Improvement Proposal: Fee Vol.4 for f(x)

[mr.masa]

Summary
To apply new fee model for f(x) Protocol to scale in the market bear cycle.

Motivation

The background of this new fee model is this: To scale f(x) protocol, we have to design fee table which motivate investors to mint xETH in the bear cycle of Web3 market.

It’s because Web3 market bear cycle will be a key growth phase for f(x) as same as UST in LUNA.

Based on the current fee structure, fETH will go into “disabled” with Stability Mode when ETH goes into the long-term bear cycle, which seriously limits the scalability of f(x).

To understand the above more clearly, we can learn important things from FRB operation model which has over 100 years history. Learning from FRB is “Standing on the shoulders of Giants” for us.

They mainly take two operations: normal mode and emergency mode.

Picture1:

From the above, the stability Mode works as same as emergency mode on FRB. FRB will not often use this mode because they deeply understand this operation model brings lots of market turmoil based on their experience. The latest case they used this model is during Corona Shock. They printed lots of USD by buying US Treasury.

In our stability mode on f(x), we use fETH on Rebalance Pool to sell fETH and buy xETH to raise CR higher than 130.6 again. I feel very uncomfortable with this operation. This is exactly as same as emergency operation model on FRB. I also expect huge conflicts will raise from fETH holders who uses Rebalance Pool. Who can guarantee their fETH on rebalance pool will move back to their hand? We must always and carefully prepare for unexpected things would happen in Web3 market which I saw many times as Web3 OG since 2013. Thus, We must minimize the usage of this operational mode.

In order to minimize the usage of Stability Mode, we should establish more flexible fee operation model as I proposed on Vol.4 fee model. This is very much as same as FRB does on FFR control like 0.0% to 5.5%, etc on their normal operation mode. The current f(x) fee table on normal mode has only single fee table, which is completely lack of flexibility.

Thus, again, how we incentivize long-term xETH holders is a key element to scale f(x) .

Proposal

To solve the above potential issues, I propose two new fee tables:
One: “Medium Mode” between Normal and Stability Mode. CR sets at <140.6 >130.6.

Two: “Bull Mode”. CR sets at >185.1.

Picture2:

The major purpose to set the above mode is to incentivize xETH holders to mint and hold during Web3 market bear cycle.

This is because to consider xETH purely as a trading product will misguide us to achieve its product market fit of f(x).

I do think considering “two year bull cycle and two year bear cycle with following every 4 year BTC halving schedule”, we should attract ETH long-term holders to invest in xETH as the long-term investment vehicle.

Because if f(x) functions accurately, those investors who hold ETH in long-term can gain bigger return than directly buying and selling ETH with the above every 4 year cycle. If we can successfully attract these long-term investors to xETH market with the accurate product understanding, f(x) can stably scale and accelerate the mass adoption.

One of the key component on “Medium Mode” is xETH Incentive Pool. The ideal APR target will be around higher than 2%. We don’t need this pool APR on Stability Mode because we make fETH disable to mint, and use fETH rebalance pool with selling fETH and buy xETH. To avoid Stability Mode is a main mission for Medium Mode.

Then, the next key question is “How we achieve higher than 2% APR for xETH holders on Medium Mode?”

I am thinking of two solutions.

A. We lock up our 10% FXN DAO Treasury with the maximum term (4 years lock up), and allocate the ETH returns to xETH holders when f(x) goes to Medium Mode.

B. To allocate the above 10% FXN DAO Treasury to FXN yield farming vaults on Concentrator.

B is back up. I prefer A is the best because the risk level of A is lower than that of B. Thus, in order to achieve higher than 2% APR on A, we should think of supplementing our DAO Treasury.

In my understanding, this 10% DAO treasury is for Recapitalization Mode on f(x). But, I think this is not the effective solution to maintain f(x) because the recapitalization mode will cause the same negative spiral which happened between LUNA and UST.

The price crush of LUNA caused additional capital flight from UST which accelerated the de-peg of UST against USD.

Thus, to use this 10% for xETH Incentive Pool is more constructive solution to scale f(x).

Then, here is additional suggestion. I think our current FXN allocation plan on Liquidity Incentive with 49% is too large.

