[Proposal] Dynamic BBD Adjustment - Capping APR and Accumulating Overhead

Proposals
1

Summary

This proposal introduces a refinement to the Buyback & Distribute (BBD) mechanism to make it more sustainable and market-adaptive.
The change would cap staking rewards at 10% APR, leaving any excess GMX on Fee Handler contract as Overhead.
Accumulated Overhead can later be released when protocol fees are low or token price is high — making GMX distribution counter-cyclical and self-balancing over time.

Motivation

The BBD program has now operated for a full year, with approximately 1,625,000 GMX tokens bought back from protocol fees (amounting ~ $28M).
However, data shows that:

  • The total staked GMX has barely increased,

  • The free float remains unchanged, and

  • Most distributed GMX is sold shortly after receipt.

As a result, BBD has not achieved its intended goal of supporting or stabilizing the GMX price. Currently, the mechanism behaves like an ongoing emission source — creating sell pressure instead of building long-term buy support.

Problem Overview

  • Emission pressure is continuous, regardless of market conditions, fee environment or token price.

  • GMX distributed as rewards tends to re-enter the market shortly after distribution.

  • No feedback loop exists to throttle emissions in strong protocol performance periods or boost yield support when activity slows.

  • Funding fee farmers exploit high-APR periods by temporarily buying and staking GMX while simultaneously shorting to collect funding fees. After high-APR phases end, they unstake and sell both the principal and accumulated rewards, effectively extracting value from long-term stakers and adding sell pressure during market normalization.

Proposal Specification

1. Introduce a 10% APR cap for GMX staking rewards.

  • When APR exceeds 10%, reward emissions are limited and excess GMX is accumulated as Overhead.

  • When APR falls below 10% (fee revenue declines), the Fee Handler should gradually release stored GMX to maintain a smoother yield curve.

2. Accumulate Overhead inside the Fee Handler contract.

  • Overhead balance represents undistributed buyback GMX held by the protocol.

  • These tokens remain out of circulation until distributed later.

3. Keep the buyback process unchanged.

  • Only the distribution logic would be updated.

  • Buybacks continue to convert protocol fees into GMX, preserving existing economic flow.

Expected Outcomes

  • :counterclockwise_arrows_button: Counter-cyclical stability — the system accumulates GMX when token price is low or fees high, and releases it when price rises or fees slow.

  • :coin: Reduced immediate sell pressure — more GMX remains locked inside Fee Handler during high-APR phases.

  • :bar_chart: Smoother yields — staking rewards become more predictable week-to-week.

  • :puzzle_piece: Self-balancing tokenomics — aligning emission pressure with real protocol performance.

Empirical Data

Below is the dataset showing the real buyback history and the modeled outcome under the proposed 10 % APR cap.

image
image1063×1692 53.4 KB

Benefits

  • Market stability: Converts BBD from constant emission into an adaptive buffer.

  • Sustainability: Prevents excessive short-term emissions while maintaining healthy yield levels.

  • Flexibility: Overhead buffer can be used strategically — released manually or via automated conditions for example random airdrops to stakers.

  • Simplicity: Minor parameters adjustment without protocol redesign, contract changes, audits or 3rd party integrations breaking.

Drawbacks / Considerations

  • Slight reward delay for stakers during high-fee periods (shouldn’t be problem for long term holders).

  • Transparency of accumulated Overhead should be ensured via public view function or dashboard integration.

  • Main concern for stakers: accumulated GMX may never be fully distributed. Therefore, a strict disclaimer must apply:

    All accumulated GMX (Overhead) may only be used for future staking rewards (in GMX or esGMX form) and cannot be redirected, repurposed, or burned for any other use.

Governance Disclaimers

  • The modified BBD mechanism should operate for at least one full year to realize it’s full potential before any new tokenomics or distribution changes are introduced.

  • During this period, no overlapping proposals affecting emissions or staking rewards should be enacted.

  • After one year, data should be reviewed to evaluate success and determine next steps.

