IIP-54: BED:ETH liquidity - Direct provision by Index Coop

Governance Proposal
1

IIP: 54
Title: BED : ETH Liquidity - Direct Provision by INDEXcoop
Status: Proposed
Author: @Matthew_Graham & @overanalyser
Created: 08 July 2021
Requires: IIP-46: Security - Operations & Investment Account

Simple Summary

INDEXcoop to provide $250K of seed liquidity to the BED:ETH pool on Uniswap v3 until 30th September 2021. This liquidity is to be administered from the Operations Account multi-sig wallet.

Abstract

INDEXcoop multisig to transfer ~$250,000 (all ETH2x-FLI, BTC2x-FLI + some DPI) from the Index Coop Treasury to the Treasury Operations Account where it will then be sold to ETH.

The Majority of the ETH will be used to attain BED via exchange issuance on TokenSets user interface. BED and ETH will be deposited into Uniswap v3 liquidity pair (0.3% Fee) with a tight range to provide deep liquidity (~1% price impact for $10,000 trade). This will occur prior to the official BED product launch.

The pool will be maintained by the treasury operations multisig until 30th September 2021 and then returned to Operations account as ETH.

The Operations Account will manage the liquidity and @Overanalyser will be added to the multi-sig and will be the primary person managing the liquidity position.

Motivation

$BED will shortly be launched by Index Coop and Bankless DAO. As a three component product, with significant liquidity for the underlying tokens, exchange issuance is expected to have a low gas price and price impact. However, relying on exchange issuance as the only source of $BED means that many familiar purchase options (metamask, zapper, zerion, 1 inch, 0x, DeBank, other wallets) which rely on secondary markets will not work.

As BED is primarily aimed at customers at the start of their DeFi journey, such a limitation to exchange issuance is not desirable.

As a passive long term product, we don’t expect large trade volumes, so there is unlikely to be sufficient trade volume to make LP’ing attractive.

Previously the coop has used Uniswap v2 for liquidity mining campaigns and if we repeated IIP-24 (MVI) targeting $5M, we would expect to spend ~$140,000 per month on liquidity mining. Note: $5M liquidity in v2 allows ~$18,000 trades with 1% price impact. For BED, exchange issuance / arbitrage will be cheaper so a 1% depth of $10,000 should be sufficient (~$2.8 M on Uni v2) which is likely to require ~$80,000 INDEX rewards per 30 day LM campaign.

With the launch of Uniswap v3 we have the option to move away from LM on v2.

Specification

Concentrated Liquidity bounds

Uni v3 allows us to set the bounds of the liquidity. For example $250,000 placed at market price +/- 12.5% results in a $10,000 depth for 1% price impact. This should allow small trades to happen and efficient arbitrage to keep the price close to NAV.

The ideal pair for such focused liquidity provision for this approach is one with minimal expected price ratio change. As such BED:ETH is a very attractive pair.

  • Over the last 90 days DPI has ranged between 0.113 and 0.225 ETH (0.169 +/- 33 %)

  • BTC has ranged over 12.3 to 28.3 (20.3+/- 40%).

  • As an absolute worse case, BED:ETH range would been +/- 24% [(31% + 40%)/3].

What are the downsides?

  • The BED:ETH ratio moves out of the range we set and so there is suddenly no liquidity available (This will result in very poor purchase experience as the market becomes very thin).
  • Concentrated liquidity also means that we will suffer a larger impact of divergence loss. This means that as the price ratio moves to the limit of the bounds we hold a larger weight of the lower performing asset. This means that we could end up holding 100% ETH or BED.
  • Rebalancing will cost management attention, multisig time and gas. If we place too narrow bounds for the pair price ratio then we will need to rebalance and reposition them quickly.

Mandate

The Treasury Operations Account muti-sig will have the mandate to manage the $250K of ETH:BED liquidity to maintain a ~1% price impact from $10,000 trades. Signers on the multi-sig account are shown below.

Roles Operations Account
Treasury Committee @DarkForestCapital, @dylan
Treasury Work Group Lead @Matthew_Graham (:fire:Fire​:fire:)
Account Manager @prairiefi
V3 Liquidity Manager @overanalyser (to be added)

Process

The Treasury Operations Muti-sig was originally proposed as 3 of 4, with the addition of @overanalyser this will change to 3 of 5.