Picture3

We can set this incentives with 30% to 39%, and move the rest of FXN to DAO treasury in order to raise the APR for xETH Incentive Pool.

The key thing we have to understand here is the market value of FXN goes up with the growth of TVL on f(x).

Based on this, to allocate the ETH return of veFXN on the DAO Treasury to xETH Incentive Pool will strengthen the network effect of f(x) Protocol.

Basically, sooner is better to start xETH Incentive Pool backed by FXN DAO Treasury because we can accumulate the ETH return with veFXN when ETH in bull cycle and we can use these returns xETH Incentive Pool when ETH is in bear cycle.

Currently, I suppose BTC 4th halving bull cycle will end in Q4 2025 to Q1 2026.

Here is also the summary of other fee adjustment.

When the CR > 185.1%, f(x) disables to mint xETH in order to secure higher return for xETH holders than directly investing in ETH in the market bull cycle. 1.5x is too weak. Higher than 2.0x is the best.

On the Bull Mode, fETH Mint Fee should set as 0.0% to increase fETH market supply while xETH Redeem Fee should set as 0.0% to decrease xETH market supply.

On Stability Mode, we eliminate xETH Mint Incentive because we should use this inventive for xETH Incentive Pool on Medium Mode.

On Bull mode, we set fETH Redeem Fee at 1.0% in order to avoid the decrease of fETH supply in the market bull cycle.

On Medium mode, we set xETH Redeem Fee at 8.0% in order to avoid the decrease of xETH supply in the market bear cycle.

One last thing, we can also apply dynamic adjusted fee table models for both fEXH minting / redeeming and xETH minting / redeeming in CR > 130.6 to < 185%. If this model does not require too much technical work and gas fee consumption for the smart contract management, we should apply this model. This automatic fee model will enable f(x) to scale faster with following the principles of market economy.

Polling Period
As I mentioned on the above, sooner is better to apply this new fee model. The polling process begins now and will end at 23:59 UTC on 12/31/24.

If Quorum is reached, a Snapshot Vote will be put up at 00:00 UTC on 01/01/2024.

Poll
For - The above Fee No.4 is applied. The New DAO Treasury is applied, and the new FXN DAO Treasury is locked up to run xETH Incentive Pool.
Against - Nothing.

[kmets]

Hey Mr. Masa! Thank you so much for writing this incredibly thoughtful and detailed post about the safety of fx. I embrace any chance I get to discuss the details of all Aladdin products, so thanks for creating this opportunity!

I have some questions to learn more about your line of thinking and understand your point of view. The first batch are philosophical and the second more detailed to your particular fee structure:

Philosophical

To understand the above more clearly, we can learn important things from FRB operation model which has over 100 years history. Learning from FRB is “Standing on the shoulders of Giants” for us.*

Am I assuming correctly that “FRB” stands for Federal Reserve Board? If so, then I would graciously ask if you could please explain why you view them as a model to emulate in terms of monetary policy. The reason I ask is because f(x) was created in direct opposition to the USDC depeg event in March, and that event was caused by the explicit actions of the FRB enacting the monetary policies you describe in your matrix. I would like to understand your reasoning for emulating the FRB in terms of system dynamics and monetary policy for f(x).

In our stability mode on f(x), we use fETH on Rebalance Pool to sell fETH and buy xETH to raise CR higher than 130.6 again. I feel very uncomfortable with this operation. This is exactly as same as emergency operation model on FRB. I also expect huge conflicts will raise from fETH holders who uses Rebalance Pool. Who can guarantee their fETH on rebalance pool will move back to their hand? We must always and carefully prepare for unexpected things would happen in Web3 market which I saw many times as Web3 OG since 2013. Thus, We must minimize the usage of this operational mode.

A clarifying point of information: By the end of the year, f(x) will have two rebalance pools:
Rebalance pool A - The classic that has been in operation since the genesis of the protocol, where fETH can be redeemed back into stETH for rebalance pool stakers in times of stability mode.
Rebalance pool B - A brand new pool where fETH can be minted directly into xETH, whereby fETH is directly moving reserves from fETH to xETH, accomplishing a double win: decreasing the amount of fETH and increasing the amount of xETH in the system.

