IIP-78 DG2 - Yield Hunter

Title: Yield Hunter Index (YHI)

Status: Proposed

Author: @overanalyser

Created: 12th October 2021

Simple Summary

To create a diversified crypto fund of ERC20 tokens that have the opportunity to capture yield in addition to the underlying token price exposure. Stable coins and wrapped L1 tokens are excluded.

Abstract

To create a diversified fund using income generating ERC20 tokens on Ethereum main net which allow income generation to the benefit of the holders. The product will be built using Set protocol with liquid tokens purchased on DEX and wrapped tokens.

Tokens can include DeFi, Metaverse and Infrastructure projects and tokens from centralised projects. However, the methodology excludes ETH, (ERC20) BTC and stable coins.

The v0.1 proposal is intended to be a Minimal Viable Product (MVP) to allow rapid launch and future upgrades to capture higher yields. The structure will be a simple Market cap (capped at 25% for any component) with a simple streaming fee for holders. Income generation will be limited to the use of AAVE, Compound and Cream and DEX liquidity.

This will be an INDEXcoop methodology maintained by contributors.

Motivation

‌Some protocols are already generating income for holders via staking, or by issuing ongoing reward tokens. However, it requires time and attention to manage such positions to ensure balanced exposure and ongoing income capture. Many potential users of our products have stated that lack of income generation is a negative when considering a purchase.

In addition, some protocol tokenomics are such that the passive holder is at a disadvantage compared to those who are able to stake/lock up tokens or wait for vested returns. While some of the tokenomic designs are unfavorable for direct inclusion of staked tokens within an index due to redemption issues, they do encourage borrow demand. This means that some income can be captured by being a lender on a liquid money market.

With the creation of Set protocol adaptors for Compound and AAVE there is an opportunity to build products that capture some income from a wide range of tokens. By forking the current adaptors for different tokens, and developing wrap adaptors for the Cream protocol, we now have the opportunity to quickly launch a product that captures yield for our users.

‌Rationale

‌The key considerations in designing this methodology are:

  • Quality and security of the underlying projects
  • Net positive income for holders at all times.
  • Accessible income generation which allows unrestricted issue and redemption.
  • Minimal development work to allow launch.
  • Community methodology to empower INDEXcoop contributors and remove friction from the launch and maintenance process.

The product is intended to be used by people wanting both price exposure and income generation from a passive product that can be used within a wider (user defined) income generation strategy including ETH, BTC and stable coins.

A underlying token cap of 25% has been applied to avoid over concentration

A free floating market cap design was also considered as it would reduce the size of monthly rebalancing. However, the overconcentration is considered a larger challenge to holders.

Alternative income generation using yield generation protocols (e.g. Yearn and xtoken) would unlock significant additional yield in the future, but is not considered necessary for product launch.

While a management fee taking a share of the income might better align the interests of holders and INDEXcoop, a simple streaming fee of 1.95% is proposed for this product launch to minimize engineering work. Likewise, a mint and redeem fee has been discounted from the v0.1 methodology. In order to ensure net income generation for holders, an income threshold for each component has been set to 2.5%.

Products with built-in burn mechanisms to improve underlying token price (e.g. MKR) are included in the methodology when they meet the other criteria.

Specification

Overview

‌This product will be built on Set protocol contracts and the v0.1 specification will use a combination of:

  • DEX liquidity.
  • Compound, AAVE and Cream wrap adaptors.

Future iterations may include wrap adaptors for token staking contracts, additional money markets (e.g. Fuse), and yield protocols (e.g Yearn and xToken).

Rebalances will be calculated by the methodologist (Coop contributor work team) and will be monthly.

Token selection will not be restricted to individual sectors but will exclude ETH (ERC20) BTC and stable coins.

Differentiation

Unlike Sector-based products ($DPI, $MVI, $BED, and $DATA) the availability of income is a key selection criteria. In addition, the methodology does not limit itself to a single category.

Unlike leveraged products, this does not use any leverage or borrowing.

Compared to the
$PAY proposal, this product captures the underlying token price fluctuations. In addition, token weights are based on circulating market cap and not on yield or protocol risk tranching.

Compared to the $LDI proposal, this product targets yield generation for all components. In addition, this product is intended to be part of a larger portfolio where the holder can complement it with exposure to ETH, BTC and stable coins based on their preference.

Competitor products, $DeFi+L and $DBI use income-generating tokens. However, they do not use income generation as a selection criteria, contain non productive components, and are restricted to DeFi.

