NEO — A Vulnerable Protagonist
NEO - Not yet as anti-fragile as Agent Smith
TL;DR. Ecology:6. Technology:5. Decentralization:2. Valuation:3. Rating:4/10
(Update 2018-12-25: Please see the semi-annual ratings review of NEO here.)
NEO is the second largest cryptocurrency rated by BD Ratings so far, and it is a very interesting one. NEO was initially named AntShares and made its debut on bitcointalk.org on 7th of July 2016 by user dahongfei, a well-known Chinese cryptocurrency pioneer. The main focus of AntShares was to become a platform for digital assets, where the proof-of-ownership issuance stayed centralized as usual but where then the mobility of the tokenized proof-of-ownership could exist in a decentralized manner not really possible before the days of decentralized blockchains. One simplified comparison to these digital assets would be regular Ethereum ERC20 tokens.
Unbeknownst to most users on bitcointalk.org back then, AntShares held a ten day long ICO in October 2015 where 2 100 BTC were raised from mainly Chinese investors and 17.5M tokens out of a total supply of 100M were reserved. The second more globally focused ICO initiated in the summer of 2016, to start on 8th of August 2016, was to reserve 22.5M tokens. This meant that a total of 40M tokens were reserved for the public. The ICO had an escrow setup to reduce risk for investors, and early ICO investors also enjoyed a 1.2 multiple token bonus, to disincentivize investors from waiting until the very last moments of the sale. Another 10M tokens were reserved for what was called early contributors while the last 50M tokens of the total supply were reserved for the development team to use on bounties and strategic investments. The 50M tokens were apparently vested for a year, after which they were to release in batches of 15M per year. With this first AntShares bitcointalk.org thread, information was also given about a growing team as well as about a new AntShares venture called Onchain that was to serve as an AntShares marketing and consultant company.
The tokens received from the ICOs were called AntShares (ANS). With them, users could accumulate a second token called AntCoins (ANC), which were to be used as utility tokens for registering assets on the chain, among other things. So as the name implies, AntShares - not divisible in units smaller than one - were equity-like tokens that gave the owner a possibility to vote on which nodes were to be delegated the trust and responsibility to create blocks and thus confirm network transactions. ANS holders would then get ANC tokens pro-rata through the block creation process. A 100M ANC were set to be produced in this way over the course of 22 years, and any ANC burned during asset registrations or other non-free on-chain activities were to be re-distributed to ANS holders. So, to summarize, just holding ANS meant you accumulated ANC proportional to your position size.
AntShares' code was open source and published on GitHub in 2015. It was built from scratch rather than taken from an already existing code base. The consensus model chosen was that of delegated Byzantine Fault Tolerance (dBFT), meaning in this case that token holders delegate the right to create new blocks to 7 trusted nodes (initially as well as right now controlled by the Foundation), ensuring that much fewer full nodes than usual actually maintain the blockchain.
AntShares ran an extensive bounty campaign that is, or at least was, the norm for ICOs. BD Ratings is as usual critical of the fact that people earn tokens by changing forum signatures, writing blog posts, retweeting tweets and so on. In an environment where already huge sums of money are allocated to cryptocurrency development teams, more marketing is the least necessary addition to a project and serves mainly to inflate the price of tokens and shift wealth from people susceptible to marketing to team members offering yet to be built products. Bounty issues also always seem to clog the discussion platforms utilized by the project. The first hundred pages of the bitcointalk.org announcement thread is full of such issues.
Releasing only 50% of the tokens to the public is mainly an issue of centralization and will be discussed in that section. But if a large amount of tokens stay concentrated in or around the NEO Foundation, that is a threat to the integrity of the dBFT consensus protocol as it makes it easier for the team to obtain "voting majority". The 10% of all tokens allocated to early contributors may also to some degree be in the hands of team members. This might affect the ecology forming around the chain since it signals to potential developers or companies that there will always be some kind of threat should either the Foundation dislike a specific NEO project, or should regulatory pressure be applied to the Foundation.
To slowly build an ecology around NEO, businesses and private individuals must find the chain useful. Examples that lowers the threshold of getting economic activity on the chain may be the NEO collaboration with Chinese certificate authorities, to help attract Chinese companies. The zero transaction fees may also act as a magnet to such activity. The cost of publishing smart contracts on NEO seems however to have been set extremely high (500 GAS, around 5 000 USD at time of writing), which restricts on-chain activity considerably. Some actions on the blockchain are even more expensive. The rationale given is that it mitigates scam ICOs or trash data/spam, which to an extent is true, but it also hinders fast, broad adoption.
