Don’t you consider Polygon a competitor ?
There the transaction cost is less than 1 cent
Arbitrum could lower transaction fees significantly in many ways at the cost of security, for example, by using a DAC or Celestia for DA. We believe one of the strongest incentives to build on and use Arbitrum One is that it inherits security from Ethereum to a high degree.
1 Like
Thanks for the proposal @Entropy!
We are supportive of the increase in L2 minimum base fee. While we don’t think an increase to 0.05 will have a material impact on organic activity, maybe it would be fruitful for both the DAO and the working group for Milestone 2 if we start with an increase to 0.03 with the goal to raise it to 0.05 after a short period of time. This should hopefully provide some additional insight into the effects of raising the L2 min base fee.
Regarding Milestone 2, we’d be in favour of Option 2. But we do have concerns about the need for constantly fine-tuning these fees and the costs that will come from funding a specialised group. We would also like to see a veto mechanism incorporated / Guardian role just for additional safety in case of any errors when updating the L2 min base fee (oSnap could also be used here).
Lastly, one overlooked point regarding the elasticity of users is that the DAO is currently spending millions of ARB in incentives to incentivise activity on the chain, so even with an increase in the L2 min base fee the expected cost will still likely be negligible for most users.
2 Likes
Thank you for proposing these adjustments to the gas fees, @Entropy. This is an area I have been concerned about since the Atlas upgrade. Increasing the gas to a minimum base fee of 0.03 and 0.05 still keeps the fees very low, so I don’t expect a significant impact on the user experience. However, I would like to understand more about how you arrived at these specific numbers. Is there a particular reason for choosing 0.03 and 0.05? What differences or benefits do these specific values provide compared to other potential options?
Additionally, we need to find the optimum point for the gas fee base to ensure long-term sustainability. I love the idea of having a working group to focus on this specific area and make adjustments based on market conditions.
Thanks to @Entropy for the proposal!
I’m glad to see suggestions in this area. I support raising the minimum base fee, as it will provide a more solid foundation for the subsequent arb economic model.
We have been made aware that per our request, risk member of the ARDC @chaoslabs is slated to look into this proposal and its implications. We will wait to proceed to Snapshot until after their analysis is complete and delegates have had time to digest it.
5 Likes
At SEED, we believe this proposal is well-intentioned and holds potential benefits for the ecosystem. We think it is a good proposal, as it can help streamline various processes within the Arbitrum network. At the common user level, the proposed increase in gas fees is not particularly significant and should not pose a substantial burden.
However, the timing of this proposal appears suboptimal, since these last couple of months we spent a significant part of our treasury (eg. Catalyze Gaming Ecosystem), so these actions feel like covering a stab wound with a band-aid. It is essential to synchronize it with the Strategic Treasury Management proposal to ensure cohesive financial planning and execution. We consider it might be a good idea to align this proposal with the Arbitrum DAO Budget proposal that you guys also created.
We are also interested in understanding how this gas fee increase might impact builders who are currently deploying or those planning to deploy on Arbitrum. Assessing the potential impact on developers is essential to gauge the proposal’s broader implications.
In the event that both this proposal and the ARB Staking: Use Surplus Fees to Align Governance proposal are approved, 50% of the hypothetical fees will go to the ARB Staking initiative. Has this potential allocation been considered?
Overall, while we support the core idea of the proposal, we believe these considerations and alignments are crucial for its successful implementation. Thank you, @entropy, for putting this proposal together and for considering our feedback.
5 Likes
Supporting this proposal is wise as it promptly addresses the immediate revenue decline by adjusting the L2 base fee while laying the groundwork for a flexible, long-term solution. This dual approach secures essential ETH inflow, sustains Arbitrum’s competitive edge, and ensures the DAO’s economic stability and adaptability in an ever-changing market environment. 
2 Likes
Thanks to Entropy for the offer!
The proposal discusses adjusting Arbitrum’s gas fees by increasing the base fee from 0.01 to 0.05. This change aims to boost the DAO’s revenue without significantly affecting user decisions due to the minimal cost difference. It also suggests collaborating with trusted service providers (SPs) for Milestone 2, including implementing KYC measures and possibly automating fee adjustments based on other L2 activities to enhance flexibility and efficiency.
Supporting
- Increasing the base fee minimally impacts users but significantly benefits the DAO’s revenue.
- Working with trusted SPs and implementing KYC measures ensures reliability and security.
- Automating fee adjustments based on other L2 activities could enhance efficiency and flexibility.
I support the proposal due to its potential to increase the DAO’s revenue with minimal user impact and the inclusion of measures to ensure SP reliability and security. However, the feasibility of automation should be thoroughly assessed to ensure its practicality and efficiency.
4 Likes
I support adjusting Arbitrum’s gas fees to increase DAO revenue. This change has a minimal impact on users but significantly improves resource utilization.
Suggestions:
-
Technical Feasibility: Ensure the technical solution for automatic fee adjustments is viable and implement it promptly.
