On Fair Token Launches
What is a Fair Launch?
Fair launches have always been lauded as the holy grail for a cryptocurrency. The problem is, how can a launch be fair when we are currently in a time when assets have absurd valuations before launch, are typically sold, and even the ones that aren’t sold are already treated with speculation?
Parties will continually argue on the philosophy of fairness, but typically come to the same conclusion of fairness encapsulating every possible individual associated with a particular system. On a macro scale, it’s every member of society, on a micro scale, it’s everyone involved with cryptocurrency to some extent. Before Bitcoin, there was no subset of individuals involved with any form of cryptocurrency (as we know them today). After the fact, that subset has been established.
An article earlier this year from Arjun Balaji and Hasu defined a fair launch as one that has a long length of issuance for the sake of price discovery, and price equality — where there’s no discount. To them, EOS was one of the fairest forms of ICO because its auctioning mechanism enabled efficient price discovery. They provided a lighter take on the subject rather than absolutism and provided an issuance axis against a price equity axis to prove their point.
Fair in my case means how many individuals have access to the token or asset upon launch, while meaningful requires an analysis of distribution. For a traditional MOE (medium of exchange), this can be the amount of addresses holding the asset, while in validation-based networks, it could be a combination of distribution plus active participation. For studies in distribution, many in the space have referenced the Gini coefficient as a means of measuring a system’s “decentralization,” and in essence — its distribution.
In my opinion, it’s next to impossible to recreate the launch of Bitcoin, but there are meaningful steps that can be administered in order to get as close to a fair launch as possible. When thinking about cryptocurrency, one of the main points of contention that comes up beside valuation models and markets is distribution, which is inextricably linked to the launch of the asset itself.
The first iteration of a fair launch came from Bitcoin’s launch in 2009. There were no coins that were premined, a two-month notice before the network launched, but more importantly, no exuberant exterior value to attach to the asset. When reflecting on the launch of Bitcoin, we typically use the “$20,000 lens” which is a detriment to any study reflecting on it.
How could Satoshi continually hoard so many coins?
Bitcoin’s early adopters made off like thieves!
The problem with that line of thinking comes from the fact that Bitcoin didn’t have “value” traditionally until 2010, when the famous pizza transaction occurred. Until then, it was simply a hobby project by a group of cypherpunks decided to send what was then digital dust and would later become digital gold back and forth. No user of Bitcoin knew what would happen to the asset’s valuation (or eventual classification as an “asset”), and how quickly its price would reach the levels it has today.
Dan Held in his article explaining this topic put it quite soundly:
Satoshi was a person like any other, not some infallible being. This was the fairest distribution he could have come up with given what he was building/timing/audience. It’s intellectually dishonest to compare Satoshi’s early mining of Bitcoin at a loss, with premining of an ICO with a positive market value (or expected positive market value).
Unlike every other founder in history, Satoshi never cashed out.
In the case of Bitcoin being a medium of exchange and store of value, the wider the pool of holders, the more meaningful the distribution. As the amount of Bitcoin one owns isn’t tied to the ability for that individual or group to add meaningful security to the network (a node without the asset), the wider the pool of holders, the better. Over Bitcoin’s various price cycles, older coins have been distributed to newer individuals through selling — effectively stretching the number of holders the asset is distributed to, and continuously spreading the asset (as seen in an analysis conducted by Dhruv Bansal).
There have been attempts to re-create the fair launch of Bitcoin as closely as possible, namely with projects like Ravencoin and Grin — both of which launched as basic PoW networks without a base unit price. However, biases can be included in each case which gave them an implicit valuation, unlike Bitcoin’s immaculate conception. Ravencoin which launched on January 3rd, 2018, was primarily funded through Overstock’s Medici Ventures. To this day, the core developers are still associated with Medici, and Overstock’s placement of ‘Millions of Dollars’ into Ravencoin allows the association of meaningful value with the launch. A launch like Bitcoin would have the asset at no implicit valuation which is quite difficult unless the network is bootstrapped after a genesis block with voluntary development.
Grin, on the other hand, was a bit different because the implicit valuation wasn’t directly associated with core development, but rather through outside parties with ulterior and promotional motives. The core development team remained independent, but the launch was extremely high profile with rumors of over $100 million in SPVs to mine it. Along with those SPVs could come promotional efforts and initiatives from those invested in order to both generate hype cycles and provide implicit “valuations” at launch from the hashpower associated with the network alone. However, distribution becomes meaningful over time as the original miners quickly become sellers to cover overhead costs.
“Fair” launch attempts after Bitcoin are typically plagued with implicit valuations and information asymmetry, even if both aren’t meant by those launching the asset. Although both come with criticism, they at least tried to provide as fair of a launch possible in 2018 and 2019 — a time when speculation has run rampant with overvalued token launches.
