How to make sense of Cryptocurrencies ? [ Part 1 ]
Frequently we hear about Bitcoin, Ethereum, Litecoin , Ripple which are the most dominant coins in the cryptocurrency market. Also Coinbase, the most popular exchange under the sun , allows us to trade only in BTC, ETH, LTC and has just added BCH. Once you start your journey beyond the popular coins, it is mind bogglingly complex world with a combination of stars, hypes, scams and junk.
Making sense of cryptocurrencies is one of the hotly debated topics now.
IMO, there are multiple ways we can get a sense of whether the coin is worth to keep in our watchlist / portfolio or not.
- Social media activity [ what coins say about themselves & what others say]
- How their price react to macro/micro economic events & news from the media
- Network analysis on the team / community influencers
- Bid/Ask , Orderbook data published by Crypto Exchanges (CoinMarketcap.com has listed 174 as of today )
- Age old Tech analysis especially for intra day trading
Doing a fundamental analysis on the crytpcurrencies is a herculean job. There are some papers available in the public domain [such as this ] but this is still a subjective area.
My belief is digital currencies will have a strong correlation to their social/digital activity. Unlike fiat currency , network for a cryptocurrency has to be built from scratch and endorsements by influential personalities [Jamie did a favourable job with his f-bomb and entered into CoinDesk 2017 hall of fame ], comments from the community etc. have a very strong correlation with their prices.
Over the vacation time, I was trying to get a handle on the wider cryptocurrency spectrum and tried to test above. This multi part article is a documentation of the crypto-fun that I had over last few days.
Fig 1. Social Activity of various coins sorted by their github, reddit , twitter and facebook activities.
I chose Cryptocompare.com as main data source for social activity comparison. Their API end point for BTC https://www.cryptocompare.com/api/data/socialstats/?id=1182 gives you wealth of social data about the Bitcoin. Once you get a handle on their coinlist API (https://www.cryptocompare.com/api/data/coinlist/) , rest is a cakewalk.
My hypothesis behind the ranking is :
- More forks in github means the project is more popular. Needless to say, Bitcoin is the most forked repository (also they have multiple repositories )
- More github subscribers means a great technical community behind the project
- More reddit posts/comments means there is a great buzz around the project
- More twitter activity means the team is very serious about getting the word out
- More facebook activity means it has reached mainstream
Needless to say, this is constrained by the data provided by Cryptocompare. I have not done validation of the underlying data. Their output has few glaring omissions like Bitcoin Cash which return no results for social activities. Perhaps https://www.reddit.com/r/btc/comments/6rj5eb/where_is_bitcoin_cash_github/ explains the github issues.
It would have been great to get market cap value from the same call so that we can co-relate the pricing , market dominance and the social attributes and also do some technical analysis on the pricing once we gather more price points. I wouldn’t crib about this as the API call is free and response time is impressive.
You can grab the entire file (ran this evening) at https://www.dropbox.com/s/0kdorwkquyjjus8/Socia%20Analysis%20of%20CryptoCoins%20-%20ver%201.1.xlsx?dl=0 and play around.
I used Anaconda Navigator for running Python Programs, beautifulsoup for scrapping data (especially from coinmarket cap which has limited APIs as of now) and SQLite for database.
Will post the source code into Github in the next few days.
If you enjoy reading this post, please leave a clap !