Next Steps

  1. Gather community feedback on proposal and APR cap level (default 10 %, adjustable).

  2. Submit final on-chain governance vote for proposal implementation.

1 Like
2

Disclaimer:

This proposal was drafted with the assistance of ChatGPT to help with structure and formatting. While care has been taken to ensure accuracy, minor wording inconsistencies or technical imprecisions may remain.

All opinions, interpretations, and conclusions expressed in this proposal represent my personal views and individual suggestions for the GMX community to consider.

I will use this post to make further non-formal personal comments on the proposal.

3

Personally dont like the idea of a cap would rather do 20% of buyback to stakers, and the 7% buy and burn it, to counter the effects of the selling overtime instead. (Effects are more instant and u dont delay stakers rewards else funds might be stuck in contract for long period that becomes effectively a burn this idea wont pass with most)

Easier to stomach for most stakers, uncertainty of when yield and overhang of GMX staying in the contract is just not sellable to most voters.

Do consider if this is a better alternative, would vote for a split 20% fees 7% burn than a delay which i feel wouldnt solve the issue.

4

for ref i think a similar proposal was given 2 years back, where they tried to cap it at 20% apr, but didnt pass / didnt push through, personally dont think is the way, but would suggest an alternative, additionally given today’s yield we aint hitting that much for a 2% apr to make a difference…

5

You misunderstood the capping. Still same 27% of fees without any cap will flow to GMX stakers, just distribution will be capped and smoothed over longer time period.

6

I understood it, i mean that given today APR its at 12% even if we cap it at 10%, it be 2% for future dispersion, previously there it a 20% apr cap proposal…

What i meant by 20% & 7% difference was an alternative, where currently all V2 Fees 27% is given to gmx buyback, 20% of it to the staking, 7% used to burn instead. Does not involve a cap, fees are still given but we have instant effects from that 7%.

Given that, on volatile times, where fees are high, then the burn effects are higher too, we dont have to rely on more GMX staked, but actl reducing supply. This scenario is better in a volatile scene, and in today prices, using 7% of fee’s as burn would reduce supply by a good %.

2 Likes
7

Here are some additional random thoughts in support of this proposal (will be more):

If fees remain at the same level as last year, there is no way the GMX price can survive the current downward trend. With the fees of $28M to stakers and the current GMX price at $9.1, ~3M tokens could be bought back in a year. Just ~800k will be distributed with a 10% cap, so 2.2M tokens will be locked for later distributions and taken out of circulation. And it is the amount of all current free float.

My prediction is that if the proposal were to be implemented, the GMX token price would diverge towards a 10% APR target in the long run and it’s price become a direct function of protocol fees. Taking $1M in weekly fees as a base, it translates to $1M * 27% / 15k GMX, which is equivalent to a GMX token price of ~ $18. So, after implementing a fixed APR of 10%, we would have a soft lock on the GMX token pricing at $18 per $1m in fees.

8

Support! Trying to picture what GMX would look like right now with 825k of the bought back tokens not being emitted throughout the year. Could maybe tweak APR capping % to be slightly higher, but overall I like this proposal.

9

If we set the APR too high, we will end up with the current status quo. The GMX price is so low at the moment that we should be as aggressive as possible with the proposed BB&A model. We also need to take into account that we had very high fees for a couple of weeks last year, which will not be repeated. To be conservative, we should brace ourselves for lower fees going forward. Therefore, a 10% APR cap strikes a balance between substantial accumulation and a solid yield for serious long-term investors.

To illustrate, hypothetical data for the past year shows that different APR caps yield the following results:

Targer APR: 20% 15% 10% 5%
Distributed: 1 476 847 1 183 756 795 573 400 618
Accumulated: 144 653 437 744 825 927 1 220 882
10

Saulius your post gave me the same “sinking in the pit of my stomach” feeling that I felt the first time I saw Q’s proposal to end our points program. These kinds of ideas only work in a vacuum. We’d have new sell pressure from stakers that will sell their $GMX in favor of better APR yields.

