The Forecasting of Cryptocurrency Price by Correlation and Regression Analysis

Abstract

In the present, the investment in cryptocurrencies is popular, both in and outside the stock market. Cryptocurrencies’ two main genera are Bitcoin and Ethereum, but they also have other several genera in the investment by investors.  In the investment, investors are at a very high risk of fluctuations of the value of the currency that will affect the profit or loss.  Forecasting of cryptocurrencies price is an interesting topic to researchers from different fields.  The data analyzed were the real price of cryptocurrencies in February 2018. Objectives of this study are to search for a relationship between cryptocurrencies and to forecast the price of target cryptocurrencies when the price of cryptocurrency base has been changed. In the analysis, we used two steps for searching the relationship of cryptocurrencies, and it was found that ETH had the highest correlation with XRP.  Then, we used regression analysis to create the function for forecasting the price of cryptocurrencies.  In the forecasting, it was found that the forecasting price of XRP varied in the same direction as the real price. So, the high price per coin (ETH) of cryptocurrencies can be used for forecasting the low price per coin (XRP) significantly.