Proposed SIP: Modify Coinbase Reward Per Block and Halving Schedule

[muneeb]

Thanks @xan for posting this! There has been some discussion on GitHub already that might be relevant. See this comment that I left today.

As I mentioned on the GitHub thread that I personally find this SIP interesting and want to explore it more. There are two main reasons:

a) Total security budget of the network: the security budget of Stacks depends on block rewards (currently 1000 STX), gas fees, and the respective market price of STX. Depending on how you count in 2021, the network had approx $60M-$70M total security budget. This number is an approximation because the STX/USD and BTC/USD price pairs fluctuates but we can use that approximation for now. Approx 1,907 BTC has been spent by miners since the launch of the network. In the long-term, if we get lots of developers and apps on the network, the total security budget might be in billions of dollars per year (Ethereum and some other smart contract platforms already have security budgets in billions).

In the initial years after launch, it’s typical for smart contract platforms to mostly rely on block rewards for the security budget and mining incentives. With increased network use, the gas fees can become more significant and with any appreciation of the gas asset the security budget / mining incentives can also increase over time.

The security budget has gone down recently with the market downturn (which was to be expected given this industry moves in bull/bear market cycles). The SIP proposes an interesting change where the total security budget can be increased by 60% in the coming years will some long-term increase in supply. Many people don’t know this but the first version of Stacks tokenomics had a 5B supply (2017), which was later reduced to approx 2B (2019) and then further reduced to 1.818B (2020). The proposed change adjusts the 2050 supply back up a little to the 1.9B range. Also, it’s important to keep in mind that Stacks is not hard money like Bitcoin. Stacks is a gas asset used to incentivize miners and to pay for smart contracts on the network. Unlike Bitcoin which doesn’t change, Stacks is expected to go through more rapid upgrades to find the right set of features and incentives for growth.

b) BTC rewards incentives during earlier years: the second thing that will result if the SIP is accepted and adopted by miners is that the Bitcoin yield in the coming years can increase for users. In the first two months after launch, the network had 2000 STX block rewards (instead of 1000 STX block rewards) and we saw significant interest from miners, developers, and users at that number. I think that increased rewards during the initial “bootstrapping stage” of the ecosystem are interesting. They give increased incentives for new miners to join, new developers to experiment with apps, and new users to explore the ecosystem and earn BTC rewards. Once the ecosystem reaches escape velocity then the gas fees will start having a significant impact on both the total security budget and the Bitcoin yield, and the increased incentives become less relevant.

This SIP proposes an important change. As a decentralized ecosystem I expect various community members to chime in with their input. The SIP process is very rigorous, it gives ample time for everyone in the community to give their input, there is an explicit voting and miner adoption phase. Voting is typically done by stackers i.e., community members and entities who are actively participating in the ecosystem. Miners are independent and they ultimately decide if the upgrade is accepted or not. Right now we’re at an early discussion phase of this SIP.

[pws]

+1 I really like the idea of us being like an adaptive system that understands itself and doesn’t sit on its hands when it should be anticipating challenges and opportunities. Really only Bitcoin can afford to be the immovable object that rarely changes.

I view Stacks rn like a Bitcoin-yielding bond that incentivizes building on Bitcoin and we have not even hit the maturity of our ecosystem, development-wise where that BTC yield and security budget has even been put to its best use. Our best years are ahead of us, and we may need to acknowledge that with a potential SIP proposal here, IMO.

[rhodan.btc]

looks like a try to make stx mining profitable in a bearmarket on the cost of the community with higher supply / inflation. I don’t like the point that we start adjusting major mechanics of the chain just how we like and feel about it, which results in a mayor loss of trust and reputation of the stx chain - investors don’t like uncertainties for possible future changes, especially not interventions in mayor topics like supply and inflation dynamics.

IMO we should start thinking about how we could make mining sustainable in the long run and especially how the mining mechanic is able to bind more miners into the process. Which probably needs a fully revision and maybe we find here also a way to stack without the need for pools.

PS: Never liked the STX supply for a Bitcoin Remora Chain, Layer 1.5, or how we call it today. Bitcoin has a 21 Million Supply Cap, why not 34 Million or 55 Million for STX as a tribute to Fibonacci, which would show the BTC | STX symbiosis quite better.

[larry]

@rhodan.btc

looks like a try to make stx mining profitable in a bearmarket on the cost of the community with higher supply / inflation. I don’t like the point that we start adjusting major mechanics of the chain just how we like and feel about it, which results in a mayor loss of trust and reputation of the stx chain - investors don’t like uncertainties for possible future changes,

For what it’s worth, this was my first impression of this as well: a STX mining company is putting forward a proposal to pay themselves more money. I don’t think the proposers have anything but good intentions, but appearances matter!

