Investigating the relationship between cryptocurrencies: An econometric approach for Bitcoin and altcoins

The rapid price increase in cryptocurrencies create new opportunities for investors. The relationship between virtual currencies and price volatility is questioned by many. Accordingly, research in cryptocurrencies expanded especially on Bitcoin and subsequently altcoins. Moreover, digital currencies have many advantages which makes them admirable to investors. The purpose of this study is to determine which crypto coins should be chosen by investors to create their portfolio baskets in virtual markets. Within the scope of the research, the daily closing prices of the top 10 cryptocurrencies, which constitute 88% of the trading volume in the virtual markets and with the highest trading volume, consisting of various start dates between 2013-2020, is taken as research source. The relationship between cryptocurrencies has been revealed in the long term and has been evaluated considering whether the risk can be minimized by making portfolio diversification with virtual currency creation. In the application part, the stationarity ranks of the series were first examined with the unit root test by Carrion-i Sylvestre (2009). As a result of the unit root test, Maki (2012) Cointegration test was performed, showing structural breaks in 28 matches. In 11 matches, the alternative hypothesis with long-term co-operation was accepted, while in 17 matches, the null hypothesis was accepted, indicating that there was no such relationship. ARDL Boundary Test was applied to the series that were stagnant at the level, and the long-term relationship was determined in 4 matches. As a result of the coefficient estimator tests performed to determine the direction and strength of the long-term cointegration relationship, negative correlations in the long term in 2 matches and positive in the long run in 13 matches were determined. In line with the findings obtained, it will be an appropriate decision for investors to add virtual currencies that do not have any cointegration and negative action in permanent consideration. Examining previous research, Polat and Gemici, (2018) Ciaian et al. 's (2018) results are similar to the findings of this study. Further research can use long-term hourly data of virtual currencies, macroeconomic developments, exchange rates, virtual money supply and demand as key figures to be considered as variables. These findigs might reveal new investment behavior to researchers and investors.