I am curious to know why you feel discomfort with these automated processes. Neither of them should require outside assistance, and both will be instantly automated and operational at the same time, along with all of the other minting/redeeming features of the stability mode processes. Stability mode should essentially be no more than a few seconds from an outside observer, and the entire f(x) system is designed to scale up and down along with the size of the Ethereum economy. Why do you feel that these systems alone are not resilient enough to ensure that the f(x) system scales at the same rate as the Ethereum economy?

Thus, again, how we incentivize long-term xETH holders is a key element to scale f(x).

Here we are completely aligned, but have different points of view. Let me explain:

One of the wonderful things about the fETH and xETH tokens is that they are both ERC-20 tokens that are composable with just about any other defi application. The core team (and also highly engaged community participants) have been hard at work building partnerships to integrate both tokens with a multitude of protocols in a variety of ways. One of the main questions we ask ourselves on the core team is “to what level must the protocol itself support these tokens”? This is an abstract question, and doesn’t have a definitive answer, but we do know that if these tokens do not have an incentive structure and support system outside of the protocol itself, f(x) will not be the paradigm shifting financial system we know it can be.

One example to think about, in terms of outside protocol integration, would be how xETH would be viewed in a lending market. Imagine if you could lend/borrow xETH. Now, in a bear market, folks would LOVE to borrow xETH! Borrow xETH, swap to stables, pay back your loan cheap? Hells yeah! Because of that, the demand to borrow xETH could be very high, and so the APR paid to xETH lenders would be very high as well. It is not hard to imagine that xETH holders would likely receive a healthy incentive for lending their xETH in a turbulent/bearish environment.

Additionally, xETH would be one of the best assets for options markets to integrate, as its volatility is so high that it would make for an excellent fee generating asset for the integrating protocol. I would imagine that xETH holders that sold call option contracts would be well incentivized during a bear market.

Incentivizing xETH holders outside of the protocol will be a great cost savings to f(x) and will actually be a positive event for the exogenous protocols/users that are incentivizing xETH holders. In my opinion, we should do everything we can to minimize costs for f(x) internally and do everything we can to maximize opportunities for integrating fETH and xETH outside the system. If you have a differing opinion, I would very much enjoy hearing it.

Fee structure

The major purpose to set the above mode is to incentivize xETH holders to mint and hold during Web3 market bear cycle.

Am I understanding correctly that it seems that your entire fee structure is centered around the protocol somehow being aware that it is in a bear/bull market? Do you happen to have metrics for determining this?

2. In medium mode where the protocol is functioning safely, why does this system have such a high xETH redemption fee? Am I correct in assuming that this would dissuade rational users from minting xETH in the first place? Bear markets have long, sideways accumulation regions, and this seems like it would hinder the adoption of xETH.

Because if f(x) functions accurately, those investors who hold ETH in long-term can gain bigger return than directly buying and selling ETH with the above every 4 year cycle. If we can successfully attract these long-term investors to xETH market with the accurate product understanding, f(x) can stably scale and accelerate the mass adoption.

Here we are supremely aligned! I 100% agree with this and could not put it better myself.

Then, the next key question is “How we achieve higher than 2% APR for xETH holders on Medium Mode?”

I am assuming here that you are referring to the new rebalance pool B, so I will address the point as such. Here I think is a bit of a pickle, as the only way we can aim for a APR for a rebalance pool is if we A) always adjust incentives to match the invested tokens (this is computationally expensive) or B)fix the size of the rebalance pool. Option A is obviously not ideal, and option B limits the size of a rebalance pool, which we would prefer not to do. Rebalance pool B can be seen as a double win for the protocol (lower fETH and raise xETH) and a HUGE win for rebalance pool B users: Not only do they get to DCA back into ETH, they get to do so with leverage and their fETH has much greater purchasing power than if they were in rebalance pool A. We want that pool to be as large as possible, and so controlling incentives for it will be tricky, as it will be up to the market ultimately to find out how desirable that pool will be, and what level of incentives will be necessary to attract fETH holders over RP A.

I want to stop here to give you a chance to respond before we go any further with this conversation. You’ve given me alot to think about, and the core team is always thinking about ways that f(x) can be improved and hardened against downside. I appreciate the great amount of effort you’ve put into caring about f(x) and the entire Aladdin ecosystem! Looking forward to continuing this discussion!