-
User Feedback: Establish a feedback mechanism to continuously collect user opinions on fee adjustments for timely optimization.
I hope these suggestions help improve the proposal and ensure its successful implementation.
I don’t understand why everyone writes that this has minimal impact on users?
First:
You want to increase the cost of gas by 5 times. This will affect the cost of gas for users also by 5 times.
If you make 1 transaction per month, then of course this does not affect anything, but for active users it will be a rollback to previous indicators.
Secondly, we all voted to reduce the base fee to 0.01 , and in a couple of months we want to roll back. Why?
Third, if we vote for ARB staking, then I can still understand the benefits for ordinary users.
But what are the benefits of this increase now?
Just show that now making a small profit is a problem. What is it, what is this money intended for?
1 Like
As part of our efforts at the Pantera Research Lab and the Pantera Catalyze Research Fellowship, we’ve explored the potential impact of a fee adjustment.
Below are findings that may contribute to the discussion here:
- A study from Pantera Catalyze fellow and UCLA PhD student Dongryeol Lee found that Arbitrum users are price elastic. This would argue against a price increase since it could drive away users. The net effect on revenue could even be negative despite the fee increase.
- Lee’s study design uses pre-4844 data. Since 4844 was a big change and may have impacted elasticities*, the finding may not be decisive.
- We conducted a follow up analysis and observed that price elasticity of demand is significantly lower post-4844. This means Arbitrum users are less elastic than Lee’s study suggests.
- One possible tie-breaker is that current thinking suggests L2 markets are winner-take-most. If Arbitrum should be targeting market share and not revenue in the short run, this would argue against raising fees.
*e.g. shifting the supply curve to an equilibrium where elasticities are lower.
Why do we care about ‘demand elasticity’ and why is it hard to measure?
The elasticity of demand predicts how users respond to fee changes, like those being considered in this thread. Given how important elasticities are, it would be nice if it were simple to measure them. Unfortunately there’s a saying in economics that applies here: “Don’t reason from a price change.” Econ 101 tells you that prices and quantities are the result of supply and demand, so we look for cases where it’s plausible that there were unanticipated shifts in supply and then measure the resulting effect on demand.
The figure below explains this in the context of semiconductor production. To discover the elasticity of demand for semiconductors, the authors needed to find sudden unanticipated changes in supply. Using the supply chain disruption caused by Covid, they could then observe the slope of the demand curve and determine the elasticity.
Here is more of the nitty gritty of our analysis:
To address the challenges of estimating demand elasticities, Lee employed a Two-Stage Least Squares regression using the plausible exogeneity of L1 gas fees as an instrumental variable. In the first stage, Lee instrumented Arbitrum’s gas fees using L1 gas fees to predict transaction values based on L1 fee variation. In the second stage, he regressed Arbitrum gas usage on these predicted values, effectively isolating the variation in Arbitrum gas fees that is independent of user demand. Below is a table from Lee’s analysis, showing the estimates from the second stage of his regression:
In order to test how applicable this finding would be post-4844, we decided to test whether 4844 had a significant impact on elasticities themselves. In our subsequent analysis, using the sudden fee adjustment of EIP-4844 as a shock, we observed that Arbitrum users became less sensitive to fee fluctuations (decreased elasticity.) This indicates that post-4844 users are less elastic than in Lee’s analysis.
In conclusion:
Our strongest evidence suggests that a fee increase on Arbitrum would cause a decrease in user demand. This decrease could be large enough to offset the revenue from a fee increase.
However, that evidence relies on pre-4844 data and we are in a post-4844 world. A further analysis we conducted indicates that 4844 did make Arbitrum users less price sensitive. We also observe that there is a good argument against raising fees even on Arbitrum in general: that current thinking suggests L2 markets are winner-take-most. That is, if the goal is acquiring and retaining users in the short run, then it may be ideal to charge them as a little as possible.
5 Likes
Hey Pantera team! I think I speak for everyone when I say how happy we are to have you guys here actively contributing in Arbitrum governance. Huge W for the ecosystem.
However, I would have like to of seen “user” defined in Lee’s study. There is a ton of bot activity occuring on Arbitrum One, and to not segregate the users into cohorts somewhat misses the forrest for the trees in my opinion. We need to know how REAL users and developers would be affected by a minimum base fee increase rather than bots, because i am of the belief that increasing fees for the latter camp is okay. What we don’t want is to lose actual people with organic demand to use ArbOne to other networks due to demand elasticity in rfr to fees. To reinforce my point, consider the fact that a project not many in the community have heard of or used (UXLink) has held a 30-60% market share over ArbOne transactions for months now (Dark Blue shaded).
