Should GnosisDAO build an IDO specific Dapp?

Knowledge Base Archive
1

Project idea: EasyAuction

EasyAuction should become a platform to execute batch auctions for fair initial offerings and token buyback programs. Batch auctions are a market mechanism for matching limit orders of buyers and sellers of a particular good with one fair clearing price.
Already in traditional finance, batch auctions have established themselves as a tool for initial offerings, and in the blockchain ecosystem, they are expected to become a fundamental DeFi building lego brick for price-discovery as well.
The EasyAuction platform is inspired by the auction mechanism of the Gnosis Protocol v1, which has shown a significant product-market fit for initial dex offerings (IDOs) (cf. sales of DIA, mStable, DMMetc…). EasyAuction improves significantly the user experience for IDOs, by settling up to 2500 - or more - bids with a single clearing price instead of roughly 28 orders and thereby making the mechanism fairer, easier to use, and more predictable.
Given the emerging regulations for IDOs and utility sales - see MiCA -, EasyAuction is intending to comply with them and enabling new projects a safe start without legal risks.

Use cases

Initial token offering with a fair price-discovery

This batch auction mechanism allows communities, projects, and DAOs to offer their future stakeholders fair participation within their ecosystem by acquiring utility and governance tokens. One of the promises of DeFi is the democratization and communitization of big corporate infrastructures to build open platforms for the public good. By running a batch auction, these projects facilitate one of the fairest distribution of stakes. Projects should demonstrate their interest in fair participation by setting a precedence for future processes by choosing this fair auction principle over private multi-stage sales.
Auction based initial offerings are also picking up in the traditional finance world. Google was one of the first big tech companies using a similar auction mechanism with a unique clearing price to IPO. Nowadays, companies like Slack and Spotify are advocating similar approaches as well via Direct Listings: Selling their stocks in the pre-open auction on the first day of trading.
Overall this market for initial token offerings is expected to grow significantly over time together with the ecosystem. Even on gnosis protocol version 1, a platform not intended for this use case, was able to facilitate IDOs with a total of more than 20 million funding.

Token buyback programs

Many decentralized projects have to buy back their tokens or auction off their tokens to clear deficits within their protocol. This EasyAuction platform allows them to schedule individual auctions to facilitate these kinds of operations.

Initial credit offering

First implementations of yTokens are live on the ethereum blockchain. The initial price finding and matching of creditor and debtor for such credit instruments are, most efficiently, executed via this batch auction mechanism.

Protocol description

The batch auction mechanism

In this auction type, a pre-defined amount of tokens is auctioned off. Anyone can bid to buy these tokens by placing a buy-order with a specified limit price during the whole bidding time. At the end of the auction, the final price is calculated by the following method: The buy volumes from the highest bids are getting added up until this sum reaches the initial sell volume. The limit price of the bid that increases the overall buy volume to the initial sell volume is setting the uniform clearing price. All bids with a higher price will be settled and traded against the initial sell volume with the clearing price. All bids with a lower price will not be considered for the settlement. The principle is described best with the following diagram:

image from nytimes
image from nytimes924×806 199 KB

(picture source: New York Times, Google IPO explainer)

Specific implementation batch auction

EasyAuction allows anyone to start permissionlessly a new auction of any ERC20 token (sellToken) against another ERC20 token (buyToken). The entity initiating the auction, the auctioneer, needs to make sure that they conform to their laws. If they are in doubt, they will get advice. To start an auction, the auctioneer has to define the amount of token to be sold, the minimal price for the auction, and the end-time of the auction. Once the auctioneer has initiated the auction, anyone can participate as a buyer and submit bids. Each buyer places buy-orders with a specified limit price into the system. The buy amount of an order must be at least 1/2500th of the auctioned sell amount.
Once the auction ends, the final clearing price will be determined via an onchain calculation. The minimum buy amount per order is set in such a way that no more than 2500 orders need to be read onchain for the price calculation. This limit ensures that the onchain price can be verified for correctness within one ethereum transaction.

