Should those untested developers be getting 10 million for that whitepaper and ICO contract? Should they be getting it all up front? What about the EOS Token Sale?
I think that after my last post, expressing some of my general worries around the EOS token sale, it might be a good time to take a deeper look into how token sales are conducted in general. After going through some information regarding sale mechanics I will revisit my comments on the EOS sale. Be sure to outline your concerns in the comments below.
Vitalik Buterin has recently posted a fantastic overview: analyzing Token Sale Models that is worth a read if you are technically minded. I will follow several points outlined by Vitalik and add my own comments; there was a lot in his analysis that directly sparked the writing of this post. Thanks to Vitalik for his reason, rationality and composure in his writing.
ICOs (previously just called token sales) are one of the most interesting concepts to emerge from the blockchain space as a whole. They are an incredibly intimate and easy vehicle into early startup funding. They allow the funding of an open-source protocols and software ecosystems to self sustain their development. We will only see more of them.
During the infancy of the last few years it has been a one-concept ICO landscape; there were capped and uncapped sales that sold a fixed number of tokens or as many as people were willing to buy respectively. Both these methods have been viewed as incomplete and potentially problematic by many members in the community. Many recent sales have tried novel mechanisms to alleviate as much criticism as possible.
It is interesting to see how responsive the space has become to the loads of criticism that aptly follows every popular ICO. Lots of this criticism has generally been of the theme of selling out too fast or that the developers are not properly incentivized towards their project.
What do people even want?
This is something Vitalik detailed extremely succinctly in his post. I agree with his assessment of the growing sample-set of ICO criticism and his distillation into the five basic desires detailed below:
- Certainty of valuation - If you contributed to a sale, you likely want certainty over the cap on the valuation. More simply you want a minimum to the percentage of the market cap your contribution grants you.
- Certainty of participation - If you properly attempt to participate in a sale, you should have a reasonable guarantee your transaction will be accepted.
- Cap on total amount to raise - This is a point Vitalik described as "to avoid being perceived as greedy" which I generally agree is important. I believe most of the rationale behind this point is because of the single sale structure. People want to see developers who modestly value their 40 lines of Solidity.
- No "central banking" - There should be guarantees that the issuer of the sale not end up with such a large percentage of the tokens that they would end up with control over the network. There should be policy involved with how the tokens can be sold and spent.
- Efficiency - Vitalik has distilled that investors do not want features of the sale to cause economic inefficiencies or 'deadweight losses'. The ticket scalping of recent sales like BAT are an example of inefficiency that arises when the equilibrium price of the number of tokens at sale is higher than the set cap.
Interestingly, the first two points can not be satisfied at the same time. The last three points are also in a place where without complex "tricks", you can not have strong guarantees about each at the same time. Vitalik goes into detail on the proofs for these two dilemmas but they don't quite fit with the theme of this post. The central theme to takeaway from these criticisms is that there must be compromises and there must be continued innovation into the structure of these crowdfunding events.
Can we even get there?
The contradictions detailed above show the need for innovation into a system with the most reasonable compromises. Vitalik has proposed the following as potential tradeoffs:
There are three more clever things that you can do. First, you can do a reverse dutch auction just like Gnosis, but with one change: instead of holding the unsold tokens, put them toward some kind of public good. Simple examples include: (i) airdrop (ie. redistributing to all ETH holders), (ii) donating to the Ethereum Foundation, (iii) donating to Parity, Brainbot, Smartpool or other companies and individuals independently building infrastructure for the Ethereum space, or (iv) some combination of all three, possibly with the ratios somehow being voted on by the token buyers.
I really like a solid implementation of this model. This way, it is not such a big deal that Gnosis team ended up with most of the tokens - the remaining value is committed to either redistribution among investors or channeled directly into improving the project itself. Now there can certainly be issues with how that "public good" is actually defined. A project would need to be very transparent regarding this process and an evaluation would necessarily require an examination of the individual specifics.
Second, you can keep the unsold tokens, but solve the “central banking” problem by committing to a fully automated plan for how they would be spent. The reasoning here is similar to that for why many economists are interested in rules-based monetary policy: even if a centralized entity has a large amount of control over a powerful resource, much of the political uncertainty that results can be mitigated if the entity credibly commits to following a set of programmatic rules for how they apply it. For example, the unsold tokens can be put into a market maker that is tasked with preserving the tokens’ price stability.
This is essentially another form of the first proposal. I think the community is quite clear that if they are giving a team access to large sums of money, they want a clear and hopefully automated monetary policy.
