FLI suite streaming fee offer - Request for comment from the community

Authors: @afromac @BigSky7 @oneski22 @allan.g @Mringz @verto0912

Request for comment from Index Coop Community

This post is intended to communicate to the community an offer for a new fee split for any future FLI products, and the reasoning for this suggested fee split. Of primary importance, we must remember that the FLI product line is the result of a successful partnership between Index Coop and DeFi Pulse.

We believe that it is in the best interest of all parties to work together. Index Coop seeks to enhance our partnership with DeFi Pulse and to continue working with DeFi Pulse on launching innovative DeFi products. DeFi Pulse are significant and important community members, and the long term strength of that partnership is without a doubt the primary objective of these discussions. However it is impossible to enter into a successful partnership - in good will - when the balance of the relationship is heavily tilted in the favour of one party. Also of importance, Index Coop was not part of the agreements that led to the creation of the FLI products.

We recognize that the path forward with our partners is based on mutual trust, and that we should have a say on the agreements that we are expected to keep. We open this discussion with the community in order to ask for DeFi Pulse’s trust and understanding, so that we can aim to redress that imbalance, and to negotiate a viable long-term agreement from first principles.

Introduction

Following on from the discussion - Request for Discussion FLI-Fee Split , the next step in this negotiation is to make an open offer to DeFi Pulse for what we consider to be a fair fee split and seek the input from the community. What has been deemed a “fair” suggestion is based on evidence which will be provided below. The current fee split is 60:40 in favour of Index Coop. This split is based on revenue and not on profit. Because Index Coop is responsible for the maintenance and gas costs of the products, the current arrangement results in an extremely inequitable outcome.

As of the end of July, the ETH2x-FLI product had generated over $400,000 in revenue. This resulted in DFP receiving $163,657 of income. Conversely, after accounting for operational expense and gas cost to transact, the Index Coop has made $5,474 in profit . During the same time-frame, the BTC2x-FLI product generated a total net loss of $54,644, while DFP received $20,838 of income.

DFP is making more from a loss-making product, than the Index Coop is making from one generating close to half a million in revenue. This is a completely unsustainable arrangement. To address this major imbalance, we suggest a new fee split whereby Defi Pulse will receive 15% of revenue from all the future FLI products.

Why a 15% fee split is appropriate

This suggested fee split has been put forward for two reasons.

  1. This kind of licensing agreement, whereby Index Coop enjoys the right to operate a product with Defi Pulse’s methodology, and pays a reasonable royalty rate on income in order to do so, more closely reflects normal practice in the software industry. DFP’s methodology for the FLI products is a unique innovation in the space. It adds value to our products. However, we believe the majority of the value add that it takes to make our products a success comes from within the Index Coop itself.
  2. This change in the fee split is suggested to more fairly reflect the operational expense required by the products to date, as well as the expense that will be required to maintain the existing products in future and to launch further products in this range.

We see expenses falling under three categories:

  1. Gas Costs
  2. Maintenance Costs
  3. Risk

Gas Costs

Average daily gas costs for the products have reduced, largely due to changing market conditions but also due to upgrades and improvements made by Index Coop and Set - such as the trade splitter and trade router upgrades. However, in recent weeks, gas costs have remained constant, and no further reduction in gas expense will be easily achieved. We are now utilizing a near optimal split between the three major liquidity pools on Sushiswap and Uniswap V2 & V3, and further engineering efforts to improve this would likely yield diminishing returns.

What does this mean? We can only reduce gas costs relative to revenue by growing the products AUM. To date, Index Coop has engaged in meaningful work and allocated resources to achieve this, but what DeFi Pulse has contributed in that regard is not clear.

Also, In this environment, the products are still vulnerable to elevated gas costs and large spikes in volatility. The vast majority of expenditure on gas is not during routine rebalancing, but during periods of market volatility, where it can be in excess of 1,000% of daily revenue. In the case of BTC2x-FLI, this is particularly relevant as the product has essentially been operating at a loss since shortly after the release. It will likely begin to operate at a profit soon, but any heightened volatility could quickly return it to a loss. Our current thesis is that future FLI products will operate with a level of success that falls somewhere within the range between ETH2x-FLI and BTC2x-FLI. However, it is equally valid that BTC2x-FLI will be our second most successful product (being a derivative of the biggest asset in the market), and that future products will produce a marginal revenue, and costs will have to be managed aggressively in order for them to be profitable. We need to construct a licensing agreement that reflects that possibility.

Crucially, it is essential to address the suggestion that gas costs will not be an issue of concern in the future. This is patently untrue, and cannot be taken into consideration. The assumption that gas costs will decline significantly is based on multiple hypothetical scenarios, which don’t reflect the actual environment we find ourselves in.

