Analýza přítomnosti bublin na trhu kryptoměn

This master thesis focuses on the analysis of the cryptocurrency market in 2016-2019 period and aims to confirm the presence of bubbles in this market. First, there are performed SADF and GSADF tests recommended as being able to detect the presence of financial bubbles as well as to indicate the starting and the end date using a date-stamping procedure. Based on the outcomes of the tests performed over the twelve major cryptocurrencies, according to their market capitalization, it can be concluded that there were bubbles present which burst around the break between 2017 and 2018 and there are bubbles started in 2019 for a few cryptocurrencies. Second, there is applied a framework called Log-Periodic Power Law model which is suggested as being able to capture the end of the bubble ex-post and ex-ante together with the price development. Consequently, the Log-Periodic Power Law model was able to capture the time of crash for different cryptocurrencies with high accuracy at the end of 2017 and beginning 2018. It can be stated that the cryptocurrency market had a price exuberance resulted in the bubble burst. For the bubbles started in 2019, the prediction with expanding rolling window was able to mimic the price evolution better compared to the single one prediction period. The critical times of crash were again determined quite precisely but the prediction horizon was short. Overall, the framework captures the speed of the price acceleration and the log-periodic oscillation which differ significantly from one cryptocurrency to the other. This means that the price evolution of cryptocurrencies has different patterns during the bubble period and the price behavior cannot be simply generalized. However, based on the results of calibration of the Log-Periodic Power Law model it is a promising framework the price evolution in the cryptocurrency market which is historically prone to high volatility.