Third, you can do a capped sale, where you limit the amount that can be bought by each person. Doing this effectively requires a KYC process, but the nice thing is a KYC entity can do this once, whitelisting users’ addresses after they verify that the address represents a unique individual, and this can then be reused for every token sale, alongside other applications that can benefit from per-person sybil resistance like Akasha’s quadratic voting. There is still deadweight loss (ie. inefficiency) here, because this will lead to individuals with no personal interest in tokens participating in sales because they know they will be able to quickly flip them on the market for a profit. However, this is arguably not that bad: it creates a kind of crypto universal basic income, and if behavioral economics assumptions like the endowment effect are even slightly true it will also succeed at the goal of ensuring widely distributed ownership.
KYC poses some serious challenges and even conflicts with some of the values held by the community such as freedom of participation. I think a more mature KYC system that is built on Ethereum in a trust-less way will enable the types of features Vitalik details. There is a lack of maturity in the short term that makes me shy away from this method until more has evolved.
Should they get it all first?
All of the problems outlined above seem to center around the developers getting all the money at once. I don't see many taking the stance that teams who can spend millions making an amazing product shouldn't get millions. Most investors are uncomfortable with giving a team access to all of that funding at the start. I am certainly opposed to untested development teams getting access to hundreds of millions for just a whitepaper and some ICO code. This time mismatch between a finished product and returns for the investors, and the developer's receiving their funding does not properly align those developers with the goals of the investors in a single sale model.
Some interesting ways to go about patching some of the single-round ICO problems is to abandon the property that causes most of the uncertainty. Investors want to know the cap so they avoid getting involved with an overvalued and unproven project. They want to see how a team spends their first million, are they buying Juiceros and swanky advertising or are they hiring developers and conducting independent audits. I am a firm believer that innovation into properly executing an ICO will come in the form of allowing for multiple rounds.
Comments again on EOS
There have been some welcome changes to the structure of the sale. I have quoted the GitHub below:
- Buyer can require that the payment apply to the intended day or transaction will fail returning funds
- Buyer can require that after adding his funds the total funds sent in for the day is below a particular threshold.
With these two changes a buyer submitting at the last second has assurance that miners will not move his order to the next day and he has assurance that if the effective price goes above his target price before his funds are applied then it will not apply.
This does not protect a buyer from having the price go above his target price after his funds are applied.
Unfortunately, buyer pays gas prices even if the transaction fails.
These give buyers much more certainty on point number 1, certainty of valuation. I think this is a welcome addition that gives investors some better protections.
I have seen lots of comments describing the EOS crowdsale as being similar to "mining" and I support that assessment somewhat. The mechanism where the tokens are distributed through this long period of time at a constant rate to those providing the highest sustained proportion of the funding - many parallels to mining. The length and structure of the ICO is very appealing as it is similar toAngelshares (AGS). I would agree with Vitalik that this style of sale needs much longer timeframes, maybe even 5 years. One year might be long enough but this is untested waters.
The early investors, who contributed in an arena with less certainty and more risk, would be exposed to lower prices than those who are trying to buy in during the times of hype. The winning strategy here is contribute a small amount over a long period of time. This is basically a form of multi-round sale, if the team is spending their funding wisely and making progress future 'rounds' might have higher evaluations. The same goes if development slows and uncertainty pools. If the Angleshares sale is any indication, the price is going to pump at the end.
The main problem I have with the sale is the 5 days at the beginning, I am not sure the rationale behind this or what effect it will achieve. I follow 
Bob cannot be sure that he'll pay less than his maximum price.
It's mathematically impossible for him to be sure. The only way he'd know for sure is if he could somehow wait until after everybody else has bid, check to make sure that the implied price is below his max, and then decide whether to contribute. But Bob's not special - and there are probably a lot of Bobs out there with a max price in mind who don't want to be caught with their pants down when all the other Bob's contribute right in the last few minutes of the sale.
The only way Bob can protect himself is to contribute less than he'd like to. So a side-effect of this auction format is that people should self-limit their contributions to mitigate the pricing uncertainty. I don't know how much FOMO will counteract this.
Another of the problems I have with the EOS Token Sale is that the developers don't have a stake in the coin, they are just getting your Ether. Now one would expect they contribute this ETH into the sale in a manner they see fit to optimize their profits should EOS be as successful as they think but you don't know. We need more clarification as to how the ETH can be recycled back into the sale. As far as AGS went this was supposed to be very transparent and allow the developers to acquire proper stake in their project - it was not without significant controversy, for good reason too. This will most likely allow whoever has access to the ETH to amass a large EOS position.
I hope to see how the EOS team responds to some of the criticism regarding this initial 5 day period. It might be able to accurately set a baseline price for the token but some of the issues that come with participation guarantees might arise near the end of that five day period.
There is more work to be done but the published plan is currently clearly labeled as a draft so it will be very interesting to watch this evolve.
Stay skeptical as always,
Kyle