Implement Akita sale/stream as a Balancer pool

This proposal came too late to be one of the options in the Akita, a path forward poll, which at the moment is leading with "Sell 10%, stream 90%" and "Liquidate 100%" so I will propose it as a further discussion on the merits of how to implement it. This was posted on June 1st as a twitter thread and gathered some positive feedback from many of the community representatives.

The issue with the two main proposals

• There's not enough market depth to liquidate all Akita tokens without losing a lot of value for Gitcoin. Even if there are serious doubts about the long term viability of the Akita community, we have a responsibility to maximize the value gained for Gitcoin.
• Locking Akita on Sablier doesn't put that capital to good use, and if the market crashes and recovers and crashes again during before the stream is over, then there was a missed opportunity.

The Proposed Implementation

1. Sell 10% of Akita in any exchange, put it in the Gitcoin treasury
2. Sell yet another 10% into ETH.
3. Create a Scheduled Balancer pool with the 10% ETH obtained in step 2 and all the remaining Akita.
4. Set up the pool controller to programmatically change the ratio over a period P so that it inverts from a 10% ETH 90% AKITA into a 90% ETH and 10% AKITA
5. After period is over, the community can decide what to do with the remaining tokens.

Those types of pools are usually called Liquidity Bootstrapping Pools and are often used to launch new tokens that don't have liquidity. In this case, the token has launched but it would still bootstrap the liquidity. You can read more about LBPs on this article about Radicle.

To make sure the market doesn't dump immediately, we could do step 1 and 2 using Gnosis Auctions: this would take off some buy pressure immediately and then make sure the starting price of the LBP is closer to what other market makers have.

The end result is that:

• At the end of the period, Gitcoin will have sold most of their Akita, at a fairer market price, similarly to a Dollar Cost Averaging strategy
• Akita community gets a much more liquid market, a "billion" dollars deep (emphasis on the air quotes). This will reduce the impact any single seller can have on the final price.
• All trades that are executed against this pool will earn fees for Gitcoin, so in periods of high volatility Gitcoin is leveraging its massive Akita Holdings to generate Yield
• There will be a constant sell pressure on Akita, but it's one that is scheduled and known

Variants and Risks

Balancer Pools are quite flexible and we could do other options like having a pool that starts at 91% Akita, 3% GTC, 3% ETH and 3% DAI. Or we could add all the dog tokens into a big billion dollar pile of dog poo money.

Pools also don't necessarily need to be locked: we can add a smart controller so that Gitcoin DAO can accelerate the schedule at a later vote, or simply change it completely and use the Pie DAO so that ratios can be changed at the will of GTC holders. Also, at any point, Gitcoin could vote to take the liquidity out of its own pool and receive it in either token of the pool.

It's also important to note that LBPs are different than just selling a token: the pool wants to keep a constant ratio at all times to reach its target, be it 10%, 25% or 90%. So it means that the pool will both sell and buy akita for ETH to keep the target ratio–that's how AMMs work. A LBP changes that target ratio slowly over time, but it also is reacting to the market in real time. It means that, if akita drops to near 0, then the pool will automatically be buying akita with ETH until it reaches the desired balance, so almost ETH would be drained from the pool. But that ETH came from the Akita sale in the step 2, and the end result would be the same if the strategy was to stream or liquidate it anyway. On the other hand, if the market goes up again, that will mean the pool will start selling the cheap akita it bought back to ETH (plus the fees it got during the period) which is why pools make money from volatility.

Disclaimer: I work for Balancer labs, but if the community wants to they can use other solutions. Balancer also has a "LBP referral program" but I am perfectly willing to either not claim it or to donate the BAL gained to Gitcoin.

Period P

This proposal leaves the P period open. The ending date of this poll is after the resolution of the "Akita - a path forward" poll, so votes should take the result of the other snapshot in consideration.

1. Stream: a LBP of duration of 2 years
2. Intermediary: make the LBP last 1 year
3. Liquidation: a short LBP over 1 week
4. Do an LBP but in another configuration or platform
5. No LBP: just lock it Sablier or dump it all in uniswap, depending on the results of the Akita poll