Bitcoin’s launch was a natural gathering in a public square, Grin and Ravencoin’s launch were widely promoted public meetups, ICO launches are mostly high priced conferences and clubs. Everyone is on equal footing in a natural gathering; meetup promoters are typically directing the show; conferences usually have a handful of high profile targets and plenty of noise.
Toward Tokens — A Note on Airdrops
It’s important to briefly touch upon distribution mechanisms such as airdops. Airdrops typically involve distributing tokens to a wide range of individuals for the sake of spreading awareness about the asset. During the 2017 ICO-craze, airdrops became increasingly common as a marketing tactic to get more individuals interested in particular projects. They were also used to grant individuals outside of private sales the ability to gain some tokens. An example of this is Polymath’s general 250 token airdrop and OmiseGo’s variable airdrop to Ethereum holders. Some projects even made followers work for their airdrop, with Quantstamp’s “Proof of Caring,” and Mainframe’s “Proof of Heart.”
Projects that included token sales weren’t the first to deploy the use of airdrops to distribute tokens — projects such as Stellar, and Decred began with an airdrop, all without selling tokens.
Note — Stellar did give Stripe 2 billion lumens in exchange for $3M, with Stripe to return net profits to their foundation.
However, airdrops with no meaning attached to them act more like a revolving door rather than an effective user acquisition method. As humans, we seek to do the minimal amount of work to extract a maximum amount of value, and most airdrops were simply sign-ups with no meaningful contribution or work required. A user could simply do the minimal amount of effort (sign-up), receive their tokens, and sell them off on an open market, never to return to the project.
What’s the incentive to stay?
Not much really — one could learn a minimum about the project, and simply find the necessary steps to acquire said airdrop before running off to return to meaningful tasks. Polymath is a great example of this — a signup granted users nearly $250 in value, and the token is a basic payment token that has questionable to no usage. Do most users remember the Polymath brand? Maybe, but they remember the $250 gain rather than the intricacies and value of the project’s offering.
Attempts with Tokens
Selling premined tokens by its very nature deviates leagues further from a fairer launch, due to the hand-selection of percentages and recipient bodies. The SPV example with Grin was a third-party collective and coordinated effort in order to use the network as intended at launch. EOS was a great attempt at a longer issuance cycle and price discovery, but was still able to decide participation thresholds and even had the power to recycle their own capital into the sale. However, token projects can make meaningful attempts to get tokens in the hands of those that can both contribute to the project, or help its growth.
The following projects are examples that have attempted to put tokens in the hands of meaningful actors and tried to avoid a pure speculation play riding on a sale as a primary means of distributing the asset.
Numerai officially launched Numeraire in June 2017 on the Ethereum mainnet. Included in that launch was the distribution of 1,200,000 tokens to 19,000 data scientists — potential users of the token, as the application involves the construction of ML models in order to make predictions on how financial markets will act. Numerai existed prior to the launch of “Numeraire,” and already had a body of data scientists that were interested in the platform. Numerai’s choice to not have a crowdsale and instead distribute tokens to potential users was an important mark in a token’s ability to have meaning in its distribution.
As mentioned prior, no token can launch “fairly,” as there is typically a central issuer and a number attached to the asset. Tokens are minted by a central party, commonly given a valuation, and then typically distributed among stakeholders in a network including investors, a team, and commonly — the “community.” I use quotes around “community” because a community is typically one that is interested in a collective commons rather than pure speculators seeking a few free tokens to unload on an open market. However, just because a token cannot mimic the immaculate conception of Bitcoin and come close to a Proof of Work genesis block launch, the issuing party can at least distribute the asset to meaningful parties and give a chance to all relevant users on the issuing party’s platform or in their ecosystem.
When Numerai was launched, data scientists were able to win Bitcoin with nothing at stake by providing the best model to the system. However, with the introduction of the token, data scientists have a bit more skin in the game and something to lose if their predictions are poor, as NMR is destroyed for bad models. NMR acts as a confidence mechanism with a “value” attached to it for the sake of proper participation.
The choice to distribute it to scientists already using the platform was a good move in defeating initial speculation, creating an information cycle on how to properly use the token, and leading to their status as a commonly pointed-to mode of distribution.
Numerai has also continued its distribution post-initial allocation to similarly relevant parties. In March of 2018, Numerai announced that it was giving away $1,000,000 worth of its token to Kaggle users in order to grow its own user-base. Kaggle at the time of distribution was just acquired by Google and had over a million data scientists using the platform. The choice to grant those scientists the asset was in another effort to meaningfully distribute it.
This initial distribution has led to NMR being one of the most widely used tokens, with an ever-increasing amount of NMR staked over time. Over time, there is a plan to cut the supply in half, and distribute the rest of the assets to necessary stakeholders as Numerai moves toward the creation of its data marketplace called Erasure. Considering that they’ve built their community over time and there’s meaningful activity with the token, bootstrapping the marketplace might come a bit easier as usage exists in harmony with speculation.