To reinforce this point, if we look at the amount of revenue being driven to the DAO on a per transaction basis - we are earning far less on UXLink “users” versus legitiment DeFi primitives like GMX and Uniswap. For UXLink, 0.000002 ETH of fees are generated for the DAO for every contract interaction versus 0.0001 ETH for GMX V2 and 0.00003 ETH for Uniswap’s universal router. That’s ~100x and ~10x differences respectively when compared to UXLink. An increase in the L2 minbase fee would force the UXLink “users” to pay more but would largely leave GMX V2 and Uniswap uneffected assuming no congestion similar to what we saw with the LayerZero airdrop that drove congestion fees extremely high. Additionally, given the fact that Arbitrum is the leading chain for DeFi - I am hard pressed to believe that a user with a $5,000 position that needs to be managed is thinking twice about making the move based on a $0.05 fee or a $0.50 fee.
In terms of other ecosystem’s strategies, specifically base - which has been gradually INCREASING the target gas per block in an effort to increase throughput and lower fees, that is indeed a viable strategy. And to your point, maybe prioritizing this short term growth in usage metrics is the better move for the long-term. My view is that we have orbit chains that can do whatever they want with gas configurations, but ArbOne is premium block space secured by Ethereum with actual security guarantees… and should be priced as such. We are handing out ARB left and right as incentives, so why make block space even cheaper when a vast majority of activity still happens on ArbOne? If there actually are certain use cases being built out today that need additional/cheaper throughput (I personally don’t think that’s true - and there are more optimal options on the market today that would be more suitable than ArbOne for extremely high throughput use cases, so why would they not build there anyways?), then we should not be thinking about increasing the base fee but instead lowering fees while juicing throughput. But I’d rather profit off the bot activity / airdrop farmers / mercenary capital for now and see if base’s thesis actually plays out. And like I said, you can still experiment at the orbit level as to not miss out on innovation at base and other ecosystems operating under the same thesis: prioritize growth and do not concern ourselves with profit.
I would be keen to see a revamped study on the elasticity of user demand in relation to tx fees, especially considering blob prices are near 0 as of now and the L1 gas spikes (outside of the blob market) have no impact on L2 fees anymore. I believe @chaoslabs is nearing completion on their backtesting on a min base fee increase as apart of their scope with the ARDC, so I will be sure to drop that here as well so you guys can check it out!
All this to say - you guys raise some great points. It becomes a strategy question of “should the dao focus a little bit on profit, or solely focus on growth?” I do not have a good answer here, but my opinion is to focus a little bit on profit so we can more easily navigate the next bear market without being forced into selling ARB tokens at low valuations.
3 Likes
I do want to insert into this discussion the following graphic courtesy of DefiLlama that to me shows a clear relation between the decreasing valuation of the ARB token and the abrupt decrease in fees after the EIP-4844 update.
I do think that the as @PanteraCapital indicated, users are price elastic but I don’t think they are at these kind of levels which is an order of magnitude lower than pre-EIP 4844.
To me it’s clear that if we revive fee income closer to the levels of pre EIP 4844, ARB token price will follow, which will in turn drive more capital into the ARB ecosystem (As ARB gets more valuable more can be lent on AAVE against ARB, the Uniswap/Camelot etc LP’s with ARB will grow larger, the STIP rewards get more valuable etcetera etcetera.)
1 Like
Before increasing fees, another idea that could improve the DAO sustainability is to stop squandering funds into pointless “grants” with almost zero benefits and which have absolutely cratered the ARB token and made it one of the worst performer of the last 12 months. Of course this would require that the people receiving the grants have no acquaintance with the people voting to distribute the grants and that seem to be a problem that many do not want to see solved.
1 Like
this might just be a case of post hoc ergo propter hoc. Linking the price of arb to this (or any) event might just make us look at all elements from the wrong point of view; and beside, the event we had in march (the huge unlock) is potentially more impactful on the price than anything.
Would take this occasion to link the study from chaos labs about increaasing arb base fee, worth reading. [Request for Support - Chaos Labs] Backtesting and Suggestions For Min Base Fee - #5 by chaoslabs
1 Like
It’s not because correlation does not necessarily mean causation that it never does.
Correlations by themselves do not prove something but they are not worthless.
Besides when it comes to economics, proofs are pretty much impossible since it’s not an exact science anyway.
Another thing I want to outline is that there seems to be severe resistance to change.
But one should see it as a revert to normalcy.
In actuality, change to the fee system was already forced upon us through EIP-4844 and we should take steps to recreate the fee market as it was prior to 4844.
1 Like
But for me it is not at all obvious. How is the price of ARB connected with the price of gas?
The price of ARB is not connected with this in any way and I do not understand where such conclusions come from. ARB is now connected only with voting.
Fees are a revenue source for the foundation, no?
(Arbitrum Foundation: L2 Treasury Timelock | Address 0xbfc1feca8b09a5c5d3effe7429ebe24b9c09ef58 | Arbitrum One)
Revenue for the foundation → value for the token that governs said foundation.
1 Like
Until the staking proposal is accepted, all these incomes will simply sit in the account and will not affect the token in any way.