Forecasting Bitcoin Prices Using Market Capitalization and Volume

The emergence of numerous cryptocurrencies has attracted individual and company investors. Bitcoin is one of the most famous and valuable Cryptocurrency. Investors and business analysist are trying to determine the best investment decisions that can result into high returns. Determining the expected price change (rise or fall) is a major factor in Cryptocurrency investment (Borgonovo, et al., 2018). The change in prices can be predicted/forecasted using the historical data. The other factor that can be determined is the relationship between price and other factors such as volume and market cap. Therefore, the research question that we will seek to answer is: Can we forecast the price of Bitcoin using the market capitalization and volume?

The data that that has been used for analysis consists of the information about the historical prices of Bitcoin Cryptocurrency. The attributes of the data include the average daily price of Bitcoin, the daily volume of Bitcoin and the daily market capitalization. All the attributes are numerical in nature and measured in terms of United States dollars. The data has 59 data points (entries) representing two months between: Jan 24^th 2019, to March 23^rd 2019.

The diagram below represents a multiple line graph of the three attributes: The price, volume and market capitalization of Bitcoin for the two periods. The multiple line graph shows the daily movement of the three attributes as well as the trend over the two months. From the graph, it is clear that the price of Bitcoin was relatively constant over the two months period. On the other hand the volume and market capitalization of Bitcoin experience a slight but consistent increase over the two months period.

The descriptive statistics analysis provides the summary statistics of the attributes. The summary statistics are shown in the table below. From the table below, it is clear that the average daily price over the two months period was $3776.48, the average daily volume was $7842095709.51 and the average daily market capitalization was $66350450021.17. The minimum daily price was $3411.09 while the maximum daily price was $4067.51.

Correlation analysis outlines the relationship between the attributes. From the results below, the correlation coefficient between the daily price of Bitcoin and the daily volume of a Bitcoin is 0.869188. The correlation is a strong positive correlation. The correlation implies that an increase in the volume of a Bitcoin by 1 unit will cause a corresponding increase in the price of a Bitcoin by 0.869188 units and vice versa (Bolin, 2014).

Description of the Data

The correlation coefficient between the daily price of a Bitcoin and the daily market capitalization is 0.9825. The correlation is a strong positive correlation. The correlation implies that an increase in the market capitalization by 1 unit will cause a corresponding increase in the daily price of a Bitcoin by 0.9825 and vice versa (Bolin, 2014).

The correlation coefficient between the daily market capitalization of a Bitcoin and the daily volume of a Bitcoin is 0.865859. The correlation coefficient signifies a strong positive correlation. The correlation coefficient implies that an increase in the daily volume of Bitcoin by 1 unit will cause a corresponding increase in the value of market capitalization by 0.8659477 units and vice versa (Bolin, 2014).

The appropriate analysis technique that has been used is the multiple regression analysis. The multiple regression analysis has been done to predict the daily price of a Bitcoin using the daily volume and the daily market capitalization. The output of the multiple regression analysis is shown below.

The summary output demonstrates the multipole regression coefficient (R) is 0.999944. The Multiple regression represents the gradient or the rate of change of independent variables (volume and output) with respect to the dependent variable (the daily price) (Bolin, 2014).

The value of R square is 0.999888701. The value of R square demonstrate that the regression model explains 99.9888701% of the population. Therefore, it is clear that our model is a perfect model that can be used for making statistically accurate inferences about the population (Bolin, 2014).

The table below is the output of analysis of variance (ANOVA). Analysis of variance is statistical tool that investigates whether there is any significant difference in the mean values of the attributes or variables. Therefore, the output below investigates whether there is any significant difference in the average daily price, volume and market capitalization of Bitcoin. The significance value is 1.2E-111 which is less than 0.05 hence it is clear that there was a significant difference in the average values of the three variables.

The coefficients table below outlines the specifications of the model. From the table below, the model for predicting the daily price of Bitcoin can be derived as follows:

Daily Price = -2.76E-09 (Volume) + 5.72E-08 (Market Cap)

The figure below is the output of normal probability resulting from the regression model. From the figure below, it is clear that the predicted values of daily price are not normally distributed. A normal distribution displays a bell- shaped curve.

Multiple regression analysis is appropriate in this case because the problem at hand is a prediction problem. Furthermore, the data set contains more than one variable and hence a multiple regression analysis is appropriate (Bolin, 2014).

The purpose of the report was to investigate the relationship between three variables of Bitcoin: The daily price, the daily volume and the daily market capitalization. The existence of the relationship among the three variables was to be used to further investigate whether a model for predicting daily price of Bitcoin can be developed.

The results of analysis clearly demonstrates that the three variables are strongly correlated. Similarly, the results demonstrate that the model for predicting the daily price of a bitcoin given the daily volume and the daily market cap is given by:

Daily Price = -2.76E-09 (Volume) + 5.72E-08 (Market Cap)

Therefore, the model can be used to solve the real-world challenge of making informed investment decisions. The model can be used by business analysists and individual investors in forecasting the possibility of fall or rise in the price of a Bitcoin. Knowing whether the price of a Bitcoin will fall or rise at any given time can lead to appropriate investment decisions.  For example, when the prices are expected fall, an investor may sell their Bitcoins in order to avoid suffering the associated loss. On the other hand, when the price is expected to rise, the investor may buy more Bitcoins in order to gain more profits.

The limitation of this data analysis is that it there are other factors determining the price of a Bitcoin that have not been taken into consideration. For example, the prevailing interest rates and the political stability of the economy which have been assumed to be the intervening variables. The recommendation for using the model is that it must be used only with daily data. Therefore, any user must be cautious not to use any other period of data as that may result into erroneous results.