[mr.masa]

@kmets

Am I assuming correctly that “FRB” stands for Federal Reserve Board? If so, then I would graciously ask if you could please explain why you view them as a model to emulate in terms of monetary policy. The reason I ask is because f(x) was created in direct opposition to the USDC depeg event in March, and that event was caused by the explicit actions of the FRB enacting the monetary policies you describe in your matrix. I would like to understand your reasoning for emulating the FRB in terms of system dynamics and monetary policy for f(x).

Yes, correct. FRB is Federal Reserve Board or Federal Reserve Banking System.

To criticize FRB and to learn from FRB are different topics. As same as our DAO operation, too.

Well-known quate: “Fools learn from experience, wise people learn from history”

Learning from FRB equals to learning from the history of monetary policy since Roman Empire.

For example, in DAO operation, we don’t take parliamentary democracy because we know parliamentary democracy malfunctions these days. But, we apply some knowledge from parliamentary democracy on our DAO operation. Boule election in AladdinDAO is a good example. The process of Boule election has the same essence of parliamentary democracy. So does monetary supply control by FRB.

Thus, what we have to apply the knowledge of FRB for f(x) are two.

1.To set two operation mode: One is normal mode. The other one is emergency mode.

2.To make the normal mode more resilient against market economy behavior in order to minimize the usage of emergency mode and to achieve better product market fit.

Autonomic fee adjustment model gives us higher scalability for f(x). Thus, I stick to set multiple fee tables on normal mode in Vol.4 Fee Model.

[mr.masa]

@kmets

A clarifying point of information: By the end of the year, f(x) will have two rebalance pools:
Rebalance pool A - The classic that has been in operation since the genesis of the protocol, where fETH can be redeemed back into stETH for rebalance pool stakers in times of stability mode.
Rebalance pool B - A brand new pool where fETH can be minted directly into xETH, whereby fETH is directly moving reserves from fETH to xETH, accomplishing a double win: decreasing the amount of fETH and increasing the amount of xETH in the system.

I am curious to know why you feel discomfort with these automated processes. Neither of them should require outside assistance, and both will be instantly automated and operational at the same time, along with all of the other minting/redeeming features of the stability mode processes. Stability mode should essentially be no more than a few seconds from an outside observer, and the entire f(x) system is designed to scale up and down along with the size of the Ethereum economy. Why do you feel that these systems alone are not resilient enough to ensure that the f(x) system scales at the same rate as the Ethereum economy?

I don’t see any problem for A. Even more welcome that I have guarantee to redeem back into stETH. Much better than the current Rebalance Pool.

But, B has a problem. What type of users wants to use this pool?

Firstly, the over-rapped area between fETH investor and xETH investor are very small. This is also our initial perception for how we design this DeFi protocol.

fETH holders prefer to gain stable return with lower volatility. If I am fETH holders, I only prefer to use Pool A simply because I can safely earn higher stETH rewards than that of LIDO and my assets are also secured with redeeming to stETH in Stability Mode. fETH investors consider stETH is lower risk than fETH because f(x) is still very early stage. stETH already have over $15B market cap run by run well-experienced Lido DAO, and huge stETH/ETH liquidity on Curve. Most users still consider fETH as higher risk asset than stETH. Only innovator and early adopters baed on the chasm theory want to use fETH as of now. And, even those investor wan’t buy xETH because they don’t like volatility. Thus, they use fETH. We have to accept this reality.

On the other hand, if I am xETH investor, I prefer to directly buy xETH, and never use Pool B because to use Pool B is totally uncontrollable for me to find the right timing to buy xETH. Most of xETH investors look for the bottom in ETH market to minimize their unrealized loss. To follow both ETH market move and CR ratio on f(x) in order to use Pool B accurately are too much work for xETH investors. Very complicated.

Again, we must simplify our product to accelerate mass adoption of f(x). Too many functions kills our potential to scale. To me, Rebalance Pool A & B model are too complicated to achieve product market fit.

[mr.masa]

@kmets

One example to think about, in terms of outside protocol integration, would be how xETH would be viewed in a lending market. Imagine if you could lend/borrow xETH. Now, in a bear market, folks would LOVE to borrow xETH! Borrow xETH, swap to stables, pay back your loan cheap? Hells yeah! Because of that, the demand to borrow xETH could be very high, and so the APR paid to xETH lenders would be very high as well. It is not hard to imagine that xETH holders would likely receive a healthy incentive for lending their xETH in a turbulent/bearish environment.