Revision following DG1

Following the DG1 vote a more detailed review of yields and token inclusion has been carried out. This was captured in the PRD supplied to the work team assessment: YHI PRD 29Sept21

The work team assessment can be found here

Following the PRD preparation the methodology was adjusted again to reduce the number of tokens included. Over all the changes compared to the DG1 proposal IIP-78 are:

  • Allowance of centralised project tokens (e.g. FFT, CEL, WOO)
  • Use of 7 and 28 day average incomes
  • Threshold increased to 2.5% income
  • Streaming fee increased to 1.95%

This has the benefit of fewer tokens, higher yields and less overlap with other INDEXcoop products.

The example composition at 10 October 2021 is as follows:

Token Weight 7 day income 28 day income
FTT 25% 6.58% 11.59%
xSUSHI 12.6% 6.24%* 6.24%
COMP 11.7% 6.77% 10.35%
SNX 11% 3.34% 2.57%
YFI 7.4% 3.57% 3.57%
CRV 6.6% 29.76% 13.88%
MANA 6.4% 8.34% 4.47%
BNT 5.4% 2.83% 13.69%
SAND 4.2% 9.49% 15.52%
1INCH 3.4% 6.00% 11.87%
WOO 3.2% 10.18% 3.77%
OGN 1.6% 2.62% 2.66%
BAL 1.4% 10.10% 10.15%
Weighted average (before fee) 7.8% 8.9%

Note * Includes xSUSHI income and AAVE rewards.

Size of opportunity

As the use of income-generating tokens makes listing on money markets or CEX exchanges more difficult, this product may have a restricted growth profile compared to $DPI, $MVI, or $DATAdue to reduced extrinsic productivity use cases/composability.

However, the capture of income for benefit will attract the attention of many more sophisticated/crypto-native traders. In addition, the inclusion of tokens not in our current products (CRV, BNT, 1Inch, FTT, WOO) will attract members of those protocols communities.

Two DeFi centric products have income-generating components:

  • $DBI with $14.5 M AUM, benefits from liquidity marketing with 41% in incentivized pools. $BDI claims a 4% income (due mainly to ySNX).
  • $DeFi+L has $3.6 AUM and captures ~ 0.6% income for holders.

The current AUM for the example portfolio yield generating tokens in the lending markets is currently $284 M. This indicates that there is a clear demand to borrow the tokens in the proposed fund and if we capture 10% of the supply we will have > $28M AUM.

Initial estimates for AUM

  • 6 Months $ 10 M with $2 M in incentivised liquidity
  • 12 months $ 30 M with $2 M in incentivised liquidity
  • 24 Months $ 90 M with no incentives

User Stories

“I want to take a step back from daily transactions to manage my portfolio. Having $YHI allows me to be exposed to safe income generation on top of exposure to overall price action. $YHI in combination with Staked ETH, BTC on Celsius and yUSD allows me to balance my risk profile and generate income”.

“Our DAO treasury has significant ETH and Stable coin diversification with both generating income. We need something we can buy and forget that gives us broader exposure to the Ethereum ecosystem while avoiding dilution. Being able to generate yield means it’s paying for itself.”

“I want broad exposure that I can forget about. Generating income without needing to manage borrowings suits my risk profile”

“Holding income-generating tokens within single wrapper makes my tax reporting so much easier”

Methodology

Token inclusion / exclusion criteria

Tokens will be selected with the following criteria:

  1. ERC 20 on Ethereum main chain
  2. Not a stable coin
  3. Not ETH or ERC20 BTC
  4. Not a security
  5. Over 6 months project life
  6. Top 250 by market cap of all tokens on Coingecko ( > $250 M circulating MCap)
  7. Annuallised income must be over 2.5% over 7 and 28 days
  8. Income generating token available by:
  • DEX liquidity
  • AAVE, Compound or Cream vaults

Criteria may be adjusted to produce a product containing 10 to 20 tokens.

Index weight calculation

Components will be selected on the basis of circulating market cap with a per token limit of 25% of the total portfolio value.

Technical on-chain specification

Technical complexity

Use of money market wrappers adds complexity in a number of areas:

  • Issue and redeem requires interaction with different protocols.
  • Exchange issuance becomes more complex and is not currently implemented for any sets.
  • Likewise, arbitrage issue and redemption contracts are more complex than DEX based products (but likely outsourced (e.g. flashbots))
  • Calculation of NAV for deposit tokens is more complex and is not currently implemented.
  • Reblances for wrapped tokens are not currently automated and so will require multi-sig transactions by devs. However, we do avoid trade size/price impact issues.

In addition for some potential tokens, the best liquidity is on protocols not currently supported:

  • Bancor (Link and BNT)

Note, many of these challenges are also present in the PAY and LDI proposals.

Costs

Cost to customer

For the v0.1 methodology, a 1.95% streaming fee will be charged.

Reward tokens available to casual users (AAVE, COMP, etc) will be captured for the benefit of holders (and are included in the income calculations above).