Fair launch? A decent shot at it. Meaningful distribution? The targeting deployed by the project’s team is quite meaningful.
Note: during the creation of this article, Numerai sold $11M worth of tokens to large investors from their treasury.
Handshake’s goal is to replace traditional certificate authorities (entities that can issue trusted certificates in order to represent legitimate online identities) with a decentralized system. As Handshake is still in development (as opposed to Numerai’s existing platform), it took an alternative approach to how it is launching its asset.
Upon the launch of its genesis block, Handshake is to have 1,360,000,000 tokens in circulation. However, a massive 77.5% of that block of assets will be granted to both contributors and open source developers. 7.5% is being allocated to direct contributors to the project, similar to the way in which Numerai distributed tokens to data scientists using their platform. However, the majority of tokens (70%) will be granted to the wider open source community. This distribution will come in the form of granting GitHub FOSS contributors, freenode accounts that are older than six months, and PGP Web of Trust members (~60,000), along with those that contact the project directly to explain what they’ve contributed over time. A nomination process is accessed after that for those that are already qualified.
Handshake did in fact receive prior investment for the 7.5% for $10.2 million, leaving the protocol’s valuation at $136 million. Unfortunately, the ‘immaculate conception’ of Bitcoin still can’t be achieved in their case but their effort to distribute 77.5% of the tokens to contributors and open source developers signals an effort for inclusion. However, all of this becomes increasingly confusing as Handshake’s primary value proposition isn’t a form of locking the asset for validation but is a means to purchase the rights to Handshake names.
A fair distribution? Fairer than most launches. Meaningful? Getting there — (7.5%). But I’ll continue to ride on the assumption that most recipients will elect to get rid of the token as soon as they’re distributed.
Livepeer is a protocol for p2p video transcoding services to “enable developers to add live or on-demand video to their product.” Nodes on the network are responsible for locking up LPT in order to provide services for the network, and are then paid for the work that they accomplish based on the demand for transcoding services.
Livepeer’s “Merkle Mine” technique released 63% of the supply (10,000,000 LPT) to Ethereum wallet holders that had at least 0.1 ETH at the time of their snapshot. Holders would then have to generate a proof to claim the LPT associated with their wallet, with a time limit of 86 days. Tokens that weren’t claimed by users ended up being pooled for active “miners” participating at the time. What this type of work enabled was the identification of potential network workers (transcoders that would have the technical acumen to participate in the network), and allow them to claim tokens to accomplish that task.
An article by a close friend of mine, Viktor Bunin, noted that the distribution of the token is in fact quite centralized, and questions the notion of any of the 800 merkle miners actually becoming meaningful transcoders when they might simply fall into the realm of speculation. One of the top merkle miners ended up receiving 19.06% of all LPT mined, and could may have well been an insider hitting the system hard as a result of possible information asymmetries.
The rest of the LPT supply was distributed among the team (12.35%) and investors (19%), and will eventually be used to transcode or delegate to transcoders (15 active slots). LPT also features a target participation rate, which sets a participation threshold as a necessary percent of tokens either being used or delegated, or risk the continual inflation of the LPT token supply.
Fair? A good attempt at opening up distribution to active Ethereum network participants. Meaningful? Questionable. Innovative? Definitely.
How do we create meaningful token distributions?
I think there are lessons to be learned from each of these projects in how to distribute assets, especially in the case of tokens. Although many of these projects haven’t succeeded yet in their original mission, or are working toward it, we can still learn from the approaches they took in asset distribution.
Build a meaningful user base, and reward existing users.
Numerai built their initial product, garnered a user base, and then distributed tokens to that base. That made their token usage that much stronger, considering that incentives were already aligned and there were clear goals outlined.
Go after parallel services and try to capture their users.
Numerai’s gift to Kaggle users was a way in which outsiders to the platform with an idea of how a platform like Numerai works can be drawn in. A financial incentive along with familiarity is a strong combination.
If you’re airdropping, make it count.
Handshake’s airdrop to the open source development community wasn’t random — it was a way in which they could potentially draw interest from developers to work on the project.
Make them work for it.
Livepeer’s model involved their Merkle Mine which required work and interaction with the network to claim. Work similar to a Merkle Mine might also be a way to distinguish pure speculators from those who will know how to interact with a protocol when it’s live.
Sell less, distribute more.
Selling a majority of tokens is a signal toward encouraging speculation rather than meaningful usage. It could also cut plenty of players out that wish to get involved with the ecosystem such as those without the means to invest. Speculation is good to an extent, but a way of achieving meaningful usage coupled with effective value capture is how a token should be crafted.