Thank you for this explanation. I understand it now much better. One maybe stupid question: Why don’t we just implement a burn mechanism. More usage = more burned, Less usage = less burned. Wouldn’t that equal to the same regulation of inflation without reinventing the whole delegation process?
Can you confirm that the AI which will be included in O will have this flag. I think it’s important to have ability to not accept co-staking.
Yes, the plan is indeed to allow bakers to chose whether or not they allow co-staking operations. Bakers will have control over a couple of parameters, and will be able to forbid staking operations towards them. A public announcement will be released in a few weeks, explaining the feature in more details, so please look forward to it
Has anyone considered how this incentivizes large bakers to drive inflation down, potentially to the detriment of smaller bakers? As these larger entities suppress inflation, they can continue to operate, possibly even profiting from negative inflation. Meanwhile, smaller bakers may be compelled to either completely exit the market or join larger bakers through co-baking, incurring additional fees. This seems to encourage centralization. Does this not seem concerning to anyone else?
Iirc ETH researched it as griefing attack. And this should be why polka curve looks the way it looks (increasing inflation slightly up to the target level to balance out)
There is already a burn mechanism, storage fees are all burned, gas fees are tiny and even if they were burned too, it wouldn’t make any difference in regards to the inflation. Even if all blocks were full, the amount of fees burned would be small compared to the 46m xtz that are emitted per year.
Looking at how much fees are generated by a chain is misleading. Supposedly we want mass adoption and as many people as possible using the chain so you would want fees to be as low as possible so the chain can be inclusive.
Like we could double our fees costs right now and then Tezos would generate double the amount of fees, isn’t that a BS metric to go by?
Don’t profit from inflation out of thin air, profit from price appreciation.
As a part of the Oxford protocol announcement, we have shared a high-level functional specification of Adaptive Issuance and Staking.
Our assessment and current plan to address the two issues recently reported on the implementation of Adaptive Issuance in the Oxford proposal :
Hello Tezos community!
At P2P.ORG we’ve been involved with Tezos since the early days and I’d like to share the respect to everyone in the community who actively participates in discussions and thinks towards Tezos improvement.
The purpose of this post is to share our perspective on Adaptive Issuance [AI] and its potential influence on baker, delegator, staker economics and collect more feedback from bakers and delegators.
We’ve made a public Oxford AI spreadsheet so you can get a grasp of the numbers in different scenarios. Feel free to make a copy and play with it, if you will find any critical issues with the calcs, let me know.
After AI activation, delegators will start earning less rewards than stakers. Initially, it will lead to delegators transitioning to the staker role raising amount of XTZ locked → lower issuance (current amount of staked XTZ is ~7%).
After staked % doubles and grows further, baker economics will be under pressure as fee gains will decrease along with AI. It becomes crucial in current market conditions. AI activation can take ~4-6 month (fixes in the next proposal + separate activation) but we can’t properly forecast the timeline for better market environment. It’s important to ensure stability in this period.
At 30% of staked XTZ APR will be ~0.66% for delegators and ~1.32% for stakers. It will cause bakers who operate as a business to raise the fee in order to remain profitable decreasing rewards of delegators and stakers even further. At high staked ratio they might raise fees up to 50-70% to maintain a profitability breakeven point assuming current XTZ price of ~0.7 (lower selling pressure won’t cause XTZ price to skyrocket x times).
Possible outcome: Some delegators can leave the ecosystem while some bakers can shut down their operations.
In order to be a baker you need to provide a slashable self-bond that can be considered as an opportunity cost versus alternatives like simple ETH staking. Even assuming that ETH APR will decrease over time, self-bond reward rate (RR) that is comprised of earned rewards + fees from delegated and staked XTZ should ideally be higher to justify running a bakery. In Oxford it’s hardly achieved at staked XTZ of >40% and might be even harder for smaller bakeries.
Possible outcome: Bakers can withdraw self-bond and deploy funds to more fruitful sources of income.
There are clear benefits of AI that were emphasized by @murbard pretty well so I won’t repeat those again. All of the above are just thoughts as in the end it’s very hard to predict the exact behavior of participants and network parameters.
We encourage bakers and delegators to calculate and share their perspective on economics under various AI conditions. We might need to fine tune AI parameters based on the feedback to ensure lower pressure on baker economics in the days of uncertainty.
Following our previous statement regarding Adaptive issuance, we at Everstake have published our next statement along with some proposed suggestions to the original Oxford proposal concerning Adaptive Issuance.
Your input and participation in the discussion would be greatly appreciated! Oxford proposal - adaptive issuance discussion - #7 by Anna
We would like to share Chorus One’s perspective on Adaptive Inflation and engage in a discussion about the current Oxford proposal.
Chorus One has been actively involved with Tezos since its early days, and it’s great to witness new insights into tokenomics and the valuable discussions taking place within the community.
According to our analysis, we have observed that the current economic situation on the Tezos network is hardly profitable, even for top validators. Introducing changes to the tokenomics and implementing Adaptive Inflation could add additional economic pressure that might force bakers to shut down their operations on the network due to unprofitable activities.