I’m afraid of saying this. but, I totally disagreed to develop such complicated derivative products for f(x). Very much like developing “Subprime Market” in Web3.

Bitcoin was invented right after Lehman Shock in 2008. Lehman Shock was caused by the crush of Subprime Market. The above idea is very much against the philosophy of Satoshi Nakamoto.

We must focus on retail investors. They are not financial engineering specialists like core members of AladdinDAO. This is my biggest concern for our current core members team formation. All members have strong financial engineering background. This is great to invent innovative DeFi protocols with the expertise. But also, this team formation has lack of capability to understand the deep meaning of mass adoption. Most of retail investors have zero knowledge for financial engineering.

The core team members only pay attention to work with DeFi professionals for their product development, and less pay attention to the user behavior of retail investors.

This product development style will kill the potential of the innovative protocol like f(x). f(x) is still a baby. We have to grow this protocol in right direction. The right direction means mass adoption. Every single Web3 users use fETH like USD. This is we want to achieve.

[mr.masa]

@kmets

Incentivizing xETH holders outside of the protocol will be a great cost savings to f(x) and will actually be a positive event for the exogenous protocols/users that are incentivizing xETH holders. In my opinion, we should do everything we can to minimize costs for f(x) internally and do everything we can to maximize opportunities for integrating fETH and xETH outside the system. If you have a differing opinion, I would very much enjoy hearing it.

I’m afraid of saying this. but, I totally disagreed to develop such complicated derivative products for f(x). Very much like developing “Subprime Market” in Web3.

Bitcoin was invented right after Lehman Shock in 2008. Lehman Shock was caused by the crush of Subprime Market. The above idea is very much against the philosophy of Satoshi Nakamoto.

We must focus on retail investors. They are not financial engineering specialists like core members of AladdinDAO. This is my biggest concern for our current core members team formation. All members have strong financial engineering background. This is great to invent innovative DeFi protocols with the expertise. But also, this team formation has lack of capability to understand the deep meaning of mass adoption. Most of retail investors have zero knowledge for financial engineering.

The core team members only pay attention to work with DeFi professionals for their product development, and less pay attention to the user behavior of retail investors.

This product development style will kill the potential of the innovative protocol like f(x). f(x) is still a baby. We have to grow this protocol in right direction. The right direction means mass adoption. Every single Web3 users use fETH like USD. This is we want to achieve.

Incentivizing xETH holders outside of the protocol will be a great cost savings to f(x) and will actually be a positive event for the exogenous protocols/users that are incentivizing xETH holders. In my opinion, we should do everything we can to minimize costs for f(x) internally and do everything we can to maximize opportunities for integrating fETH and xETH outside the system. If you have a differing opinion, I would very much enjoy hearing it.

This is not the right way to accelerate mass adoption while keeping our product competency. All mass adopted products in our world starting from Google Search and iPhone always control their product’s intrinsic value and risk vertically. The above idea is similar to Google outsourcing their search engine algorithm development and maintenance or Apple outsourcing iOS development and maintenance.

If we outsource the incentive mechanism of xETH holders, we lose our control the product’s intrinsic value and risk. We must control these elements by our selves. This is a must item.

fETH with Rebalance Pool A is perfectly designed product. I love it. The current desing of fETH with Pool A can cross the chasm. I am very confident on it. But, f(x) has another side: xETH. As we know, the relation between fETH and xETH is two side of the coin. And xETH is still too weak for mass adoption. xETH needs its own incentive mechanism to accumulate xETH holders in ETH long-term bear cycle.

[mr.masa]

@kmets

In medium mode where the protocol is functioning safely, why does this system have such a high xETH redemption fee? Am I correct in assuming that this would dissuade rational users from minting xETH in the first place? Bear markets have long, sideways accumulation regions, and this seems like it would hinder the adoption of xETH.

Simply because we want to accumulate xETH holders. Thus, we need to demotivate xETH holders to redeem with this high redemption fee, which enables us to minimize the usage of Stability Mode. This is the basic philosophy behind medium mode.