For most blockchain startups, the initial distribution of their tokens matters tremendously. The reason is that selling tokens is both a way to raise funds and also a way to build an initial community of users, developers, and contributors to the project. Because there are strong network effects (i.e., people will use and contribute to the project only if other people also do so), getting these people on board at the beginning is crucial to a project's long-term success. However, despite several attempts (ICO, IEO, IDO, various types of airdrops, …), I would argue that we have not yet found a suitable token distribution mechanism, at least one that can be applied in general.
This article has two goals. First, to introduce several properties that a token distribution mechanism should satisfy. Second, introduce a new mechanism that meets such properties. This mechanism boils down to a sequence of auctions. Tokens sold in the first auction are immediately liquids; tokens sold in the subsequent auction vest over a given period (or better, in this context, are released slowly over a given period); tokens sold in the next auction have an even longer vesting period; and so on.
What do we want from a token distribution mechanism?
A token distribution mechanism has three main objectives:
(1) Raise funds to be used in the project's development.
(2) Make sure that the project’s core team has the right incentives: it should maintain a sufficient stake in the project and is committed for the long run.
(3) Engage the broader community of potential users/developers. This has to do with how the token is initially sold and what happens afterward.
There is also an important secondary objective is
(4) Get professional investors/speculators to participate
The reason is that professional investors and speculators have the resources/skills to do due diligence on the project and, by their activity, help discover the initial price of the token. I think this is an important element for achieving objective #3: if you have no idea what the correct price of a token is, then it is very hard to incentivize and motivate users and potential contributors properly.
I think most of the issues with existing token distribution mechanisms arise from the fact that, although (4) is necessary for (3) because it provides price discovery, in a very practical sense, there is a tension between (4) and (3) because professional investors and speculators may crowd out (or front run) other buyers. A somewhat naive solution is to airdrop (i.e., give away for free) the token to people who may contribute. But that, in practice, often means rewarding past behavior rather than incentivizing future contribution (airdropped tokens can be sold immediately or as soon as market conditions turn unfavorable).
A better token allocation mechanism
It seems to me that a natural way to meet all the above objectives is to use different vesting periods: users and contributors may be willing to hold tokens for a few years or even longer while doing so would be very costly for speculators. My idea is, therefore, to screen the different types of buyers who may be interested in buying tokens (i.e., investors/speculators vs. contributors) by imposing different vesting periods.
Imagine creating different "vintages" of tokens, each corresponding to different lockup periods. Hence, vintage 0 is immediately liquid, vintage 1 can be traded only after 6 months, vintage 2 can be traded only after 1 year and so on. You then run a series of auctions, one per vintage, starting with the more liquid one. The first one should attract speculators and therefore helps with the price discovery element. All subsequent auctions are less valuable for speculators and more valuable for users. The price should progressively go down (due to the inconvenience of holding the token), allowing users and contributors to purchase the token cheaply. Finally, in general, auctions are an excellent mechanism to raise revenues. I expect this to be the case here as well.
This mechanism is also helpful in achieving objective #2. The tokens allocated to the core team should vest, and the length of the vesting should be significantly longer than the longest-vesting tokens sold at the auction. I don’t know what the right numbers are—perhaps the longest-vesting tokens sold at the auction should vest over 3 or 4 years, and the core team ones should vest over 5 or 6. In any case, the point is that this mechanism also helps to put some structure on how the core team is incentivized.
One can, of course, embellish the mechanism in various ways. For example, the token could start trading in a DEX after the first auction (the liquid one). Any new info related to the project (or even aggregate movements in the crypto market) should affect the price of the token on the DEX. Users who purchase tokens with a longer vesting period get a higher discount relative to the DEX price. This does not necessarily mean that the sequence of auction prices is decreasing, but I believe it nonetheless captures the idea that those who hold the token for longer get a larger reward.
Concluding thoughts
When considering token allocation mechanisms, most people worry about volatility in the price of tokens after the initial allocation. The reason is that a random drop in the token price may cause token holders to panic and sell their tokens, depressing the price and inducing further sales. The resulting spiral can kill a project precisely because all those who are supposed to contribute to the project (per objective #3 above), end up not holding tokens anymore and therefore not having any stake in the project.
The mechanism described earlier does not directly address this issue but makes it irrelevant: project contributors’ are supposed to purchase tokens that vest over a period of time. Short-run variations in the price of tokens should not affect their commitment to the project. If anything, they may want to contribute more when the price is low as a way to increase the token’s future price (when they expect to sell).
Any comment? Please leave them below.