Cost to mint / redeem

For a 12 component fund split between the money markets and DEX swaps from ETH:

Protocol # Gas Total
AAVE 3 195 k 585 k
Compound 3 220 k 660 k
Cream 3 185 k 555 k
DEX trade 12 110 k 1,320 k
Issue 1 170 k 170 k
Total 3,290 k

At a gas price of 50 gwei 3,290 k would cost 0.165 ETH

Notes:

  • 12 approvals would add a further 600 k (0.05 ETH at 50 gwei)
  • MVI exchange issuance for 15 tokens uses ~ 2,500 k gas.

Rebalance frequency:

Monthly following methodologist instruction.

Manual Rebalance magnitude:

Income redistribution between the underlying tokens is expected to be 1 to 2 % per month.

Rebalancing due to underlying token issuance is expected to be 1 to 2%

Having a 25% cap may introduce additional rebalancing when the capped token under/overperformed the average. This is estimated at 5%. [Note DPI with a 25% cap on $UNI has $UNI rebalances of 1.55%, 5.53%, 4.55%, and 2.53% since April 2021 ].

Token addition or removal is expected every few months and may be up to 25%.

An average of 5% to 10% is expected each month.

Fee split

All streaming fees will go to INDEXcoop. All methodologist bonus rewards will go to INDEXcoop.

Methodologists will be paid by INDEX coop contributor rewards/working group rewards.

Meta / intrinsic productivity

If available, metagovernance rights will be held by INDEX holders.

Reward tokens available to casual users (AAVE, COMP) will be captured for the benefit of holders.

Protocol level rewards (e.g Yearn Affiliate rewards) will be captured by INDEX coop.

Liquidity

An $YHI:$ETH pair is proposed as default as YHI is expected to be volatile vs stable coins.

A xy = k pool or a concentrated liquidity pool may be considered.

Market support

Seed liquidity and any liquidity mining incentives would need to be provided by INDEXcoop.

A $5 M xy=K pool (uni v2 or sushiswap) is suggested. This would have a LM cost of ~$300,000 per quarter.

‌Author Background

I (@overanalyser) have been an active member of the INDEXcoop since its launch in October 2020. I have a keen interest in on-chain index products and liquidity and I was the joint Product working Group lead (with Punia) between April and September 2021.

It is my intention that if successful, I will form a small work team (myself with 2 or 3 others) to support the launch and provide ongoing maintenance. This team would be managed by the PWG and be paid via the PWG contributor budget.

Marketing support / distribution / partners.

All marketing would need to be provided by INDEXcoop.

Revision history

The initial proposal for discussion - 12 August 2021.

Typo fixes for DG1 IIP vote - 20 August 2021

Updated DG2 proposal - 12th October 2021

Links

Initial proposal

DG1 snapshot

Community call and slides

Work team assessment

Product requirements Doc

Copyright

‌Copyright and related rights waived via CC0.

10 Likes

Some back test data for the example composition based on prices and average income 23 Aug to 10 Oct (48 days) and 13 Sept to 10 Oct (28 days).

Token Weight (10Oct21) 28 day price change 28 day income 48 day price change 48 day income
FTX 25.0% -24.0% 0.89% 11.7% 2.94%
Sushi 12.6% -7.7% 0.58% 4.0% 0.99%
COMP 11.7% -18.2% 0.79% -27.7% 1.11%
SNX 11.0% -7.1% 0.20% -6.6% 0.31%
YFI 7.4% 2.1% 0.27% -13.9% 0.47%
CRV 6.6% 20.2% 1.06% 19.8% 1.32%
MANA 6.4% -1.0% 0.34% -5.0% 0.36%
BNT 5.4% -2.0% 1.05% -5.9% 3.76%
SAND 4.2% -1.7% 1.19% 22.3% 1.19%
1inch 3.4% 7.0% 0.91% 22.1% 2.71%
WOO 3.2% 106.0% 0.29% 49.6% 0.38%
OGN 1.6% -10.5% 0.20% -24.8% 0.33%
BAL 1.4% -18.2% 0.78% -27.3% 1.35%
$YHI calculated -5.4% 0.68% (8.9% Annual) 1.6% 1.55% (11.8% Annual)
ETH 0 5.0% 0 10.6% 0
wBTC 0 19.3% 0 11.9% 0
DPI 0 -4.9% 0 -19.9% 0
MVI 0 15.8% 0 10.4% 0

Overall price change has been worse than ETH, BTC and MVI, but outperformed DPI over both periods.

Over both periods the yield captured is smaller than the impact of price fluctuations.

Note: I’ve not modelled the effect of price changes / rebalances. e.g. The example composition 28 days ago would have had a larger weight of COMP, so the 18 drop in price would have had a larger impact on the %YHI price to the 10th October.

2 Likes

Hi,

Can this IIP be queued up for vote Monday 18th?