Like if this is legitimately a concern of yours, incentivizing stakers to look for “better than 10% in the short term” opportunities would exacerbate the token’s price decline. All the more so with the uncertainty of ever actually receiving the “overhead” reserved GMX. Essentially the protocol would be taking staker’s property and refusing to give it to them until such time as the protocol sees fit. Again, maybe this would work if GMX was the only game in town. But it would further the catastrophic price decline we’ve witnessed over the past two years. Looking at the Empirical Data spreadsheet you posted, it’s pretty clear that GMX’s only saving grace at this point is the routinely 15% to 20% APR (or higher) it yields. Given that the token itself has failed as a “store of wealth” since its value keeps dropping, the decent APR’s are the only thing that are keeping the token on life support.

This will send long term stakers packing when they can park whatever value their GMX tokens remain as into dozens of competing products, or even competing products on the protocol, like GLV pools. The token can’t survive cascading sell pressure from that many yield-seeking holders all selling. As best I recall, every DAO vote that messes with the tokenomics has been followed by another decline in price.

I feel like we’re treating symptoms rather than the disease. You highlighted the problems:

  • distributed GMX get’s sold right after its distributed,
  • and the amount staked has barely budged in years.
  • Plus a third I’d like to point out - receiving yield in GMX token rather than ETH has cost stakers that would have held one or the other MASSIVELY, as ETH still performs well enough while GMX reliably loses 50% time and again.

It’s significantly less risk to the project to reintroduce a points program (we did great when we had one) and make the real-yield distributed as ETH, and then see how those affect the token price. I believe several people in the telegram channel and this governance message board have proposed various iterations of those. Let’s treat the disease and not the symptoms, and get non-bots (actual humans!) incentivized to buy $GMX tokens again.

If this proposal were to pass, I honestly believe it will result in the token price going to below TGE value (so $2 to $3) even if floor funding is enacted. At the very least, wouldn’t you accept a compromise of 20% cap? Many of the top level stakers would still regard it as theft, but they might not insta-liquidate their holdings over it, which I know with certainty they would under this current proposal.

My goodness what happened to your project, @xdev_10? You had so many great ideas, and they worked! But other people started trying to fix things that weren’t broken, toying with the tokenomics, to the point where:

  • The 4th biggest holder is trying to enact auto-compounding explicitly so he doesn’t have to see the price
  • The guy who took away all of the amazing incentives you put forth (MP’s, real yield) was able to vote himself into receiving a $100k a month salary
  • The one person in this community that I always saw as wise & levelheaded is proposing a 10% cap on APR with a “maybe” attached to the distribution of anything above. Saulius, you were the one who lead the fight to keep the cap at 200%!

There are ways to get out of this hole we’ve found ourselves in. But this proposal isn’t it. There’s too many other opportunities out there for buyers to waste their time on a project that keeps removing its incentives to actually stake the token. The GMX token was doing so well for years under the original tokenomics plan that it launched with. Why can’t we just go back to that?

11

If this is the case, shouldn’t we be making attempts to increase the fees & amount staked? Again, let’s treat the disease, not the symptoms.

And Saulius, oftentimes you’re one of the only voices of reason in this space. You’re time is so much better spent trying to come up with solutions to stop the bleeding, rather than simply slowing it down.

The fact that you’d even put this idea out there before sponsoring something like initiating a points incentive plan or changing the yield to a token that people want is really disappointing.

Please don’t favor simply slowing the bleed. Let’s at least try to cure the disease first.

12

How can you call 10% APR theft then in times GMX was prospering with $50+ token price and MP soft-locking, APR constantly stayed near 5% or lower? Do we want monstrous APR with declining token price or lower (still solid) APR with stable or rising token price?

And don’t forget this exec capped yield is not stolen but just spread (shifted) over time. For long time holders ir doesn’t change anything. I’m speaking as holder from day1 myself still claiming top holders list.

13

How does this relate to the disclaimer you posted about the accumulated overhead perhaps never being distributed?

Whoah whoah whoah - please don’t misunderstand me. It’s not me thinking it - I’m a shrimp compared to the whales. What I’m saying is that the high quantity stakers would see it that way, that the APR methodology they’d agreed to and that keeps them here in the ecosystem is getting rugpulled. By altering the deal, they’ll seek out greener pastures.