Once the price of an auction has been verified onchain, everyone can claim their proceedings of the auction. The auctioneer can withdraw the bought funds, the buyers being matched in the auction can withdraw their bought tokens. The buyers that were not matched in the auction can withdraw their initial deposited tokens as well.

Comparison to Dutch auctions

The proposed batch auction system has a number of advantages over Dutch auction.

  • The biggest advantage is certainly that buyers don’t have to wait for a certain time to submit orders, but that they can submit orders at any time. This makes the system much more convenient for users.
  • Dutch auctions have a very high activity right before the auction is closing. If pieces of the infrastructure are not working reliably during this time period, then prices can fall further than expected, causing a loss for the auctioneer.
    Also, high gas prices during this short time period can be a hindering factor for buyers to quickly join the auction.
  • Dutch auctions calculate their price based on the blocktime. This pricing is hard to predict for all participants, as mining is a stochastic process.
  • Dutch auctions are causing a gas price bidding war to close the auction. In contrast in batch-auction, different buyers will compete against another by the gas price instead of the limit price.

Embracing regulation

More and more regulations for utility, governance, e-money, and security tokens are on the horizon. Many jurisdictions are investigating a proper framework for these crypto natives. While this might introduce some friction, overall is great news, as it will give many start-ups certainty about their business and their tokens.
EasyAuction will conform to regulations and is intending to ease the legal procedure for startups and token sellers. It will advise token sellers of their duties and all processes will be built such that they conform to the regulations. Given that utility tokens are allowed to be sold without KYC requirements in many jurisdictions, regulation and user experience seem to be reconcilable. This approach will allow teams to focus on their product rather than the complications of a token sale.

And that’s it. Hopefully, this platform will allow other teams to raise capital, get traction, and make our world a better place.

Should the GnosisDAO initiate the development of such a product?

References:

Smart contracts proposal

8 Likes
2

Here are the 4 main reasons why it should be built:

  • Proven product-market fit via IDO on GPv1
  • Profitable business: Token sale volumes will likely increase significantly in the coming years
  • Product fits well into many mentioned GnosisDAO visions
  • simple design, quick to build and to iterate

There is one concern:

  • There are no technical network effects for the platform. All network effects need to be generated from the branding and reputation of the platform.
3 Likes
3

Thank you for this very detailed and thought out proposal!

A few questions:

  • I assume the business model would involve a protocol fee as a percentage of the transaction volume?
  • Is there any way to prevent someone to buy almost all tokens within the last block by diluting everyone else? This would be most likely against the interest of the seller interested in a wide token distribution.
5 Likes
4

Yes, I think so, unless someone has a better idea.

This would be quite hard to implement without KYC, as this rich person could use many accounts.

2 Likes
5

Really nice idea to implement “direct listings” on the blockchain. I agree that there is likely going to be a 9-10 figures market value of IDOs per year (large cap US IPO value was $46 billion is 2019).

Given the simplicity of the matching and the lack of network effects (each sale basically starts from 0), I wonder how we can guarantee that the value is captured for GNO holders and projects don’t just use the same contracts with a smaller or 0 fee?

Would the main selling point be regulatory consulting that is hinted at in the post? Who should conduct this service? Gnosis Services GmbH?

2 Likes
6

I believe that new projects planning to do an IDO are seeking for reputable platforms instead of copy-cats. Imagine you would want to fundraise and gain your investors’ trust. How would you make sure that investors trust you, there is as little confusion as possible, everyone has legal and tax safety? I think the obvious answer is by choosing reputable platforms.

Just one example, if you wanna do an IDO in Germany, among others you have to tell the bafin where you want to list the token. If you want their approval, people have to choose good players.

yeah, maybe at the beginning, the Gnosis Service GmbH, later one specifically hired lawyers.