  1. Call Fee Collector contracts on FLI products
  2. INDEXcoop multisig transfer $250,000 of FLI products + DPI from the Treasury Wallet to Treasury Operations Account
  3. Swap to ETH via most liquid DEX pool
  4. Convert majority of ETH balance to BED via Exchange Issuance on TokenSet user interface
  5. Deposit BED & ETH to Uniswap V3 across a narrow range as detailed in specification section above
  6. Fees are claimed and liquidity repositioned (with swap trades / issuance to rebalance)
  7. Retrieve liquidity from liquidity pool
  8. Swap BED for ETH. ETH is held in the Operations Account

This Liquidity provision will stop on the 30th September 2021.

IIP vote specification:

FOR

  • Transfer $250K to Treasury Operations Account for use as BED:ETH liquidity on uniswap v3 until 30th September 2021

AGAINST

  • No change.
9 Likes
2

Love the move to protocol owned liquidity think it’s a positive change in approach and a great way to save on LM costs in future.

Can you explain a bit more about the management side of this? Looks like by using the data we can specify a range that shouldn’t require too much interference, but how do you intend to keep track of the price, how often will you check and is it only when price goes ‘out of range’ that a shift will take place?

3 Likes
3

The advantage of BED : ETH is that it should be a fairly stable pair. So we can do some back testing on how the price would have behaved over the last 30 or 90 days.

When watching the MVI launch I saw premium to NAV of up to 20%. (With bots mimicking any exchange issuance and then selling in the same tranaction) I think mainly driven by people wanting to get tokens to stake for LM. So I would probably launch the pool with more BED than ETH in it so people can buy BED and there is more depth. So maybe we put 70% of the Liquidity as BED and 30% as ETH for the launch.

I would plan to have at least two positions for the liquidity. Then as they approach the limits, we withdraw one, swap to rebalance, and then redeposit. Then we know we won’t pull all the liquidity from the pool… (I did a similar reposition for uni v3 and it took me 40 minutes, [Higher gas would have made it quicker, but a multisig will slow it]).

Two identical pools would work with this approach, but something more complex may be better e.g.:

  1. $80,000 over +/- 20% gives $1,951 per 1%
  2. $80,000 over +/- 10% gives $3,810 per 1%
  3. $90,000 just +10% gives another $8,182 per 1%

So we get something like:

image
image933×582 14.8 KB

Which means we have $14,000 depth for 1% buy price impact and $140,000 of BED purchases to move the price 10% (but only $6,000 of BED sell to move 1%).

Position 1 should be safe for a long time, 2 and 3 may need more frequent management.

Monitoring
Initially, I think we will be monitoring it multiple times a day. Ideally, I would want to get to a point where we only move things once a week and we can schedule it so we can avoid weekends / unsociable hours.

Uni v3 positions are pretty easy to monitor as anyone who knows the NFT number can see it online. This is a ETH:DPI position in range

image
image880×852 109 KB

And this one is out of range:

image
image857×879 104 KB

I think BanklessDAO will be doing something similar, and individuals may join the pool so we may see it get deeps, so once we observe live bots issuing and arbing the pool, we can move to wider ranges and less management.

2 Likes
4

Hi All,

Just highlighting here this forum post relies on IIP-46 being progressed to snapshot. The only people with Governance Forum editor privileges and SnapShot admin access are Set Lab employees.

@dylan, @anon10525910 & @puniaviision - may we please have IIP-46 progressed to snapshot to enable this draft IIP to progress soon thereafter ?

1 Like
5

Good Morning Matthew :wave:

Can you please respond here and let me know what day you want the IIP-46 snapshot to run?

Also, per IIP-26 (99% FOR), there is a 48 hour waiting period before a vote can be called (Gitbook).

I am noticing this was posted 5 hours ago. [Edit] Happy to do the IIP editor work and queue this up to kick off Saturday, if you & OA so please!

6

Following IIP Step-by-Step - Index Coop Community Handbook I would request that this goes to snapshot on Monday 12th July 2021.

I’ll highlight any changes to the proposal during discussion.

2 Likes
7

I think Monday would be great @anon10525910 :slight_smile:

1 Like
8

I’m very opposed to this aspect of the proposal. I would much prefer to keep treasury as is vs holding BED.
Debt is still the cheapest way to buy liquidity with Liquidity Minting. If we want LP positions on treasury balance sheet might as well make them maximally profitable.