The City of Zion - Alive and thriving like the NEO community
Despite smart contract fees, it seems there has been a decent number of ICO's conducted on the NEO blockchain, and the NEO Reddit community with close to 100 000 subscribers is active as well as open to discussing sensitive issues. Red Pulse took in more that 400 000 NEO in its sale in October 2017, which also successfully stress tested the network. Neon Exchange (NEX), an ambitious decentralized exchange project, is soon running an ICO on NEO as well. Many more have been conducted and was planned mainly before the cryptocurrency hype died in January 2018, but there is still a lot of ICO-related discussion on the NEO subreddit. Most, if not all of these ICOs, are issued in the NEP-5 standard, the NEO equivalent of Ethereum's ERC20 tokens.
The network seems to handle around 30 000 transactions per day on average right now (earlier this year it was closer to 100 000), which is acceptable as the hype surrounding cryptocurrencies has died off. What has to be considered though is that many of these transactions are free as Consensus Nodes have not implemented any transaction fees yet. A large number of developers are contributing to the ecosystem, according to Da Hongfei, which obviously is very positive for this Ecology grade. Among these are semi-independent community developer groups as well. NEO has a good ecology growing around its chain and there are multiple indications it may take economic activity from Ethereum should that platform be congested again in the future.
Reasons: Large community. Independent community developer teams. Reasonable on-chain activity. Low transaction fees but high smart contract fees.
The NEO GitHub is active but the number of contributors seems quite low. The metric may be misleading however, and it is a fact that developers are spread over at least three different groups: NEO Global Development (core developers), City of Zion and NewEconoLab. A 3 month long audit was conducted on the code base by security research company Red4Sec.
Through the on-chain elections of Consensus Nodes, NEO will soon have in place a governance system that very well may be flexible enough to give NEO holders the chain parameter changes that they wish for. On-chain governance is an area explored too little in the cryptocurrency ecosystem, and despite the system being criticized by prominent researchers like Vlad Zamfir, BD Ratings see the advantages of resolving conflicts by voting to lessen risk of hard forks. We have to wait to get more data on the practical implications of on-chain governance.
Due to the very small number of Consensus Nodes, NEO can achieve up to 1000 transactions per second (tps)according to Da Hongfei. This reddit thread paints another picture. The 1000 tps claim seems to be more theoretical and would need changing of some chain parameters by the NEO Foundation. Also, if the maximum throughput would be tuned up to those numbers, the blockchain size would increase considerably each year. The relatively high NEO throughput capacity right now of around 30 tps is not achieved due to superior technology but rather by moving along the centralization/throughput spectrum. Since only a handful of full nodes need to store the blockchain and synchronize transactions, throughput is higher than Bitcoin or Ethereum. Trinity, a second layer scaling solution similar to Ethereum's Raiden, is being worked on as well, which might mitigate the need of increasing on-chain throughput somewhat.
One thing that distinguishes NEO from other large chains is that transactions are free for users. It is a philosophical decision, where chain maintenance costs are pushed mainly to corporations in the future as the Consensus Nodes decentralize. But even if chain maintainers are somewhat willing to carry these costs, having a zero transaction fee creates all sorts of problems. It is a massive attack surface where malign users can attempt to spam the network with useless transaction in order to congest it and at the same time force full nodes to store unnecessary data. A zero fee policy is also troublesome from a Consensus Node's economic point of view. A blockchain ought to have economic incentives in place for the nodes that are actually carrying the network forward. Instead, NEO seems to hope that the Consensus Nodes are run by philanthropic entities, or at that they are content with indirect benefits like future NEO appreciation. This sadly is building in a need for trust, in what ought to be constructed as trust-less as possible.
NEO was pushed in to the spot light earlier in 2018 when known cryptocurrency advocate Eric Wall went on a tirade against NEO in a Twitter thread. The thread was initiated after a long block delay occurred on the NEO blockchain. A bug caused one of the seven Consensus Nodes to fall out of sync, resulting in the other six nodes settling in a state of limbo so that the finality-focused chain came to a halt. Misunderstanding the situation somewhat, Eric Wall seems to have drawn very harsh conclusions that the NEO developers lacked even fundamental knowledge in applied Byzantine Fault Tolerance and that they hadn't prepared for Consensus Nodes going offline. The bug however seems to not have been related to the consensus algorithm, but other logic. NEO Consensus Nodes are able to carry the chain forward as long as at least 66% of them agree on the transactions. Eric Wall later apologized but retained some of the criticism of NEO. The reason BD Ratings bring this whole thing up is that many still seem to think NEO had a fundamental weakness in its dBFT algorithm. The bug however was serious and it is quite concerning that one unstable node could cause such a mess on a hugely valuable network.