Monstrous APR will be what we get since the price of the token will keep dropping. But that’s the rub - that APR is the only thing keeping us from a cascading liquidation crisis as stakers see their wealth continue to evaporate. Is there any number other than 10% you’d be open to? Why so firm?

And (unfortunately but practically), have you talked to Q about this? Ultimately he’ll be the one making the decision. If he pushes for 20% would you be open to the 15% to 17% range? Or on a different note, why not institute this gradually so there’s less shock to the ecosystem. Start with a 20% cap and next month make it 19% cap, next month 18% cap and see where the happy balance is.

Freaking a, our cap used to be like 200%…

1 Like
14

it is disappointing how things turned out, it is not 100% certain whether things would have been better had nothing been changed, when the original MP capping proposal was created, token price was already not doing well, of course it is much worse now

for what it is worth, at the time of the MP capping proposal, i was unsure of what the effect would be, there were pros and cons for both sides, and i was also not against the change

i was considering whether the issue is having too many differing opinions and something like tokenomics should be delegated to a focused team

but i think that would be very difficult, something as sensitive as tokenomics would lead to a lot of unhappy people if just a single team were making the decision on it

i recall at the time of the previous tokenomics change there was again a lot of differing ideas, the ideas were consolidated and the community voted on their preferred solution to focus on

so what we have now is not ideal, but i don’t see a better way to do this, we either continue to allow the community to choose and experiment with the tokenomics (taking up development time for this) or we have a vote to decide not to change the tokenomics again and focus on the protocol

to me personally, not allowing changes seems like it would be the less ideal route

5 Likes
16

agree，concentrate on gmx itself，The cryptocurrency market pricing is still immature, but we are mature.。

同意，专注于 GMX 本身，加密货币市场定价尚不成熟，但我们平台已经成熟了。

During the decline, I continued to buy using my scholarship for this month.

下跌期间，我继续用本月的奖学金进行消费。

2 Likes
17

This is already the default based on the way the DAO and Tally are set up. It just happens to be a team of one guy.

It’s not the community. It’s one guy with the lion’s share of voting power.

In any case, I’m not advocating for not holding votes. I’m advocating for dropping Saulius’ idea because it’s just another case of trying to fix the symptoms (low token price) rather than the disease (people don’t want to buy/stake the token).

If you look back at the way the MP burn argument was put forward by SniperMonke, it was argued that getting rid of our points program would incentivize new buyers and more stakers. That people seeing other holders with 200% MP would be dissuaded from buying the token. Well we now have years of data to show that everyone having the same MP hasn’t changed the quantity staked - it’s been virtually static the entire time.

So why not go back? Q mentioned that the sole benefit of BB&D was that it helped stem the loss of the giga-whales that unstaked and dumped the token since they didn’t have MP’s caging them in anymore. If a lack of APR cap (or very high cap) worked great for GMX, even during the Nov 2021 to Aug 2023 bear market for bitcoin, doesn’t it make far more sense to go back to these mechanisms that worked? Experimenting for sake of experimenting is what got us into this situation in the first place. MP Burn and BB&D were solutions to problems that didn’t exist.

I posted the question in the telegram chat about “If the GMX platform launched today rather than 4 years ago, knowing what we know now, how would the protocol & tokenomics be different?” @xdev_10 would you still have set up the DAO to where a single holder can decide all outcomes? (Including the transfer of treasury funds to him)

Would you have stuck with the original tokenomics or would you have modified it to be similar to how it is today? The pre-Q tokenomics got us to $92 USD per token, and now we’re at $8 after implementing all of his plans.

It’s just a thought-exercise but it could lead us to a solution. The car industry does this every generation - they ask “If the car didn’t previously exist, and had to be invented right now, how would it be different?”

Saulius’ plan punishes the long-suffering stakers even more, to the point where we will encounter cascading liquidations. Treasury & fees will have to get eaten up by the floor price plan, to the point where even Q’s monthly stipend may have to be IOU’s.

