Root
Press Crypto
Do your own research - How do I analyze the on-chain behavior of Bitcoin?

Do your own research - How do I analyze the on-chain behavior of Bitcoin?

Do repost and rate:

Rated:131

In our search for new knowledge about Bitcoin, data analysis can serve us well. In this article we will do a little data analysis together. The focus is on the relationship between the Bitcoin rate and the transaction volume.

Do your own research is a popular motto in the crypto field.

Own research also includes a sober look at the raw data. The responsible crypto citizen should always be a little data analyst. In the last article we took a first step in this regard: In the first part of this series, I explained how interested parties can download and work with course data. With TradingView and APIs like Coingecko's, investors and analysts have come a long way. You can view Bitcoin & Co. courses according to the current rules of technical course analysis and study new metrics and indicators. You can also study investment strategies such as dollar cost averaging (regular investments in smaller, fixed dollar amounts) and the use of seasonality or create a portfolio based on several cryptocurrencies.

However, the course does not take into account a special feature of cryptocurrencies: Thanks to the public nature of the blockchains, every analyst can take a look on-chain himself. How many transactions were there on the blockchain? How full are the blocks? Analysts can clarify such questions on-chain at a glance. In this article we would like to explain a few ways in which you can get such data using APIs.

A second goal of this article is to do a little exploratory data analysis together. We therefore start with the data acquisition and play together with the corresponding data. Let's see what we find out.

Sophisticated Bitcoin data from Coinmetrics

Coinmetrics is a good service for a first look at data beyond the mere price events. Coinmetrics is a company that focuses on crypto-relevant analyzes. Their analyzes have been discussed in various articles in the past. In addition to these insights, Coinmetrics offers the community a large amount of data that can be the first step for many analyzes.

Even those who don't want to get started with R right away can get started on their platform. Under this link, interested parties can access a frontend in which various data can be viewed with regard to their development over time:

As you can see, a wide variety of other cryptocurrencies can be examined in addition to Bitcoin. The above illustration shows the development of the Bitcoin exchange rate. However, sizes such as the realized capitalization, the NVT ratio or the current block size can also be selected. If that is not enough for you, you can use “Formula” to create combinations of the sizes mentioned and plot them graphically. The future Bitcoin analyst can go very far here and model simple trend forecasts, for example.

But what if you want to use other models or look at the Coinmetrics data together with others? Or if data should not simply be analyzed in terms of its time dependency? Let's say that we want to analyze the relationship between Bitcoin's number of transactions and its rate in more detail. We are not interested in the temporal context, but what an influence the number of transactions has on the Bitcoin rate.

Here we can use the Coinmetrics API. For those who are not fans of the JSON format, the data can be output in CSV format. We can receive our data on the number of transactions and the Bitcoin rate via this API call:

https://community-api.coinmetrics.io/v2/assets/btc/metricdata.csv?metrics=PriceUSD,TxCnt

Actually, you can almost read the API call above. We request data for the Bitcoin (btc) asset in the form of a CSV file (metricdata.csv), specifically the Bitcoin rate (PriceUSD) and the number of transactions (TxCnt).

Tidyverse for beautiful illustrations

So far, so easy. How do we now represent the above relationship graphically? In principle, we could use the plot command as we did last week. However, I want to introduce aspiring analysts to the Tidyverse.

As mentioned last week, the Tidyverse is a collection of libraries that make working with data in R easy. Especially in the field of exploratory data analysis, i.e. if you want to play something with the data first, the Tidyverse is a very pleasant aid. With it, the preparation, analysis and visualization of the data is very easy.

In order to be able to work with the Tidyverse, we must first call the corresponding libraries. To do this, we proceed as in the last week and call the Tidyverse using the library command:

library (tidyverse)

By the way: If there is an error message here, you will most likely have to install the appropriate libraries. You can do this with the command install.packages:

install.packages (tidyverse)

Now we can work with the data. With a simple command we pull the data:

read_csv ("https://community-api.coinmetrics.io/v2/assets/btc/metricdata.csv?metrics=PriceUSD,TxCnt")

We could drag the data into a variable like we did last week:

data <- read_csv ("https://community-api.coinmetrics.io/v2/assets/btc/metricdata.csv?metrics=PriceUSD,TxCnt")

Bitcoin rate and number of transactions under the microscope

Now we can play around with the data. There is a direct way to do this in Tidyverse: With the operator%>%, data can be passed on from one command to the next. Let us use this to show the desired context:

data%>% ggplot (aes (x = TxCnt, y = PriceUSD)) + geom_point ()

We use it to produce a rough diamond. At the moment it still looks like a rather wild point cloud, not to mention the axis labels etc. But we are still at the beginning of our analyzes:

In both cases - Bitcoin's rate and number of transactions - we consider sizes that span several orders of magnitude. In order to have a better view of both smaller and larger values, a logarithmic representation is recommended. This can be done with commands called scale _ * _ log10, where x or y can be entered for *. The good thing about this command is that we can use it to implement slightly better axis labels and set limits. Overall we come across this extension:

data%>% ggplot (aes (x = TxCnt, y = PriceUSD)) + geom_point () + scale_x_log10 ("number of transactions", limits = c (30,1e6)) + scale_y_log10 ("Bitcoin rate in US dollars")

on a very interesting chart:

Overall, that looks like an exciting linear relationship. The Bitcoin price increases with the number of transactions. He follows a corridor. This visualization confirms a little of Metcalfe's Law: The value of a network increases with the users.

Again, this article should provide a small introduction to how you can better understand Bitcoin & Co. with the help of data analysis and visualization. And this article is also intended as a start: How do such relationships look with other crypto currencies? Should the price development be considered separately before the introduction of the Lightning network? Can the relationship visualized above even be modeled? Question after question. As is so often the case in research, there are more locked doors behind an open door.

I wish every reader a lot of fun opening further doors.

Source: Shutterstock

Don?t Miss MY SWISSBORG PROMOTION!

Swissborg gives everyone who register to the wealth app and deposit 50 € (depositing FIAT is recommended) a reward ticket with a random value ranging from 1 to 100 € in BTC.

But only, if you use my links and follow the described steps below. If you do so:

I give you 50 % of my reward ticket i am getting from your subscription for free!

All you need to do is this:

1. Download the Swissborg Wealth App (mandatory)
2. Deposit 50 €, deposits on Swissborg are free so you can withdraw them later. (Withdraw costs about 1 €, with the earning from the reward you'll gonna make either a plus or a break even) (mandatory)

Note: If you only download the Wealth app you only get 10% of my ticket

3. Download the Swissborg Community App (mandatory)
4. Apply the Code: OA62FUI (mandatory)
5. Contact me via Discord (Overmind#0055), Telegram (@SwissborgSwarm) or mail ([email protected]) (preferred) for verification and to get your BTCs! (mandatory)

Note: If you only download the Community App and apply the code i say thank you!

Have fun earning free BTC!

Do not miss the chance to get up to 150 € in BTC for free!

This promotion begins on January 2021-01-01

I Paid every participant: Check in the comment section of these posts for a proof of my payment to them:

Proof 1

Proof 2

Proof 3

Proof 4

Proof 5

Thank you for participating!

Note: I pay in any currency you like :). And the withdrawal fee from Swissborg Wealth App to your wallet are free of fees. I pay them. Happy free Crypto earning!