EOS’s unconventional ICO is raising uncomfortable questions
There's no hard evidence of impropriety – just strange design choices.
EOS's initial token sale, which might not technically qualify as an ICO by all definitions but will be referred to as one for the sake of clarity, was unique in a few different ways. Most distinctly, it was enormous, having raised $4 billion in Ethereum by some measures. It also ran for a full year and allowed trading of EOS tokens prior to the completion of the crowdsale.
It also operated more like a thoroughfare for funds rather than a repository for funds raised. In fact, 90.8% of all ETH raised by EOS was withdrawn before the end of the token sale, according to an analysis from ConsenSys. This element in particular is raising uncomfortable questions.
What not to do
The reason ICOs generally don't want to withdraw funds before the end of the crowdsale is because it raises natural concerns around insider sales, the potential for price inflation and the consolidation of a large proportion of a project's tokens in the hands of one party.
The EOS constitution says no single entity should own more than 10% of all EOS tokens, but this is impossible to enforce because it's easy enough for a single entity to hide it simply by splitting their funds across multiple anonymous wallets. Also, the EOS constitution says a lot of things, such as no vote buying, which are most realistically construed as an unenforceable suggestions rather than actual rules. It's also worth noting that 10% of EOS's funds were initially allotted to Block.one, so if at any point it got its hands on more EOS tokens without dispensing of some first, it would have violated its own constitution.
So, an unscrupulous ICO operator would simply withdraw their own funds and then re-use those to buy more of their own token, then withdraw those funds again and repeat as desired. This is sometimes referred to as "contribution recycling" or more straightforward wash trading and would be evidenced by frequent withdrawals during the token sale period.
While there's no evidence one way or the other that EOS creator Block.one did or did not do this, there are several reasons why an unscrupulous ICO founder would do so. First, it's free money – they can just keep buying their own tokens with the same funds over and over again. If that token is unconventionally also available on an open exchange prior to the completion of the ICO the way EOS was, then it might be even more profitable.
So the process might be as follows:
- Buy one's own token with Ether
- Withdraw that Ether from the crowdsale contract and use it to buy even more of one's own tokens, and repeat as desired
- If possible, sell those free tokens for real money
The process might also have the effect of creating artificial scarcity among the general public and creating an artificial appearance of demand, both of which might also help drive up prices. The ability to get near complete control over the governance power of the project is similarly worth a lot, allowing an entity to designate itself as one or more highly-paid EOS block producers or sell its votes to other parties.
The reason some people might accuse Block.one of recycling contributions is because EOS tokens were withdrawn from the crowdsale contract at an alarming rate compared to most other ICOs. More than 90% of all funds raised were withdrawn during the ICO, and a withdrawal occurred on average every 3.8 days.
It's hard to "fairly" compare this to other ICOs because it's unusual for a typical ICO to run for a full year the way EOS did, but the average rate of withdrawals is extremely high compared to the norm by any reasonable measurement.
The withdrawals also started extremely quickly after the token sale commenced. Typically, a project will wait until the initial sale period has concluded before they start withdrawing in case they need to refund investors for any reason or if something goes wrong, but Block.one started pulling Ether out of the contract just five days after the ICO kicked off.
Also, the day Block.one began pulling Ether out of its ICO contract was the exact same day EOS was listed on Bitfinex. If it wanted to use the contribution recycling method to freely purchase its own tokens for resale elsewhere, then that would be a sensible time to begin because it could start selling EOS tokens.
It would also make it much easier to hide one's tracks. Rather than transparently showing obvious token recycling on the blockchain, it would instead look like a large portion of the tokens being withdrawn from the contract ending up on an exchange.
A large portion of the tokens withdrawn from the EOS sale contract went to Bitfinex, ConsenSys noted, although it cautions that the identification of wallet owners is based on EtherScan labels which can only be assumed to be, rather than guaranteed to be, correct. It indicates the flow like so:
The yellow arrow indicates the portion of funds that, through multiple other wallets, ended up going to a single Bitfinex wallet, designated Bitfinex 4 above. By the time the crowd sale concluded, the third-richest EOS wallet was also identified as a Bitfinex wallet, which received over 1 million EOS from Bitfinex 4 during the course of the ICO.
What does it all mean?
There is no concrete evidence that Block.one has been recycling contributions, wash trading EOS, violating its own constitution or secretly giving itself complete centralised control over the EOS ecosystem.
It's possible that it's all just a coincidence, and that the ICO structures which look like they could be used to facilitate wash trading are simply unfortunate design choices. Similarly, one would be leaping to unfounded conclusions by assuming that Block.one is only in it for the money just because it ostensibly raised up to $4 billion but still ended up with an unusually buggy network, a governance structure that failed to account for the possibility of dishonest participants and a platform that was dependent on voting but didn't even include official voting tools or instructions.
Similarly, the reason it's relying on third-party developers to do most of the work, despite raising billions of dollars, might be because it's just really passionate about its enthusiastic community.
Similarly, it might also be a coincidence that the Chinese government's Ministry of Information and Technology has strongly endorsed EOS even as it moves to eliminate decentralised networks. One may or may not construe it as a sign that EOS is centralised under the control of Block.one or Bitfinex, both of which are based in Hong Kong.
Plus, these kinds of concerns have been stewing for a long time now and were being aired even as the EOS ICO was still underway. That's why Block.one promised a public audit to alleviate these concerns.
On 3 August 2017, it said "Block.one is finalising an engagement with a third party auditor to release a public audit designed to provide comfort that block.one has not participated in the EOS tokens sale in anyway, with any funds." That was almost a year and a half ago and there's been nothing yet, but it's still theoretically possible that there will be an audit at some point which will clear things right up.
It's worth emphasising that even if one assumes the worst and that all the allegations are true, there's likely nothing explicitly illegal about any of it.
The main point of ConsenSys's analysis was that there are many ways of preventing the appearance of this kind of impropriety, and that with a few simple ICO decisions and more transparent contracts, there's no reason for an upstanding project to be saddled with these kinds of rumours.
But that's alright. Everyone's first multi-billion dollar ICO is a learning experience.
Disclosure: At the time of writing, the author holds ETH, IOTA, ICX, VET, XLM, BTC and ADA.
- Economist: Stablecoins are doomed to be either too fragile or unscalable
- Coinbase going public: A new way to get broad crypto exposure?
- This week’s big altcoin moves have been cooking since June
- IOTA launches Pollen update: End of an era, start of another
- 5 curious findings from a new government cryptocurrency survey