Bringing back MP’s and ETH yield is significantly less “experimental” and stands a better chance at putting us back into the pre-Q “upward momentum” period than this plan that Saulius is putting forward.

I can’t imagine anyone saying that if GMX Platform was launched today, the native token would have an APY cap. The project would last a week before being shut down. Virtually everyone WOULD say that it would include a points program, just as the dozens of competitors that are leaving us with a smaller & smaller piece of the perps pie have.

1 Like
18

Thanks X, and yes focus on growing the protocol, the only thing that matters!

That’s why the only precious coding time should be fee reductions for certain amounts of GMX staked (including a few royal slots at 1M GMX staked)… That gives both real traders incentives to actually buy the token and keep it, and at the same it builds the protocol volume.

19

1.6M $GMX was bought back, which breaks down as follows:

  • 933K $GMX compounded (57.27%)

  • 408K $GMX sold (25.05%)

  • 288K $GMX dormant in the rewards contract (17.68%)

Equivalent to BB&D contributing 1.2M $GMX in buy pressure. However, since the launch of BB&D, GMX has experienced three large-scale sell-offs:

  • The first was a 300K $GMX whale dump,
  • the second was nearly 500K $GMX dumped due to the GMX V1 attack,
  • and the third was close to 400K $GMX dumped from price manipulation.

These total 1.2M $GMX in sell-offs, essentially completely offsetting the effect of BB&D.

Essentially all marginal buying pressure for $GMX comes from BB&D, yet you’re engaging in simplistic attribution and even blaming BB&D. With BB&D, $GMX is at its current price; without BB&D, removing 1.2M $GMX in buy pressure, what price would $GMX be at right now?

As for your mention of GMX rising from $2 to $92 and calling it the pre-Q era, that’s interesting. You further attribute the decline in GMX’s market share to changes in the economic model, rather than competition from all sides. A direct piece of evidence is the LP side, which is unaffected by the economic model—if you switch to TVL in ETH, you’ll see whether the inflection point aligns with the economic model change.

GMX has done a lot and done it well, but not enough, and not well enough. It will do more and do better. Before spraying nonsense, explain the data above instead of your imagination.

4 Likes
20

agree that we should not experiment just for the sake of it, and if anything, we should try to learn from the previous round of discussions and votes that led to the changes

i believe we had already attempted to approach the previous discussion carefully, and the only thing we can do is to be even more careful this time, to help ensure that the community is provided with sufficient data, evaluation and analysis to come to the right decision

i would reiterate that i am not 100% certain whether the tokenomic changes over the years are truly a significant cause affecting gmx’s current price

i personally can’t fully ascertain if the previous tokenomics are better or if the current one is, the previous tokenomics had some interesting features like multiplier points, but they also came with their drawbacks, the current tokenomics to me are simple and elegant in their own way, if i recall correctly, gmx’s token price was already on a downward trajectory before most of these tokenomic changes

if i had to guess the true cause, it would be a dilution of attention with the increase in perp dexes, and due to that, as you mentioned, attempts to change / fix the tokenomics yet again are again trying to treat the symptoms instead of the cause

so i feel we should approach the discussions with that focus, what can we do to bring attention back to gmx, would reverting to the old tokenomics bring back attention, i personally do not think so

we either need something innovative, or even something simple like was suggested of setting up a floor price fund, we need to evaluate these based on the environment and participants today, whether these tokenomic changes would make gmx as a token unique and attractive

and also whether it is even tokenomics that would have the most impact, or are there other focuses that would bring back this attention

internally within labs we are also working based on this

about DAO voting being controlled by just a few people, i would also prefer more people to be delegated to, but i don’t feel it is a large issue that requires focus at this time

DAO delegates are specified to only vote based on the results of snapshot votes, for snapshot votes distribution of voting power is much wider as it is based on token balance instead of delegation

3 Likes
21

I have a sense that many are in favor of creating fee tiers also using staked GMX as a factor. Is it maybe wise to create a seperate topic on this and it’s implementation. Just to discuss seperately?