From my convos with coop members about DEXs I don’t think we have the experience to easily and effectively manage a v3 position. With such a small $ amount i don’t think the operational/educational expenses would even be worth it.

1 Like
9

This has been upgraded to an IIP and announced in Discord.

Snapshot vote has been called for Monday, which is after the 48 hour waiting period.

As such this will be queued for Snapshot vote on Monday.

10

The following line was added to the body of this proposal, as it is understood IIP-46 is indeed a requirement.

Correct me if this is wrong.

1 Like
11

I like this proposal in spirit, but I have several questions/requests:

  1. I can’t make an informed decision regarding how to best use a portion of treasury assets without an accounting of what’s currently in the treasury overall. I’d like to know what percentage this represents on a vested and overall basis - this would effectively be a loan so knowing what our other commitments are would be useful.
  2. Will IC be launching a bot to keep NAV and spot aligned? I ask this because from my perspective the primary goal of creating a market would be to keep the UX positive by not having spot stray from NAV, with slippage as an important secondary consideration; the seed pool only seems to address the latter.
  3. Do any aggregator or AMM protocols currently plug into Set’s exchange issuance module for any of our other products, and could they easily do so for BED? Meaning can we push people toward the more efficient options?
  4. What is the estimated cost of issuance given that it’s only 3 tokens? (for example, if I’m supplying ETH I’m only swapping for two tokens and wrapping, I cant imagine that costs much more than say 4x a 1for1 AMM swap) - I’m thinking unlike DPI, pushing people directly to the E.I. module at launch may be the preferred marketing strategy.
  5. What are Bankless DAO doing to incent liquidity and how does this strategy align?

I know I’m asking a lot at once, but this seems to be moving quickly. Thanks @overanalyser

  1. Afterthought - would a multi-asset zero-slippage pool of BED and wBTC/ETH/DPI make sense here, or does that not solve the depth problem to the extent we’d need it to?
3 Likes
12

@anon10525910 @overanalyser @Matthew_Graham why are we pushing this vote so fast when hardly any conversation has happened and BED isn’t even launched yet?

You haven’t even detailed if or how we will be adjusting the liquidity position if it gets out of range which is a pretty important detail because if we aren’t in range this whole thing is useless.

2 Likes
13

as an editor, I am following the governance process and the authors’ direction.

unless one of the authors posts publicly here to postpone the vote, it will go forward.

@overanalyser, @Matthew_Graham, please advise.

1 Like
14

We are pushing for a quick vote because we want the liquidity in place before the BED token is announced. Launching a product aimed at people new to DeFi without some secondary market liquidity will risk some people buying at a large premium.

For MVI we launched liquidity mining on uniswap v2 at the time of launch. This approach avoids doing any liquidity mining for BED.

The liquidity will be repositioned when required. The exact parameter used / procedure to be followed have not been fully researched / written. The IIP is intended to ask INDEX holders to grant:

  1. permission to use coop resources ($$$ and peoples time)
  1. A mandate to manage the liquidity on behalf of the coop.

I don’t think that we need to have a fully scoped out implementation to allow INDEX holders to make a considered descision on the IIP.

15

Hi Mel, you raise some great questions, and I don’t have good answers to all of them, but this is my best shot:

INDEXcoop Multisig (recipient of INDEX from vesting contract, and streaming fees) is 0x9467. and contains $46 M. This is currently controlled by SET team members. Part of this proposal is to transfer the $250 k to the Treasury Operations Multi sig (detailed in IIP-46) which will manage the liquidity.

I’m not sure about whether we will have an arb bot using issuance running for launch.

For MVI, there are bots running that issue and sell / buy and redeem MVI. This recent trade was ~6.25 ETH ( ~$13 K) trade and captured ~ $95 profit. The exchange issue and trade cost 1,706 k gas at 12 gewi ($43 fee).

When MVI launched, I don’t think there was direct arbitrage. However, watcher bots would copy some exchange issuance (including one of mine) and then sell in a single transaction to capture premium. E.g.:

  • This trade minted 118 MVI from 5 ETH. @ 165 gwei
  • Was front run 2 blocks earlier by this trade (@265 Gwei) which burnt CHI, copied the mint (118 MVI from 5 ETH), and did a simultaneous sale to capture ~0.6 ETH profit.