1 Like
7

Speaking from experience, beware of the regulatory hell that regulation and KYC/AML invites. Imagine been in meeting after meeting with people dictating how your dapp must run without any knowledge of how decentralised applications are inherently different from centralised ones. No one knows how to regulate this and those that try just impose existing regulation so the Dapp looks like, well, an app. It kills the very spirit of decentralised applications and you might as well build the entire app on a centralised architecture.

That said if you decide to go that route and have customers lined up who are dying to run their permissioned, KYC’d IDO and a bunch of investors who would prefer to buy tokens with a padded lining of regulatory compliance, well there is an entire tiered automated KYC/AML app designed for decentralised apps at your disposal :slight_smile:

8

I agree with Alex if the fee is small enough (0.3% seems reasonable) and the platform is trusted it seems unlikely to me that the contracts will be forked for a single-time event.

Regarding execution: I think the reason why IDOs were conducted via Mesa was not only the design of GP but also because of dxDAO operating it. We should evaluate all options but I definitely think it is an option to have dxDAO being the operator of this product. The protocol fee could be split.

5 Likes
9

To my understanding, KYC is not needed according to the MIcA proposals out there for selling utility tokens. If KYC is needed, then this would indeed change the picture.

Here are some regulations required according to current circulating MiCA proposals (my current understanding):

  • Must set rules on admitting and suspending crypto-assets sent or received by platform.
    (This can be implemented straight forward and be managed by a DAO or legal entity, these are similar rules as USDC has, and still, everyone uses them)
  • Privacy tokens that deter user identification and transaction history are prohibited.
    (Auctions with privacy tokens could be stopped by an owner of the smart contract)
  • Operators may not trade on their platform with their own accounts.
    (Easy to fulfill)
  • Platform must continuously publish bid and ask prices as well as transaction volumes
    (Such information is accessible from the blockchain)
  • Transactions must be settled on DLT on the day they happen
    (Easy to fulfill)
  • Prevention of market abuse (Title VI)
    (The design prevents practically all common market abuse scenarios, like front-running etc)

I think the landscape changed quite some since your efforts, GraemeB.

But I am not a lawyer and this is just my current understanding.

10

hi All

I just jump in, also because we have a discussion with alex on our keybase chat right now and he did link this post.

I do product ownership for Mesa/DXdao. I started last week and did have a call last week with chen.

I also have been on the community call about GPV2.

We should have GnosisDAO x DXdao call next week about our future collaboration and ideas how to move forward together. I suggest 4-5 pm CET sometimes next week. I would prefer tuesday, but also have to check if the relevant person from us can take part.

I had plans to do a forum post for this only, but now things are rolling. I sugest I still can do it because this post is not the right place for this.

Made a post on its on about this: DXdao x GnosisDAO Partnership Discussion

2 Likes
11

Mesa.eth has traction as an IDO platform. Given the difficulties sales have encountered and now with Gnosis officially saying GPv1 is not recommended for IDOs, I think there is an immediate need to adapt Mesa so that it can continue to serve IDOs. Would be great to work out a partnership between GnosisDAO and DXdao. I think this is just one of several potential points of collaboration. Omen, governance infrastructure, and Swapr/GPv2 are other places ripe for collaboration.

For reference to historical volumes on GPv1: Gnosis Protocol

I would need to go back and check dates but I believe the spikes are a result of specific IDOs:

Screen Shot 2020-12-03 at 10.29.09 AM
Screen Shot 2020-12-03 at 10.29.09 AM1260×808 44.5 KB

5 Likes
12

Not sure if you are speaking about forking the contracts or the platform (meaning front-end) or both… but in our fascinating world of open-source, you can assume that forks will happen for sure.

The contracts maybe, but the platform/front-end for sure will be forked. We have already been seeing it with Mesa. Definitely for “geo-blocking” reasons. Potentially for fee reasons.

So, where the fee is built into the system matters. However, like you and others have mentioned, there is also a brand and services that be can built around and packaged into these events, so stickiness can be created.

