A comparative research study of the Cryptocurrencies’ volatility using GARCH-model analysis

Abstract

This thesis examines four exchange rate pairs of fiat currencies in comparison to four of the main cryptocurrencies based on market capitalization. The primary goal is to obtain new estimates for the cryptocurrencies based on the use of the GARCH (1.1) model, the EGARCH (1.1) model and simple historical volatility (SHV). The volatility is regarded as an essential risk measure in the management of financial flows, which makes it reasonable to conduct a multilateral scientific analysis and a comparison of this new digital phenomenon. The calculated volatility estimates are based on the Banks of International Settlement’s data on currency exchange rates, together with the cryptocurrency data retrieved from Yahoo finance. Concretely, the aim is to reveal the significant differences between the volatility of the following exchange rates EUR/USD, GDP/USD, CNY/USD and YEN/USD in comparison with the closing prices of bitcoin, ether, xrp, and xlm. Furthermore, the period of observation is divided into an in-sample and an out-of-sample period. The models are estimated in the in-sample period and then used to forecast the volatility in the out-of-sample period. This thesis uses the squared returns as an unbiased approximation of the latent volatility. The forecasts are evaluated using the mean squared error as a loss function. Estimation of volatility shows that the volatility of the cryptocurrencies, in general, are significantly higher than the fiat currency exchange rates volatility. In this regard, it can be concluded that the recognition of cryptocurrencies as a medium of exchange is premature, concerning their violation of the essential requirements. The findings of the thesis show that both the GARCH (1.1) model and the EGARCH (1.1) model predict fiat currency exchange rates more accurately. The EGARCH model has a better accuracy of predicting the currencies overall in the sample. The simple historical volatility was ten times higher on the cryptocurrencies compared to the fiat currency rates. Among the cryptocurrencies, there is an ambiguity to predict the most robust model, since no superior model could be identified by the mean squared error estimates. This also involves further research into the prediction among the cryptocurrencies themselves.

Popular Abstract

This thesis examines four exchange rate pairs of fiat currencies in comparison to four of the main cryptocurrencies based on market capitalization. The primary goal is to obtain new estimates for the cryptocurrencies based on the use of the GARCH (1.1) model, the EGARCH (1.1) model and simple historical volatility (SHV). The volatility is regarded as an essential risk measure in the management of financial flows, which makes it reasonable to conduct a multilateral scientific analysis and a comparison of this new digital phenomenon. The calculated volatility estimates are based on the Banks of International Settlement’s data on currency exchange rates, together with the cryptocurrency data retrieved from Yahoo finance. Concretely, the aim is to reveal the significant differences between the volatility of the following exchange rates EUR/USD, GDP/USD, CNY/USD and YEN/USD in comparison with the closing prices of bitcoin, ether, xrp, and xlm. Furthermore, the period of observation is divided into an in-sample and an out-of-sample period. The models are estimated in the in-sample period and then used to forecast the volatility in the out-of-sample period. This thesis uses the squared returns as an unbiased approximation of the latent volatility. The forecasts are evaluated using the mean squared error as a loss function. Estimation of volatility shows that the volatility of the cryptocurrencies, in general, are significantly higher than the fiat currency exchange rates volatility. In this regard, it can be concluded that the recognition of cryptocurrencies as a medium of exchange is premature, concerning their violation of the essential requirements. The findings of the thesis show that both the GARCH (1.1) model and the EGARCH (1.1) model predict fiat currency exchange rates more accurately. The EGARCH model has a better accuracy of predicting the currencies overall in the sample. The simple historical volatility was ten times higher on the cryptocurrencies compared to the fiat currency rates. Among the cryptocurrencies, there is an ambiguity to predict the most robust model, since no superior model could be identified by the mean squared error estimates. This also involves further research into the prediction among the cryptocurrencies themselves.

• author: Linder, David
• supervisor: Thomas Fischer
• organization: Department of Economics
• course: NEKN01 20182
• year: 2019
• type: H1 - Master's Degree (One Year)
• subject: Business and Economics
• keywords: Volatility Forecasting, Conditional Variance, GARCH, EGARCH, Simple historical volatility, Cryptocurrencies
• language: English
• id: 8975310
• date added to LUP: 2019-05-13 13:25:54
• date last changed: 2019-05-13 13:25:54