So at launch there was premium to NAV worth multiple ETH at times (I think I saw the uniswap pool sitting at a +20% premium).

Non-copy trade issue arbitrage appears to start with this trade @08:04 on 07APR21 (~1 day after the uniswap pool was created) : which looks like a bot targeting the MVI pool.

Arbitrage bots will only strike when they can make a profit which is related to both the depth, and the % premium to NAV. So we need bopth depth and a premium. If it takes $10,000 to move the price 1% off NAV, then there is less than $100 available to arbitrage. (Actually, I think it it’s linear over short ranges, it could be ~$50 as the first $1,000 of trade will only yield $1 as it’s 0.1% price impact). So, gas costs will make it hard to keep clost to NAV.

Following this logic, I think that a $50,000 trade with 5% price impact (i.e. five x $10,000 per 1%), will have $1,250 profit available.

Not at this time, but getting trade aggregators to include exchange issuance is something the BD team are working on.

Lots of people still use the uniswap interface directly, andI don’t see us ever getting Uniswap to integrate exchange issuance (we have lots of educating to do).

MVI Exchange issue and sale (x15) ~1,800 K
DPI Exchange issue (x14) ~900 k
CGI issue /redeem (x3) ~200 k

V2 Uniswaps are ~100k (and my memory tells me that additional tokens for exchange issuance are ~80 k). So, I would say we are looking at 400 k to 500 k for Exchange issuance of ETH to BED. At 14 Gwei we are looking at $15 to issue.

I’ve discussed liquidity mining with BanklessDAO in their forum, but I’m not involved in the goto market discussions. Hpowever, it’s my understanding that Bankless are planning to provide liquidity via the same Uniswap v3 pool. So what we are proposing should be the minimum liquidity available. If we implemnet this IIP, then I would plan to talk to them to discuss our plans and co-ordinate.

I’m a fan of having multiple liquidity pools available so such an approach is attractive. (There is no reason not to include INDEX and MVI in such a pool). However, I think we need a single main (deep) pool on L1 for the issue arbitrage bots to target (and the other bots trading between pools). The last tiem I looked DPI issue arbitrage was via the v2 pool only, thenthe other pools mathed that pool.

For the moment, everyone looks to Uniswap first so that is the simple route to market.

I should note that this approach does have risks. If the BED:ETH price moves out of range, then we will be holding the lower performing asset, and to reposition we will need to buy the other, so we could easily lose 10% of the value compared to just holding 50% ETH and 50% BED in a cold wallet.

1 Like
16

Chiming in, I think this is crucial for new people to have a good experience with IndexCoop. It is pretty clear that both DAOs want this product to be directed towards new users as an introduction to Crypto investing. If I was to recommend this product to friends and family, I’d like them to avoid technical issues, astronomical fees and high slippage.

I love the spirit of the proposal, thanks to @overanalyser. Just got to find a good implementation.

1 Like
17

Isn’t that the major function of PWG?

Doesn’t sound like you are confident Bankless will provide liquidity either. Isn’t one of the key factors for a methodologist to provide sufficient liquidity?

18

This is now queued for Snapshot vote that will begin tomorrow, July 12

https://snapshot.org/#/index-coop.eth/proposal/QmU22KuNvzr289Pd5K2wo97hkiE4m24R1D6GF3hQ1tAotJ

2 Likes
19

Reposting from the #delegates channel in discord (Index).

While generally supportive of this IIP, will not be voting until either the authors (@overanalyser @Matthew_Graham) or someone from the GTM team (@LemonadeAlpha ?) can give a concrete answer on whether or not Bankless Inc or Bankless DAO will be providing launch liquidity and if they are, what the details are.

This information is essential for an informed vote and I look forward to it being posted here and in the #iip-054 channel in Discord.

5 Likes
20

Since drafting this forum post, I have learnt that there is a need to do a bit of home work on the legal front to determine if Index Coop risks attracting regulator attention by being a market maker for its own products.

Neither @overanalyser or I have sufficient knowledge to know if Index Coop by being a market maker of its own products would be an issue or has the potential to be an issue later on.

From my point of view, I don’t know enough to draw a conclusion on the legal ramifications and therefore, I believe it would be prudent to seek some advice before progressing further with this proposal.

4 Likes