Correlations of Cryptocurrency Indicators and Google trend data

Correlations of Cryptocurrency Indicators and Google trend data

In this article we discuss the correlation of cryptocurrency indicators regarding price and volume and google trend data. Furthermore, it is extension of an article I was writing about a year ago.

Additional disclaimer: This is no investment advice or encouragement to buy Crypto. Please, do your own research before investing. This tutorial is only for educational purposes and therefore meant to explain simulation concepts using Python. Furthermore, all mentioned cryptocurrencies are just for illustrative purposes.

Google Trend data

Google trends is a Google service that provides information which search keywords are entered. Furthermore, the results are normalized with the search volume. Hence, we can use the service to analyze trends regarding the popularity of search queries.

Correlation Coefficients

First of all, we use correlation to measure the relationship or association between datasets. Therefore, correlation coefficients are used as a numerical relationship measure. Practically this means, a correlation of +1 would represent a perfect correlation. Additionally, a correlation coefficient of -1 represents a perfect negative relationship. Furthermore, a correlation coefficient of 0 represents a very weak relationship.

In the following, we are using the Pearson correlation coefficient.

For further details, I refer to this article I wrote in Mai 18.

Python

You can view and download the entire Python code in this Github repository.

Simple Moving Average Indicator

In the following we smooth all the date with a 14 day simple moving average.

Correlation of Price and Trend

The following analysis is conducted on the 4th of November 2018, with 4 different time periods (50, 100, 150, 250 days). Furthermore, the cryptocurrencies BTC, ETH and XRP are considered in Dollar and Euro on the exchange Kraken.

First, let us have a look at the cryptocurrency prices and the Google trend data (100 days, Ethereum (ETH) traded in USD).

Furthermore, the data of 24 data sets is entered in the following table.

Time period
in days
Fiat
(Kraken)
Crypto
(Kraken)
Correlation
Coefficient
50EURBTC-0.1387
100EURBTC0.4928
150EURBTC0.5011
250EURBTC0.6312
50USDBTC0.4927
100USDBTC0.4766
150USDBTC0.5183
250USDBTC0.7192
50EURETH0.574
100EURETH0.4799
150EURETH0.2593
250EURETH0.6611
50USDETH0.6505
100USDETH0.4761
150USDETH0.2536
250USDETH0.6957
50EURXRP0.3894
100EURXRP0.589
150EURXRP0.2676
250EURXRP0.678
50USDXRP0.4718
100USDXRP0.6097
150USDXRP0.2697
250USDXRP0.7113

Observations
First, we usually observe a better correlation coefficient with longer time periods. As we can see, there is a positive correlation. But, that can not be concluded in general. Furthermore, the mean and standard deviation is calculated for all data in the table. In conclusion, there is a mediocre positive correlation with a large standard deviation.

  • Mean Correlation Coefficient: 0.489
  • Standard deviation Correlation Coefficients: 0.198

Correlation of Volume and Trend

In the following graph you see the exemplary volume and the Google trend data of 250 days, Bitcoin (BTC) traded in EUR.

Furthermore, the data of 24 data sets is entered in the following table.

Time period
in days
Fiat
(Kraken)
Crypto
(Kraken)
Correlation
Coefficient
50EURBTC0.809
100EURBTC0.9386
150EURBTC0.933
250EURBTC0.953
50USDBTC0.9368
100USDBTC0.9365
150USDBTC0.8764
250USDBTC0.9659
50EURETH0.9853
100EURETH0.6984
150EURETH0.633
250EURETH0.7718
50USDETH0.9785
100USDETH0.607
150USDETH0.5493
250USDETH0.7924
50EURXRP0.9952
100EURXRP0.9902
150EURXRP0.9866
250EURXRP0.9159
50USDXRP0.9938
100USDXRP0.9848
150USDXRP0.9797
250USDXRP0.9526

Observations
The mean and standard deviation is calculated for all date in the table. It is obvious, that the mean correlation coefficient of all data is higher compared to the price comparison. Furthermore, the standard deviation is smaller.

Additionally, the correlation between volume- and trend data is stronger for Bitcoin (BTC) compared to Ethereum (ETH). Furthermore, the strongest correlation coefficients were observed for Ripple (XRP).

  • Mean Correlation Coefficient: 0.882
  • Standard deviation Correlation Coefficients: 0.136

Conclusion

  • The overall correlation coefficient regarding volume (average: 0.882) is significantly higher compared to price (average: 0.489).
  • Therefore, Google trend and volume data is clearly stronger correlated. Hence, we can use it to grasp the cryptocurrency market sentiment.
  • The magnitude of the standard deviation is an indicator for the overall quality of the correlation coefficient. Hence, the larger the standard deviation the more unreliable the correlation.
  • Generally, you can use simple moving averages (SMA) to smooth out data.
  • A different time window, different currency or even a different Google trend keyword can alter the results significantly.
  • Furthermore, fiat currency fluctuations can have an impact on the result.

What do you think?

I’d like to hear what you think about this post.

Let me know by leaving a comment below and don’t forget to subscribe to this blog!