Looking forward to seeing this discussion progress.

13

What about a modification - basically implementing the logic you described as a “smart contract order” for GP v.2.

The idea would that it is always settled over GP v.2 but the internal logic of the contract is executed as well and the clearing price that is achieved by the described auction logic is the MINIMAL price the seller should get.

So we are turning this into a general tool for people to do protected market orders on GP. That can be an IDO, but it could also be a e.g. Maker liquidation.

The advantages with connecting it to GP are (for the seller) in the case of already liquid tokens access to other on-chain liquidity and in the case of IDOs it would allow buy-in with other tokens (ring trades).

3 Likes
14

Hi Alex thanks for writing this up.

I really like the idea of developing something where we have encountered product market fit. I believe our strong advantage is around fair price finding + frontrunning resistance.
I believe Gnosis has been a pioneer of fair price finding since the days of the Gnosis ICO and the DutchX.

Whether we build a separate or do it with GP V2, I think it is important to have IDO’s as a main use-case.

3 Likes
15

Could you explain this more in detail, please? How exactly is the price determined, how binding are bids in this auction model?

One could definitively make this auction model atomic: One could allow for “last-second orders”, being submitted and settled together with the price-submission + verification. These “last-second orders” could come from a gp-v2 solver and include trades with ring-trades. Are you referring to this setup?

If this is the idea, then yes, we could combine these two products, eventually.

16

So in my view, the rules could be very simple:

Settlement always happens from GPv2.
This contract will refuse settlement if:
a) the price for the seller is worse than the calculated price based on all the bids
b) all bidders above the clearing price need to be considered

17

I agree with you, it really depends on the infrastructure around the contract itself to prevent forks of contracts. For the business case, it is important to have no forks of the contract as the fee occurs there. I see the UI itself already as part of the infrastructure around the contract, which can make forks less likely. But as we saw with Mesa this in itself is not sufficient. In the end, the best way to make contracts irreplaceable are other smart contract integrations.

2 Likes
18

Maybe one way to protect from forks would be to include slightly more involved infrastructure which is harder to clone. As an example VDFs could be used to seal and later on automatically reveal bids. Running a VDF revelation service is likely not something a team would easily clone for their IDO.

Generally I also like the idea of including this use case in GPv2 for the reasons Martin mentioned (ring trades allow people buying the asset with any token, same concept can be used for liquidations).

3 Likes
19

I think this is a complex decision and listing the pros and cons of both solutions would be helpful.

Advantages of integration into GPv2:

  • Only one system for everything: combines liquidity and can leverage on-chain liquidity.
  • Ring trades allow people to buy assets with any token.

Disadvantages:

  • Presumably more complex to implement?
  • Amount of participants per batch is smaller but there could be multiple batches.

Advantages of a standalone contract:

  • Can have up to 2500 participants in one batch.
  • Last/closing order could also leverage on-chain liquidity.
  • Very simple contract architecture and dependencies.

Disadvantages:

  • One buyer could dilute everyone placing his order last.
  • No network effect.

Am I missing something? Maybe it would make sense to start an external document (hackmd?) to collaborate on a complete list.

IMO, looking at the IDO use-case only, leveraging on-chain liquidity is unimportant, and afaik in past IDOs most used the same token for buying assets, and ring-trades didn’t add many benefits.

The upside of having IDOs on GPv2 would be that it is more likely that liquidity will just stay on GPv2 after the IDO has been completed.

Looking at the liquidation use-case on-chain liquidity becomes much more important to get the best price.

2 Likes
20

I think we can all agree that the IDO use case is very specific and it makes sense to have a solution that is optimized for that scenario.

The difference between 2,500 participants in one batch and the 30 or so participants in one batch on GPv1 is HUGE.

Finding a fair clearing price among all (or most) of the participants in a sale is just not possible when only 30 trades can be included in a batch and you leave out the smaller notional buyers on top of that.

2 Likes