Primarily:

  • A large portion of the AUM of the products is likely to remain on L1 and not on L2. Yes, we have the capacity to launch FLI products on L2 and gain the advantages of doing so. However, as of today, L2 introduces additional centralization, and changes the risk profile of the product. For this reason, the FLI products will still have a large allocation to L1.

  • Optimization on L1 is possible, but is entirely contingent on the network first being upgraded to PoS, and then upgraded again to enable sharding. These are not trivial upgrades, and may take years to occur and for the results to enable cheaper transactions. If demand for the Ethereum network continues to scale, it is perfectly reasonable to assume that future upgrades to enable more bandwidth may lag increased demand. Conversely, it is unreasonable to suggest that the (at best) nebulous roadmap for the ethereum network determines that gas will be a trivial expense within any kind of reasonable time-frame.

  • Cryptocurrency is an incredibly volatile asset class. The majority of our gas costs are incurred to protect our product during periods of heightened volatility. This will be a feature of both the market and of our products for the foreseeable future.

Maintenance Costs

The FLI products have proven to be extremely complicated products to manage post-launch. This has been a point of contention within the Coop on many occasions, as these products have been allocated resources to the extent that it has obstructed other projects and initiatives from proceeding. Overcoming these challenges and equipping ourselves to grow other aspects of our business has been an enormous hurdle to overcome, and took concentrated effort by many teams over several months. This included massive infrastructural upgrades to the underlying tech by Set, the development of new UIs and interfaces in order to scale safety parameter management, an innovative rebalancing algorithm to enable optimized transactions on multiple DEXs, and the hiring of a full-time team of in-house members to manage risk, general operations and communications on the product, of which I am a member. More info on the extent of our work is included in the original fee split discussion post.

It is difficult to quantify some of the above costs, but costs relating to the FLI team are straightforward to produce. To date, we have been paid $52,850. The PWG has budgeted $20K-$35K p/m for the next quarter. Annually, we can expect to spend $240K-$420K on this team. Preemptively, I want to state that as our capabilities stand today, it would be extremely challenging to grow the product range to its potential. It would either require the hiring of a full time member to monitor and report on risk, or the automation of much of our risk management, and communication around that. We have selected to go with the second route, and the team is currently developing an update that will automate risk management. This encompasses design, project management and engineering all happening exclusively within our team.

Longer term, we anticipate that competitors will enter the market, and ensuring that the products continue to be successful will be contingent on Index Coop to optimize and improve upon what we have today. This will involve the future creation of new technology, and the hiring of technical experts to fill emerging roles. For this reason, we strongly suggest that it is inaccurate to assume the IC’s costs relating to R&D and maintenance of the products are fixed and can be amortized across future products. Instead, these are real costs that will continue to grow as we scale the product range and the market matures.

Risk

In the event of a major product failure, DeFi Pulse will not assume responsibility. DeFi Pulse is a data company. Index Coop manages the product and assumes all responsibility for any negative outcomes. There are multiple risks associated with managing these kinds of products, and Index Coop shoulders the full burden of responsibility for navigating them. As such it is unreasonable to suggest that Defi Pulse can expect 40% of revenue, while having zero exposure to the underlying risk.

These risks include:

  • Dependencies such as Compound or USDC could fail.

  • Regulatory Risk - Circle or Coinbase could be forced to shutdown USDC

  • Liquidation of the product due to an oracle failure - MakerDAO 2020

  • Smart Contract exploit

The Leverage Indices Pod is ready to launch products

The leveraged indices pod has spent the last 4 months successfully managing both FLI products, and in that time has developed the skill and knowledge to be able to launch our own range of simple leveraged index products that do not use the FLI methodology. The team outlined their vision in this post.

Next Steps

The purpose of this post is to communicate to all of our stakeholders, and to our partners in DFP, what we believe is a fair outcome of these discussions. The intention is to present this evidence and reasoning to all of the wider community to invite further debate and discussion. If we need a community call to discuss this and align as a community we will schedule one.

This issue is central to the wider debate about autonomy - and the nature of the methodologist role - that is already taking place in the community. We appreciate the sensitivity of these discussions, and welcome the concerns of any community members. It is our responsibility to address those concerns. We promise to uphold our central principles of empathy and understanding in this conversation. All voices will be heard. To reiterate our primary objective from the start of the post, we believe that this is a crucial step in restoring equilibrium in our partnerships, and to secure and strengthen our relationships with Defi Pulse going forward.

17 Likes

Hi @verto0912

If the gas costs for maintaining the product were built into the product itself, a design feature requested by the methodologist during the pre-launch build, the $147K rebalancing gas cost would have not been incurred by Index Coop.

Furthermore, if this gas cost was socialised across ETH2x-FLI holders it is around 18 cents per unit of circulating supply. The gas costs like the maintenance costs can be recovered from the COMP rewards residing in the vault. Again, this is due to another design feature not being built out.

