Ether Historical Data Brief

This brief highlights the following 3 themes that I find interesting in Ethereum right now.

  1. Price Change and Transaction Count
  2. Transaction Count and Days of the Week
  3. Transaction Fees

These themes highlight the cryptoasset Ether and its blockchain of Ethereum. Updated on and data up until 12/13/2017.

This brief is based on a more full report of earlier data exploration here: http://rpubs.com/RichardJamesLopez/Ether_Historical_Data_11_2017
========================================================

1 - Price Change <> Transactions

The relationship for # of transactions to price change is a slight twist on volume to price analysis. As a sort of an unadjusted proxy for volume, I propose a # transactions. The notional amount of Ether can rise disproportionately than to the # of transactions and can overemphasize the volume as a result. For this reason, I think the # of transactions is a good proxy.

To start off this analysis, I subset the data. 8/11/15 is a proper starting data in order to remove the noise of the data from the first month of trading and indiscernible % changes.

A scatter plot for the data subset shows that there is clear behavior of price change for each day with more than 600,000 transactions.

In another plot, we zoom in on the dataset and remove 5% of each tail:

With this new scatter plot, a new pattern emerges. With the intuition that the # of transactions is increasing over in time as a result of popularity of Ethereum and the evidence of past Ethereum data, there are a handful of data points that indicate positive % change for particularly active days. I’ve gone ahead and boxed in the data for transactions above 600,000. From this analysis, there appears to be a threshold for # of transactions that have positive indications. This is a trend worth monitoring, especially as Ether has any drawdowns.

Of note:

  • this is not an exact proxy for volume, this is just the sheer count of transactions for the day
  • both positive and negative tails of 5% have been removed
    • however there were no negative tails realized in this time period (see first chart in section with all transactions)
  • these are end of day transaction counts

Transaction Count data is relatively transparent and its availability is one of the advantages of a public blockchain. In the reference section there are a couple of useful sites for real time Transction data. Please suggest other useful sources at @RichLopezNY or richardjameslopez[at]gmail

2 - Transactions <> Day of the Week

This chart takes an earlier chart that had Block Size and simplifies the relationship by focusing on Transactions.

I come back to this chart because it gives me a rough bifurcation of high and low transaction count days of the week.

In particular, the plot corroborates 2 things:

  • the 75th quantile is lowest for Saturday and Sunday, and
  • the highest amount of transactions for both Saturday and Sunday are lower than the highest amount of transactions for all the other day of the weeks.

We can also make observations about the busiest days being Thursday and Friday based on the 75th quantile points as well as the 2 data points with the most transactions also being on both of these days. However, what stands out to me is that there are less transactions on the weekends and this is useful for a couple of reasons:

  • if there is particularly high count of transactions on a weekend, there may be a peculiar dynamic taking shape for a deep dive
  • for those individuals who want to add to their positions on relatively calm days under normal circumstances

Pairing this data with the observation that more transactions can correlate with higher price volatility, an individual can use this day of the week analysis in a couple of different ways.

3 - Transactions Fees

Based on some earlier feedback from the Twittersphere, @SGBarbour suggested that I check out Transaction Fees. To start off, I decided to look at some correlations to develop an understanding of the variable dynamic.

From these correlations charts, there is only one of these variables that has a correlation more than 0.8 and that is the # of transactions themselves. While that may be interesting, the Transaction Count has already been analyzed earlier. Plotting Transaction Fee over time may be another starting point.

This scatter plot highlights the growth of traffic on the blockchain. Towards the end of 2017, the amount of Ether used in form of transaction fees is regularly creating new highs. Some may want to use this as a sign of robustness for the network, others may want to blame the cryptokitties. I do neither.

(side note: I cannot blame the cryptokitties for 2 reasons- 1) I do not fully understand Breeding Fees that may play into Transaction Fees , and 2) I cannot disparage cuteness in any of its forms).

To normalize this analysis on the network, averaging the Transaction Fee data over the amount of transactions for a day gets the following scatterplot.

This scatter plot helps orient Transaction Fees on an average basis. There has been discussion on the viability of the blockchain if Transaction Fees are not reliable and this analysis adds color to the discussion. While not 100% predictable, Transaction Fees appear to show a reliable state between the range of 0 and 0.005. This may be reassuring for some or untenable for others. Should the Fees for 12/2017 continue to increase into 2018 at this rate, I may do a deeper dive into a possible trend.


Of Note:
To those who may comment that these Transaction Fees are actually substantially higher because Ether is rising in USD$ terms - this is correct. However, in the true spirit of being crypto adopters, I maintain that it makes sense to view in terms of Ether itself. If an individual converts all this data into USD terms (or local currency terms), a very different analysis takes place and a lot of this data gets skewed. In any case, I can see both sides, but lean towards valueing the Transaction Fees in Ether terms.

The main takeaway for this analysis is that Transaction Fees are something to be monitored going forward. With the most recent popularity and increasing volume of Ether, it may seem that the network is being pushed to its limits and the fees will increase as a result. However, based on Ethereum’s history, Transaction Fees are not quite at the highest rates it has seen in the past. In the coming months, fees may pass past highs, but are not quite there yet.

About

I foresee updating this analysis every month or so. I am curious to see if these patterns, behavior and relationships hold over time as this data set continues on. Reach out to me at @RichLopezNY or richardjameslopez[at]gmail if you have any ideas on what I may be missing.

Full Disclosure - this analysis is not meant to be used for investment recommendations in any way. This is simply exploring data.


Donations kindly accepted at:
0x76276793c2A8dEB6F51eb75C43a9aDfCBb78C195

References

Current repository of dataset - https://etherscan.io/charts
Necessary primer to understanding Cryptoassets - https://www.amazon.com/Cryptoassets-Innovative-Investors-Bitcoin-Beyond/dp/1260026671
Glossary - http://ethdocs.org/en/latest/glossary.html
More Technical Glossary - https://bitsonblocks.net/2016/10/02/a-gentle-introduction-to-ethereum/

Sources for Transaction Data:
https://www.etherchain.org/
- gives latest data on a TPS (Transactions Per Second) basis
https://www.ethernodes.org/network/1
- provides granular data for nodes that host transactions
https://coinmetrics.io/charts/
- lets the user play around with their own correlations