@gregdocter @Mringz @Pepperoni_Joe

3 Likes

Hey OA! I will be handling the IIP DG2 process.

Congratulations on making it to DG2! Look forward to the first yield based Index for the Coop.

3 Likes

Hi @overanalyser!! I am incredibly excited for YHI and your lead on the methodology here. It is apparent productive assets are what the current market wants. Considering the complexity and income possibilities a streaming fee of 1.95% is a very compelling product for even the most sophisticated crypto investors. At the current point in the market I believe we are underestimating some of our user base, including more knowledgable investors or “degens” or maybe even zoomers like myself. This has been shown through the performance of the FLI series. Furthermore, while having discussions with DAO treasuries, productivity and yield have been clear interests.

I am curious, however, how you see this product evolving? Considering the changes since DG1, do you believe higher and higher yield would lead this product to go further down the risk curve? Not that this would necessarily be a bad thing, but curious how we can set risk parameters; particularly around staked tokens. Either way, I strongly believe the market is looking for a simple index like this that takes advantage of DeFi’s composability.

Errbody hunting for that yield, $YHI makes it easy :bow_and_arrow:

5 Likes

Hey @overanalyser

Yield Hunter Index Has been queued on the 18th October 2021 here for vote:

https://snapshot.org/#/index-coop.eth/proposal/QmapbbsGeaqs2dU7Up51UXKZdt2jHuCZ7qkaok3apTouS8

2 Likes

Hi @funkmasterflex

Good question on getting higher yield vs increased yield.

My design choices were generally around security and simplicity with the goal of being simple to build and explain to holders at a cost of (slightly) reduced yields. This was done with the intention of getting to market quickly, and allowing further developments once we have demonstrated customer demand.

For example, the methodology uses 3 protocols to generate yield (Cream, Compound and AAVE). There are many others out there. The other lending protocols have less AUM and time deployed so I’ve discounted them from launch.

There are also Yield protocols that can give higher returns:

Token Yearn xToken Vesper
UNI 9.77%
LINK 4.35% 3.77%
SNX 3.99%
COMP 7.72%
BNT 34.32%
1INCH 18.59% 9.39%
Alpha TBD
AAVE 8.07% 8.71%

These could offer the benefits of higher yields or more token diversity within the product. However, using the yield platforms would require two things:

  1. more (and different) smart contracts prior to launch.
  2. more detailed understanding of the native token liquidity within the yield protocols.

#1 adds complexity and work prior to launch and so slows down this (and subsequent coop) product launch.

#2 relates to how easy the product can be redeemed (generally by Arb bots).

The lending protocols generally use similar models to calculate interest.
image

50 to 80% utilization generates 4 to 10% lending APY, and over 80% utilization is discouraged by high rates. So the pool will typically sit below 80%. This means YHI can be redeemed and tokens removed from the lending contract under the vast majority of trading conditions.

The yield protocols may be using the same lending protocols (and farming stable coins), or they may be depositing inyo staking contracts. E.g. stAAVE has a 14 day (?) cool down period, so there is a risk that if an arb bot wanted to go ETH → YHI → yAAVE (–> stAAVE) → AAVE → ETH in one transaction, it could be blocked.

The yield protocols try to manage this and provide liquidity (within the vault, or as DEX liquidity). But it would take time, coordination and engineering to managed. I didn’t think it was worth doing such work for an extra +0.x% on YHI.


Note, Other than brief comments, I’ve not discussed the engineering lift directly with the engineers. Edward confirmed that EWG are looking at AAVE and Compound wrappers, and I have assumed that Cream will be straightforward. The PRD identifies future upgrades for more yield, but these are not a launch requirement.

The work team analysis (which includes EWG) didn’t flag any technical blockers.

If the DG2 vote passes, then I expect we will launch with the three protocols as described.
If we get good AUM, then the cost / benefit of incremental improvements will be discussed (and prioritised).

1 Like

Hi @overanalyser

Can you please advise if Index Coop is expected to provide the following for YHI:

  • Seed Capital - if so, how much.
  • Liquidity Mining Incentives

Both of these costs are significant and should feature in the proposal.

What is the revenue split with the Yearn vault kick backs ? Are there any plans to negotiate such deals with other asset managers ?

@Matthew_Graham

As per the DG2 proposal:

I would expect that the coop might decide to use Uni v3 as the secondary market as this should be lower cost to seed / LM. We would also look at Wintermute / Onsen etc.


The v0.1 methodology would not capture any kick backs. As we build out integrations we would seek any such opportunities (for the benefit of INDEXCoop)

2 Likes

Decision Gate 2: Yield Hunter Index ($YHI)

Failed with 244.54k INDEX (68.42% voting AGAINST) :x:

Link to the result: Snapshot

2 Likes