Reasons: Open source. Diverse developer teams. Novel dBFT algorithm. Unnecessary need to trust Consensus Nodes built in to the system.
With the delegated Byzantine Fault Tolerance consensus rules, NEO has chosen to have 7 Consensus Nodes. The NEO Foundation controls all 7 of those, stating that they for now value efficiency more than decentralization. The founder Da Hongfei even goes so far as to say that decentralization is cryptocurrency political correctness. Under 2018, the number of Consensus Nodes are to stay between 7 and 13, effectively ensuring at least a majority control of the chain by the NEO Foundation unless they give up some nodes. Even if the total number in the future is raised closer to 1024, which is maximum, the vast number of NEO owned by the NEO Foundation ensures that they arguably have power over who the new Consensus Node owners will be. This is a Ripple-like situation not fondly seen upon by many cryptocurrency enthusiasts. Centralization mitigates the whole point of a blockchain. NEO developer Malcolm Lerider does not agree, and astonishingly calls the 7 Foundation controlled nodes a balance between decentralization and performance (full control is not in any measure something balanced). It is also interesting to see him definitively state that NEO is not a currency. If the network is as decentralized as he proclaim it to be, then he has no authority on defining what NEO is or isn't. No, NEO definitely ought to let go of the control and go for painstaking consensus building instead; the risk of not convincing unknown Consensus Node operators to follow a general strategy/roadmap ought to be quite low as it will be proposed by NEO experts.
The situation gets worse as it seems Consensus Nodes are encouraged to ID themselves as well so that they can be held legally responsible for attacks. Presumably, a Consensus Node election runner-up will not get elected by NEO holders unless holding an ID certificate. This is constructed to elect known, accountable entities but also makes it easier to shut them down should such orders come.
Cypher - Betraying everyone on Nebuchadnezzar despite being trusted
In the Ecology section, the high contract deployment costs were discussed, but the issue touches on the subject of centralization as well. The expensive smart contracts on NEO are often in contrast compared to the cheap, permissionless nature of their Ethereum cousins. Critique of this setup is often met with the argument that a solid idea can always ask for funding from the NEO Foundation or other community groups. This creates, in effect, a permission culture - something that blockchain projects ought to be allergic against.
Given that the NEO Foundation and other independent NEO developers have understood the importance of decentralization for a chain's long term survival, BD Ratings find it probable that they will take measures to relinquish control of Consensus Nodes. When there are more evidence of this actually occurring (not just on the testnet as is the case right now), the Decentralization rating will be higher. A positive step already taken is the support of community developer groups like the diverse City of Zion, and the NewEconoLab in China. Also, Consensus Node election infrastructure is slowly coming together, laying the foundation for future elections.
Reasons: NEO concentrated in the hands of the Foundation. NEO Foundation controls all Consensus Nodes. Multiple developer teams. Election infrastructure is being built.
Valuing this project first need some short reasoning on the relationship between NEO and GAS. BD Ratings conclude that GAS has inherent value on the chain for its usefulness as payment to Consensus Nodes to actually confirm the smart contracts, asset registrations and so on. NEO then has inherent value as "GAS producing" tokens as well as having in-built voting rights. How the value of GAS and NEO ought to stand in relation to each token is another discussion, but for this ratings analysis, BD Ratings will sum the total market valuations to get to a relevant number. All NEO tokens are already distributed (although some are still vested a couple of more years). Most of the GAS tokens are still dormant, to be released/produced continuously in the block creation process. BD Ratings include all 100 000 000 of them mainly to be conservative. Only including the GAS in circulation would be more misleading as there currently is a rather high GAS inflation.
At time of writing, NEO is valued at USD 3.14B. GAS is valued at USD 1.05B. Total valuation then is USD 4.19B. That is clearly putting it in the top 10 cryptocurrencies based on market valuation. It is too high for something that is still so centralized. As soon as the NEO Foundation proves that it can let go of the control, NEO will obtain better rating grades for both Decentralization and Valuation.
Reasons: Chain still centralized.