A decrease in inflation could also result in an increase in fees from bakers to remain profitable. However, this is not a sustainable solution either, as it would place economic burdens on delegators and stakers as well. Adaptive inflation and reduced inflation do not necessarily guarantee an increase in the token’s price: in some networks, it has shown adverse effects.
Considering these factors, we believe that the present market conditions are not favorable enough to implement uncertain tokenomics changes that could place additional economic stress on stakeholders.
We encourage stakeholders to share their opinions and actively participate in discussions to collectively arrive at a solution that benefits the entire Tezos community.
Why will fees be decreased? They are not adaptive like the block rewards afaik
This is an assumption. Besides the “for profit” bakeries that are also multichain, the operational costs to run a bakery are really really low.
You are assuming that the reduction in issuance won’t have any effect on the price, why?
Another possible outcome: Some delegators will choose to stake instead of delegating to benefit from the extra rewards. Some other might decide to start their own bakery as well.
Again, smaller bakeries have much lower operational costs, with the current price action of Tezos it is already an oportunity cost to run a Tezos bakery compared to staking ETH. That’s why tokenomics need to change and fix the price action. An eternal issuance that keeps adding to the selling pressure from exchanges and big bakers and people that have the tez liquid at all times is a terrible way to keep going.
Thanks for joining the conversation!
For example, public bakers charge fees from rewards. Amount of XTZ rewards depends on issuance. If issuance decreases n times, amount of rewards is expected to decrease in a same pace. Thus, the same fee will in % will result in lower amount of XTZ in absolute terms.
There are 100+ bakeries operating as a business and cost structure includes personnel besides infra so yeah, costs are higher but this group represents a significant share not all delegators are comfortable with running infra by their own. It would be harder to justify economic incentive to run a bakery versus an option to simply stake ETH.
Price of a token is just a function of its fully diluted valuation. Current FDV of Tezos is ¬620M. After AI implementation issuance may decrease 4-5 times influencing baker economics. If you assume that price will make 4-5x to compensate this decrease (that would result in the same $ denominated rewards) you’d get Tezos FDV of ¬2.5 billion USD, which is unrealistic to be achieved only by decreasing selling pressure.
Not everyone is born as a baker. The ones who believes in the long term success of XTZ might stay, the ones who stake for rewards might leave. That’s not necessarily bad but it may cause fluctuations that only add to the public baker economics uncertainty.
I am not against the concept of staking and AI in general, there are clear benefits that you’ve just mentioned. I am concerned about baker’s p&l, not sure what costs solo bakers have. I would be interested to see some calculations of solo baker financials to understand it better. Overall idea is to ensure that from a baker perspective it is still reasonable to lock funds in a self-bond versus ETH staking. From what I observe for many bakers operating as a business it’s not the case under current AI parameters.
I do not agree with your argues at all, you can check my profile and former messages to know why.
In any case, please share your decision through your twitter so your delegators can decide keeping their delegation on you (cause may share your vision), or not. Last phase you vote Yay, which has been the common vote for other proposals/phases, so it would be fair to do that communication as soon as possible.
Then you should vote no for Adaptive Issuance specifically, using the toggle vote. Instead, you voted no for Oxford, which does not enable Adaptive Issuance. Why are you blocking Oxford?
I am a public baker as well. A small one. I run my baker on an Orange Pi.
I had one time Hardware costs of <400$ and running costs of ~2$ per month for electricity. I do not count DSL-Connection because I have it anyway.
Even with just 60.000 tez my baker is profitable.
The amount of work I put in is ~ 1 hour per month for updates and votings.
Of course I can’t make a living of this bakery, but it pays for itself.
I run a ghostnet baker on a raspberry pi with over 1million tez. That one has some lost attestations and a few lost baking assignments but there is no significant loss. So even for a bakery with higher stakes, lowcost equipment is still an option!
Why is it unrealistic? 2.5B FDV is not that much if you compare to other blockchains that have as much or even more.
I thought in your initial post you were talking about the fees included in blocks but you meant the fees bakers charge delegators. I misunderstood that part. Thank you for clarifying.
Thanks for sharing, that’s super interesting!
In our case we can’t save a lot on infra. If delegators lose rewards due to significant downtime we will have to compensate and the idea is to minimize the probability of that event. The most severe risk here is a loss of reputation, as it might affect perception of us in eyes of delegators and in other networks.
Infra is not the highest item in terms of the costs, personnel is. We want to provide good staking/delegation experience (DevOps to manage infra, support, governance, etc.) so P&L looks slightly different.
At these levels of FDV, growth in such multiples can be derived from the improvement of fundamentals. FDV by itself is a function of network utilization and expectations over its future dynamics.
AI by itself does not increase that by 3-4x. AI brings positive properties that could create such indirect influence but it won’t be explosive and can take a decent time. During that period public bakers might continuously incur loses in $. If many public bakers will shut down their nodes it’s not a positive thing IMO, having different types of bakeries is good.
P.S. I don’t think that many networks that have higher FDV are priced fairly.
Always happy to give insights. My delegators did not lose much rewards either. My efficiency in the last 50 cycles is ~ 98.5% My actual ROI is calculated with 6.29%