that cryptocurrencies tended to be statistically insignificant determinants for both positive and negative coexceedances. Our third contagion analysis using a VAR system further revealed that cryptocurrencies tended to be statistically insignificant determinants explaining the current Asian financial market change variables. Our results suggest that cryptocurrencies do not possess a systemic risk to the Asian financial markets.
Original language | English |
---|---|
Pages (from-to) | 416-429 |
Number of pages | 14 |
Journal | Research in International Business and Finance |
Volume | 50 |
DOIs | |
Publication status | Published - 1 Dec 2019 |
Keywords
- Contagion
- Cryptocurrencies
- Multinomial logit
- Systemic risk
- Vector autoregression