Bloom Tokenomics Research

Objective

We want to see if there are discernable patterns between a project’s tokenomics and their short and medium term outcomes.

For projects with a meaningful market capitalization:

  • What was their initial token distribution method? (ICO, airdrop, other?)
  • What was the starting number of tokens?
  • What was the initial price?
  • What was the initial market cap?
  • What was the governance model?

Can we link the characteristics to useful outcomes?

  • Market capitalization (24 hours, 1 week, 1 month)
  • Number and growth in the number of users
  • On-chain metrics (transactions, contract interactions)
  • Community support (Twitter or Discord growth)
  • Locked Token Value or other Staking metrics

Data Curation

To begin, we had to establish a set of tokens. This set includes, among other things, protocols, “ETH killers”, exchanges, etc. We chose to ignore airdrops for this analysis. With the initial set of tokens established, we also had to identify the characteristics of each that we wish to collect. These characteristics include: A sample of the data is shown below

Now we can begin a deep dive in order to compare various metrics across the factors we’ve collected.

Analysis

Initial Exploration

We immediately get our first set of results that show there is a concrete differentiation between the market adoption of ICO, airdropped and mined coins. Airdrops start with higher market caps, but overtime fade.  That was true for all except Uniswap, which was the outlier in the last graph. ICO/Public sales start off with the highest market caps, dip, then may recover over time.  On average, their marketcaps are always the highest. Liquidity Mined coins start off the lowest, as they struggle to gather holders and liquidity, but over time begin to out perform with increasing market adoption.

Social media adoption differs across the groups, and there is a clear relationship as to how a coin’s socials is correlated to their price. Airdrops have the second most followers after ICO supported coins, yet this method has a higher average followers than those coins that start through liquidity mining. This offers further evidence that liquidity mining is the equivalent to bootstrapping a new community from the ground up, considering they have some of the best outcomes over time.

While this was a great start, some of the characteristics were too onerous to track across every coin. Some of these metrics included the number and growth in terms of users, on-chain metrics like transactions, and locked token value as well as other staking metrics.With this in mind, we took a detour to do some tooling that would not only allow us to investigate other relationships within tokenomics, but give us actual software that we could open source and share.

Holding vs. Market Capitalization

What can we say about the relationship between the number of wallet holders and market capitalization? One of the main points of an initial distribution is for the tokens to get distributed to as many of individuals as possible.  The idea is that the token’s price and network will be more secure due to the distributed and decentralized nature of the participants. To look at this, we looked into Etherscan’s token wallet list to see if we could find some quality data. Etherscan just so happened to have some relevant data: https://etherscan.io/tokens

We collected the data from this page and cleaned it into a parsable CSV, a screenshot of which can be seen below.

With this data, we were able to create a 2D matrix that compares market cap with number of wallet holders. We divided the chart into quadrants according to the following:

  • Quad 1 (Blue) – lots of holders, high mcap
  • Quad 2 (Green) – lots of holders, low mcap
  • Quad 3 (Orange) – few holders, low mcap
  • Quad 4 (Red) – few holders, high mcap

First the graph does confirm what we’d hope and suspect, the most decentralized projects (as measured by the number of holders) are indeed the same ones with the highest market capitalizations.  We see some of our well known pals there: UNI, USDT, BNB, and LINK.

We deployed an interactive version of this application available here: http://crypto.omnianalytics.io/apps/etherscan/ at this URL, you can explore individual coins, and add lines of best fit to see empirical and theoretical relationships between market capitalization and the number of holders, and more. The plot highlights the positive correlation between the two factors, and is segmented into four regions colored according to the perceived risk to the token holders. We assume that tokens with few holders but a high market capitalization are more susceptible to a sudden reduction of liquidity.

Generalizing the Analysis

What if we could expand the power of our inference by generalizing the analysis? To do so, we began investigating the initial liquidity provided by the brand new Uniswap listings. We want to see if there are discernable patterns between a project’s tokenomics and their future market outcome. Instead of focusing on 20 coins, can we define, collect and then track information on a larger population of tokens?

To do so, we created a smart contract listening bot that collects information on new contracts deployed to UniSwap. Note, however, that we’re not really covering airdrops, ICOs or liquidity mining distribution approaches anymore, but looking solely at new liquidity pools.  A pool could have been created for any of the above three purposes, that’s why the analysis from here on out will be general and not specific to any one distribution type. The dataset captures information on the pair, the amount of liquidity, the number of tokens, and other information around transactions and volume.

Now, for each token tracked we have a very clear picture of:

  • What was the starting number of tokens?
  • What was the initial price?
  • What was the initial market cap?

Which will allow us to track…

  • Market capitalization (24 hours, 1 week, 1 month)
  • Number and growth in the number of users (as measured by available liquidity)

Uniswap Analysis

Contract deployments likely follow some generalized gamma- Poisson  model where the number of contracts initiated follow some gamma distribution with an estimated Poisson rate •That rate parameter is estimated by the average number of deployed contracts per day, 73.96.

In terms of the daily number of contracts deployed over time, the overall trend since the end of February has been increasing, but there was a clear peak in the late part of April. Since then there has been a steady decline in the number of contracts deployed per day, but with a recent rebound in the middle of May.

The average amount of daily total liquidity per contract is a heavily right skewed distribution, with a few contracts adding a significant amount of liquidity, but the majority adding a much smaller amount. This is reflected in the measures of centrality, where the median is significantly lower than the average.

For the daily amount of total liquidity deployed, overall there is an upward trend in the amount of liquidity added in deployed contracts per day, this is despite the decrease in the number of contracts deployed. The daily amount of liquidity added onto UniSwap spiked early at around 85 million one day at the end of February. Since the start of our data tracking we’ve seen on average nearly 18 million per day.

The distribution of token amounts per deployment shows that token amounts at initial distribution vary wildly from 1 to more than a few Quadrillion. Most projects deploy between 10,000 and 1,000,000,000 tokens at the initiation of their contract. This analysis, however, only accounts for the tokens deployed on UniSwap, but not the total in circulation which could far outweigh the number deployed in the liquidity pool.

Finally, we can compare stable coins by looking at the number of contract pairs deployed over time with a pairing to a stable coin. On UniSwap, WETH has by far the most throughout the duration of the data.

Summary

  • Intuitively, higher market capitalizations are positively correlated with the number of wallets holding a cryptocurrency.
  • Within the sample we tracked, cryptocurrency projects with initial coin offerings typically had higher long term market capitalizations. Second were tokens dispersed through liquidity incentives and last were tokens distributed via an airdrop.
  • Contract deployment on Uniswap is growing, showcasing the increasing popularity of AMMs for liquidity provisions.

Leave a Reply

Your email address will not be published.