Revenue was $690K on ETH2x-FLI at the end of July.
Methodologist received $271K during this same period.
Income before operating overheads is $420K.
Gas Costs are $147K to date. ie: 35% of Income

Please advise, what would the new fee split be with the gas cost being incurred by FLI holders ?
(Eliminate over 35% of the costs, so 35%*(40%-15%) = 8.75%)

Would this mean the new fee split starts the discussion at 76.25/23.75 or rounding say 75/25 just by implementing a design feature requested prior to launch…

4 Likes

this is lovely :heart: when we approach problems like this, we are sure to win together.

8 Likes

Jumping in here only because a specific claim about the product process was made that the broader community may not have context for.

I’d be curious to learn where you got this information from. Please point me towards any discord message or forum post where this happened.

From the product and engineering perspective, this was never brought up as a requirement during the 3 months of R&D, auditing, testing, and implementation.

i suggested this on a call that i believe you were on (but maybe you were not?). this bounty for public function call is a reasonable flow and something i have previously designed for a different project that is currently in production. incentivizing keeper behavior flow is not some crazy bleeding edge design space. it was immediately dismissed by set labs. the plan at that time was for set labs to cover the gas from rebalancing. i do not know when, but at some time it seems this burden was shifted to the index coop?

furthermore, we initially offered to cover these costs completely for a higher rev split going our way. another suggestion that was immediately shot down by set.

i think a major problem in the current approach is that rather than engaging about to how to improve, set labs and some set labs contractors being tasked with index coop contributor work immediately move to a zero sum mindset. the gas costs problem is likely quite easy to ameliorate with a bit of design work and COLLABORATION.

but to get to real solutions, you have to stop thinking these are zero sum games. bc when you are playing zero sum games, and not respecting agreements you have made, you lose the trust needed to find real wins.

the current approach is just not the way.

5 Likes

@scott_lew_is I am curious how you think a reasonable flow for a keeper system would work.

We have considered many options for an incentive system (hence we have a ripcord feature), but there are 2 problems with rebalancing for FLI products vs. traditional products (e.g. yield farming):

  1. Incentivizing normal daily rebalances is typically not an issue, as gas prices are low and they are not time-sensitive. In fact, the cost of daily rebalances is typically $100 per day per product and can easily be outsourced to a keeper system. However, Leverage Indices are all about safety and most of the cost occurs during times of high volatility where gas prices can spike to 1000 gWei+ and numerous rebalances need to happen in succession. Thus, you need incredible amounts of liveliness during these periods of time and normal keeper systems like Chainlink / Keeper network (based on fast gas oracles) don’t work too well.
  2. When it comes to these edge case scenarios where delevering is critical, it is often the case that we are willing to pay significant gas costs to prevent liquidation from happening to maintain brand and the products’ safety. Thus, we leave the attack vector that keepers can submit absurd gas prices during rebalances and drain the leverage index holders’ funds.

We are thinking about these issues and if you have a good design - we are game as we are VERY pro using keeper networks and doing gas splits totally on-chain.

4 Likes

So, as we outlined above, gas costs are one of the three expenses / reasons for the proposed fee split. If we take out gas costs as an issue, there’s still maintenance costs and risk. In comparison with DPI, where there’s ongoing maintenance of the metholdology by DFP, the FLI suite of products is entirely maintained by the Coop. I, personally, see the FLI suite relationship as us licensing the formula from DFP.

If we look at your equation here, it’s based on the assumption that gas costs are the only reason for the proposed fee split. That’s not the case, they are 1/3rd of the reason.

At the end of the day though, it is up to Index Coop, as a community, to decide what the appropriate fee split is for licensing the FLI formula. My feeling is that even if gas costs are not a consideration, 75/25 is still not equitable.

3 Likes

The ratio used above was to say gas costs represents 35% of historical cost, which I think is based on an elevated cost base. If we pivot from a 60/40 split to a 85/15 spit due to three factors then the 25% swing is attributable to three factors. This is the logic used to justify the amended proposed fee split.

If 35% of the cost can be eliminated, then at 35% x 25 (40-15) = 8.75% is a reasonable adjustment. If we think about this a little deeper, prior to launch an assumption for gas cost would have been made and it probably wasn’t zero. And this would have been built into the 60/40 split. Unless gas cost was an oversight… Then the one of three points for changing the fee split is the difference between expected gas costs and actual gas costs. The same budgeted versus actual cost and risk question can be extended to the other 2 of 3 factors driving the choice to change the fee split.

If 147K is materially different from what was budgeted for pre-launch, then it would be great to know what was anticipated in $ terms. Then we can say what the difference is between pre-launch assumptions and what has transpired since. Do you have any supporting data from the 3 months of R & D cited in an earlier comment.

Can we propose a new fee split on the assumption gas costs are built into the product ?

2 Likes