How Much ICP is Being Burned

An analysis on the computational burn

Is the Internet Computer ecosystem growing? And, if so, at what rate?

TL/DR - the ecosystem is growing linearly now, but that could change quickly.

Background

Applications on the Internet Computer require burning ICP into cycles to pay for computational resources. This reduces the supply of ICP and also provides a great metric to measure the growth of the Internet Computer ecosystem since the growth of the ecosystem should be closely correlated with the burning of ICP for computational purposes.

High Level Burn

The Internet Computer Dashboard provides an easy way to see the total ICP burned by day. That chart, copied below, can be found on the Circulation page. (Note - all data in this article is as of 26APR22)

From a high level it looks like January was an explosive month of ICP burning. This was caused by the launching of Sonic.ooo, which allows burning ICP into cycles for DeFi liquidity pools. Since the new year, Sonic has burned 50k ICP!

However, Sonic’s use case shouldn’t be included in the analysis if we are trying to measure the computational increase of the Internet Computer network as a way of measuring the ecosystem’s growth. Therefore, it’s necessary to remove the burning occurring from Sonic. Once Sonic’s burning is removed, the total computational burning looks like the chart below. (Note - data does not include the last 5 days of April)

Notice the huge amount of burn in the summer of 2021… that’s most likely caused by various cycle faucets used to attract new developers to the ecosystem. Looking to more recent data, it’s clear burning has been increasing from November until April, with an average month-over-month increase of about 500 ICP. This is great news and suggests that the ecosystem is growing in a somewhat linear fashion.

A Better Analysis

Unfortunately, total ICP burned is not a great metric (even with Sonic’s burn data removed) because the conversion rate of ICP to cycles is not fixed. Instead it fluctuates with the price of ICP. When the price of ICP is high, it converts to more cycles. When the price of ICP is low, it converts to less cycles.

We can account for this fact by charting the total USD value of ICP burned per day (basically the total ICP burned on a day multiplied by the price at the end of that day). When doing that, a truer computational burn metric is formed. (Again, data does not include last 5 days of April)

Accounting for the exchange rate, it’s still clear that the Internet Computer ecosystem is growing in a linear fashion so far in 2022.

What will cause more growth?

The best way to think about the Internet Computer ecosystem is on two dimensions: (1) the number of applications within the ecosystem and (2) the amount of usage within the average application in the ecosystem. There is a platform network effect between applications on the Internet Computer (the more applications there are, the more likely new applications will be built) and a system network effect within applications (the more a particular application is used, the more likely it is to be used more). Both of these network effects could easily produce an exponential growth curve. The fact that there exists the potential for exponential growth along two dimensions makes me believe that it is unlikely that the cycles burned metric will continue to grow linearly into the future. In fact, it’s likely that exponential growth in either dimension will lead to exponential growth in the other dimension. All this to say, once things start moving, they will be moving fast.

It’s key then to have metrics to measure both of these dimensions so that we can identify direction and signal. I still need to do some more thought on these measurements, but a few easy metrics stand out. For dimension 1 (number of applications) it seems logical that smart contract growth could be a rough metric. That chart can be found on the IC dashboard homepage.

Another more nuanced way to measure ecosystem growth would be the “smoothness” of the daily burn rate. Ideally there would be dozens or hundreds of applications burning ICP every day and the largeness of that pool would smooth out situations where an application or two burn a significant amount of ICP in a given day. The 7-day moving average curve (below) shows a considerable amount of choppiness, which suggests that the majority of the burn rate is coming from just a handful of applications. I’ll be watching this chart as time goes on to gauge the number of well-used applications in the Internet Computer ecosystem.

More Coming…

Keep an eye out for more information from me on a new product I am building intended to provide it’s users even more data and analyses for making informed decisions regarding their investments in the Internet Computer ecosystem and the ICP token.

In the meantime, feel free to share or subscribe.