@jude wrote:

This entire argument assumes that if you mint more STX per block, the STX will have the same valuation as if you had not done so.

The security budget of the Stacks blockchain depends on both the amount of BTC spent per block, and the number of block races an attacker must win to reorg the chain. The number of tokens emitted per Stacks block is orthogonal to this. If you want the amount of BTC spent per block to increase, then the aggregate value of the Stacks block rewards must increase first. Simply sub-dividing the aggregate worth of the block into more units would not achieve this by itself.

Put another way, why should I believe that e.g. 2x-ing the coinbase will 2x the amount of BTC spent on it, when basic economics would suggest otherwise? If you believe that there is a linear relationship here, then why not increase it a billion-fold?

I think @jude is spot on here. The basic economics of this don’t make sense as explained.

@muneeb wrote:

When I’m bringing up mining incentives I’m mostly concerned with total security budget of the network. Miner decentralization is a secondary concern for me. The total security budget for the network has gone down recently and can go down further if markets continue to drop and/or we don’t see significant growth in apps/users in the next 1-2 years.

The way to increase the stacks security budget in real terms is for STX to be more valuable. That’s a function of the market for STX and requires an increase in demand of stacks compared to the supply. Assuming demand stays constant, putting more STX on the market increases supply without any obvious mechanism to increase demand.

If the concern is increasing the security budget in STX, there are other ways to do it besides changing the money supply. For example, the system could charge accounts that haven’t made a transaction in a given amount of time a fee and pay that to miners.

@muneeb

The SIP proposes an interesting change where the total security budget can be increased by 60%

Again, it’s important to note that this 60% increase is only in STX terms - not in real (btc/usd/bars of gold/loaves of bread cost to attack) terms.

@xan wrote

I think you bring up an incredible point that applies not just to this SIP, but to the SIP process in general. There needs to be a forum that is easier to make proposals on and for discourse to happen for those that aren’t developers.

I agree with this. Good thing we have this forum that runs discourse.

I use Github every day, all-day and I didn’t see these proposals. I must not have one of their complicated notifications set track issues in that repository because I’m not a member of the organization. There’s probably a way to solve it, but .

---

As others have noted the SIP itself isn’t very clear on what problem this is intended to solve. This feels very much like a solution looking for a problem. Several have suggested some problems this might solve: increasing security budget, attracting more users, increasing stacking yield, etc.

There’s a lot of speculation about what effects it might have, but the effect of this proposal we can say for sure is that it slightly transfers ownership of the total STX supply from STX holders to miners and to do so more to earlier miners vs later miners compared to the current system. Is this good or bad? I can’t really tell.

Why is 60% more STX for security budget today going to lead to our ecosystem’s success while the current STX budget won’t?

How is transferring a larger percentage of the STX supply to our ~5 whale miners going to attract more users to our ecosystem?

Are any of these things 10x-better-enough to be worth the credibility hit that comes with changing the money supply? I’m not convinced.

[muneeb]

Let’s look at the first two months after mainnet launch. The 2000 STX reward resulted in a higher BTC APY, which was resulting in more demand for Stacks and for locking STX. In fact, we saw a slow down in demand for stacking after the APY dropped. So the proposed change is not only increasing the short term supply, it is also increasing demand because of the higher BTC APY.

Obviously if you suddenly flood the market with too much supply there would be negative implications but the proposal is talking about a curve adjustment based on real world data that we’ve seen before.

I totally agree that randomly changing supply has a cost in terms of how much users can trust that the supply won’t change in the future. Hence the rigorous SIP process and these discussions. I don’t think any change can go through that easily but we should be more open to changes in general than Bitcoin (Stacks is not a store of value asset and hard money like Bitcoin).

I think it might make sense to even separate out the supply change. Let’s assume there is no change to the year 2050 supply. We’re only discussing the emission curve and debating adjusting the curve to (a) increase total network security and (b) interest in stacking for the coming years.

@larry

Assuming demand stays constant, putting more STX on the market increases supply without any obvious mechanism to increase demand.

Demand does not stay constant. The increased Bitcoin yield would give you a different demand from what you have today.

@larry

There’s a lot of speculation about what effects it might have

I don’t think it’s speculation. It’s based on the real-world experience of the boosted rewards of the initial two months. When those boosted rewards ended a lot of us (in private discussions) wished that the boosted rewards period was designed a bit better and it should have lasted longer.

