Falling Cryptocurrencies-Which Is Resistant to Market Trend? Find Out Using Data Analysis
Find out how Top 200 Cryptocurrencies are Resistant to the overall Crypto Market Trend using Data Analysis.
You must have come across a lot of articles on how the cryptocurrency market is performing poor and skeptics triumphing on the falling crypto market. Some may argue that there is a bounce back from Bitcoin however, this article is NOT to debate how good or bad the crypto market is performing.
This article is to understand how currencies are dependent over the overall market and find out which currencies are most resistant to the general market trend. I will share the code also, you can carry out further analysis as well.
Why is it Important to find which currencies are most resistant to the market movement?
If you visit https://coinmarketcap.com/ you will notice that all the currencies have a similar trend visually unike other financial markets where you see quite a lot of disorder. Possible reason could be that the same market trends affect all cryptos to a greater or lesser degree.
The vast majority of them are heavily dependent on the performance and price of bitcoin (BTC). So its good to know which do not follow the market to manage risk if you are a crypto investor.
In this article we will try to explore the following-
- How Crypto Currencies are correlated to the market
- How Crypto Currencies are correlated to bitcoin(BTC)
- How Crypto Currencies are correlated with each other
If you are interested in the code you can jump and view the Repo Github
Few assumptions before we go ahead-
We’re only going to look at coins on CMC that have at least 120 days of historical data. We will be only focusing on the top 200 coins by market capitalisation.
We will be considering the data for the time period of July 1, 2017 to July 1, 2018. You can carry out this analysis for a different range of time period as well.
How Crypto Currencies are correlated to the Crypto Currency Market?
We don’t want to double count the Crypto Currency’s market capitalisation, so we will subtract Crypto Currency’s market cap from the total market cap.
Rest of the Market Cap = Total Market Cap — Individual Market cap
We can calculate the Pearson correlation coefficient between the two over the entire time period.
A positive coefficient reflects positive correlation & negative coefficient reflects negative correlation.
We need to calculate the correlation coefficient between the top 200 currencies and the corresponding rest of the market cap. Then we can plot a histogram and density plot with all of the correlation coefficients.
From the plot you can clearly see that most coins in the top 200 are highly correlated with the market.
- 75% of the top 200 coins have a correlation higher than 0.6
- 50% of the top 200 coins have a correlation higher than 0.8
Though we are interested in the coins with the least correlation. The following observation was made-
How Crypto Currencies are correlated to the Bitcoin?
Similarly, we can calculate the coefficient of the top 200 Currency coins with the Bitcoin(BTC)
The belief is held True. The top 200 coins by market cap are closely correlated with Bitcoin (BTC).
- 75% of the coins in the top 200 have a correlation higher than 0.4
- 50% of the coins in the top 200 have a correlation higher than of 0.6
We can derive the coins with the least correlation as well. The following observation was made-
How Crypto Currencies are correlated with each other?
For this we will pick up the top 20 currencies and plot a correlation table with gradient colour to get a rough idea how dependent the currency market caps are dependent on each other.
The following table represents the correlation between Top Crypto Currency’s Market Caps-
Key Takeaways -
These observations suggest that when the market goes up, most coins are also likely to go up. And when the market goes down, they are bound to follow. A diverse portfolio should have holdings of crypto coins which follow the opposite.
Caution : Correlation changes over time
The Pearson’s correlation coefficient derives a single number for the entire time period for which we considered. It does not account for changes over time, a rolling average comparison might give more deep insights. However, This observation is good enough to take describe the behaviour of the cryptocurrency coin.
I will try to do a follow up article to carry out more analysis and give deeper insights.
