The Economics of Filecoin
At Blockstack, we’re thinking very hard about crypto-economic system design, particularly as we approach the release of our own mining system and token sale for the Blockstack network.
Bitcoin, Ethereum, Zcash and Tezos have been the networks that we’ve studied most extensively, given that they are truly independent systems with differentiating factors, strong teams, and very promising market opportunities.
Beyond these three, Filecoin particularly intrigued us as we see it shares the aforementioned properties as well. And with the Filecoin sale coming up, I thought this was a great opportunity to share some analysis we’ve been doing on the economics of the system and how it compares to Bitcoin, Ethereum, Zcash and Tezos.
Supply
The Filecoin supply has four components: the miner supply, the Protocol Labs supply, the Filecoin Foundation supply, and the sale supply.
The miner supply follows an exponential decay curve with a half life of 6 years. We’ll explore that some more below.
The Protocol Labs supply and the Filecoin Foundation supply each vest over a period of 6 years.
The sale supply, meanwhile, has variable amounts of vesting. The main sale has vesting over a period of 6 months while the advisor sale allowed participants to vest over periods of 1, 2 or 3 years. For the purposes of this graph, I averaged the vesting out at 1 year.
The total number of Filecoin that will ever be created is 2B.
Supply as a % of Supply After N Years
As mentioned earlier, the Filecoin miner supply schedule follows an exponential decay curve with a half life of 6 years. This is similar to the Bitcoin supply schedule, but with two important differences.
First, the half life of Filecoin is 6 years while the half life of Bitcoin is 4 years. Second, while Bitcoin’s exponential decay is sampled every 4 years, Filecoin’s is sampled every month, leading to a smoother drop-off in the rate of coins mined in each block.
You can see from the graph below that Filecoin is a smoother variant of the Bitcoin exponential decay curve. Note that the graph uses yearly sampling, so some of the smoothness of the Filecoin curve could not be adequately captured.
Note that while Filecoin, Bitcoin, & Zcash all define a finite number of coins that will ever exist, Ethereum does not have a defined cap on the number of coins. This means that “% of final supply” is not a metric we can use reliably to compare anything against Ethereum. For this purpose, we decided to use “% of supply after N years” instead.
Annual Supply Growth
The growth in the supply of Filecoin is approximately that of Bitcoin for each year that passes, with two exceptions. First, the curve is smoother as it is sampled every month instead of every 4 years. Second, the Filecoin supply growth includes the amount that is being sold in the sale and distributed to Protocol Labs and the Filecoin Foundation, which means that a greater amount is released in the first year.
You can see the annual supply growth in the chart below:
Please note that the Ethereum supply numbers are completely speculative after year 3. Empirically the miner supply increased the total supply by about 11M coins in year 1 and 7M coins in year 2. This number has been decreasing and rapidly approaching 0 as a result of the fork bomb. After Ethereum has migrated over to proof-of-stake, Vitalik has said that the annual inflation rate will be somewhere between 0.5% and 2.0%. I’ve chosen to go with 2.0% for now.
Creator Ownership
The Filecoin creator ownership (factoring in the amount going to both Protocol Labs and the Filecoin Foundation) is somewhat higher than the known creator ownerships of Ethereum and Zcash and the estimated creator ownership of Bitcoin (Satoshi’s holdings).
Whether that’s a good thing or a bad thing is up to the community to judge. One could say it’s not good for the creators and associated entities to take up such a sizable portion of the protocol tokens. On the other hand, this could be a very good thing for the system if a large portion of the Protocol Labs tokens are reserved for a future financing round, as it would further capitalize the business and provide them the extra fuel they need to compete with AWS.
You can see creator ownerships over time graphed below:
We can break the creator supply amount down further and see that 25% of this amount goes to the Filecoin Foundation while 75% of this amount goes to Protocol Labs. That means that Protocol Labs will end up with 15% of the total final supply while the Filecoin Foundation will end up with 5% of the total final supply.
Coin Price
As a bonus, here’s an analysis of the Filecoin price and what it means for your ownership of the network, as benchmarked against the figures for Bitcoin, Ethereum, and Zcash.
Assuming a Filecoin price of $2, it would cost $37.22 to buy 1 one-hundred-millionth of all of the Filecoin tokens that will be minted over the next 20 years.
Meanwhile, it would cost $42.43 to buy the equivalent percentage of the Zcash network. Similarly, to buy equivalent percentages of Bitcoin, Ethereum, and Tezos, you’d have to spend $684.92, $345.31, and $3.18, respectively.
Here’s a graph that further illustrates the difference:
Conclusion
The Protocol Labs team has taken quite a few learnings from its predecessors and incorporated them into the design of an entirely novel protocol.
Protocol Labs has taken the exponential decay curve of Bitcoin and improved upon it. They’ve created some interesting dynamics with their token sale. And they’re setting a new precedent with CoinList.
Disclosure: I participated in the Filecoin advisor sale and I’m a big fan of Juan and the Protocol Labs team. The article above contains my personal analysis only, not investment advice. Information I have about Filecoin and Protocol Labs may be inaccurate and should not form the basis for any investment decision.
If you enjoyed this article, please take a moment to recommend it. Also you can follow me on Twitter if you’re interested in more thoughts on tokens, blockchains, and decentralized apps.
Thanks to Juan Benet, Zooko and Patrick Stanley for reviewing this post.