[muneeb]

Let me explain a bit more.

Security budget = (block-rewards + gas-fees) * STX-price

The average value of a stacking slot follows a similar equation with the same three variables.

I think that we’ll all agree that a high number for security budget and a high number for average value of a stacking slot are good things for the health/growth of the network.

Out of the three variables, we don’t know how (a) gas fees and (b) STX price is going to play out in the coming years but we can assume that both these numbers can grow in the coming years with more network growth (more developers building apps and users using apps).

The variable of block rewards was pre-determined before mainnet launch and is the only variable where we can predict how it is going to play out.

The reason why the emission curve is currently scheduled as 1000 STX dropping to 500 STX and so on is that there is an assumption there that in the coming years gas fees will compensate for the drop in block rewards. And also potentially STX price can also increase with network growth increasing the security budget.

The argument for this SIP is simply that higher block rewards are more needed in the early years than later on. If the gas fees were already becoming significant (say on average 500 STX or more per block) then we wouldn’t be having this discussion.

So the discussion that we should be having is how many years of bootstrapping time will the network require where the block rewards will be the main variable in the equation above? My guess is at least 2 years, but potentially more hence I’d be supportive of adjusting the emission curve to have higher block rewards in the next 2-6 years. I care less about block rewards later on because if after 6-7 years of operation the gas fees and/or STX price is still low then there is something fundamentally wrong and you have bigger problems. However, it’s important to give the network the time to reach escape velocity and optimize the emission curve for the coming years.

@larry

For example, the system could charge accounts that haven’t made a transaction in a given amount of time a fee and pay that to miners.

This is an interesting comment/idea. An increase in supply does effectively charge the inactive accounts (the folks who are not actively stacking). If you are an active stacker today (among the 35% of the supply) then you are effectively not getting diluted but in fact are earning proportionally more BTC than the net inflation per year (because the inactive accounts are taking the hit). So an increase in supply actually does cost the inactive accounts, and does not dilute down the active stackers (they’re earning more than the net dilution per year with current market dynamics and can always use a subset of BTC rewards to buy more STX say at the rate of inflation).

My comments above assumed no increase in supply but just adjustment of the emission curve. However, you can see from the reasoning above that even an increase in supply might be worth discussing (although harder to get enough support for). BTW, it’s amazing to see how decentralized this ecosystem is. I can tell from these discussions that the bar to making any change is very high and several different actors (community, miners, developers etc) will need to be aligned for any change and follow the SIP process. That’s a healthy thing for decentralization!

[jude]

I would just like to point out that nothing stops the STX whales from just giving miners 1000 extra STX per block for the next two years. They don’t need to go through the SIP process to do this. They could instead create a smart contract that stores enough STX to adjust the total block reward curve to their desire, and doles it out to miners in a self-service fashion (recall that Clarity exposes the address of the miner who mined each block via get-block-info?, which makes this possible). People who like this idea could contribute STX to this smart contract to keep it going, and no one has to accept higher inflation than what they’ve already priced in when they acquired their STX.

I’m actually serious. This sounds like a win/win – folks who don’t want to adjust the token supply stay happy, and folks who want to increase the block reward can do so (but they’ll have to pay for it).

[muneeb]

Great idea Jude it sounds very similar to the hypothetical drivechains/spacechains like works where people discuss that why don’t folks just donate BTC to a sidechain and use those BTC as mining rewards.

I think a requirement for discussing any SIP should be how practical an idea is and how realistic it is in a real-world setting. Optimizing protocol emission rates is a practical thing, these emissions are meant to incentivize mining and provide a security budget and were designed by humans (me actually). Relying on donations from STX holders sounds more like mockery given how unrealistic this is. We’re dealing with a tragedy of the commons situation here and donations are never a sustainable path out of a tragedy of the commons situation.

[jude]

No mockery was intended.

The uncomfortable reality of the situation is that there is likely to be a swath of STX holders who are adamantly opposed to any inflation whatsoever. This is especially true given the current inflationary climate we’re in with fiat currencies. This would have significant consequences on the ratification of this SIP – even if there were enough votes to pass it, this would not prevent a large-scale flamewar nor prevent lots of toxic behavior poisoning the community during the ratification. I’d like to avoid that if possible. We both lived through the Bitcoin Block Size Wars and the DAO Disaster, and we both know how bad it can get.

What that suggests to me is that we first at least consider different ways of solving the problem that are less controversial. My suggestion was one option. Another option could be to introduce a SIP-010 token that rewards miners for mining blocks. Yet another option could be to get existing or upcoming SIP-010 token projects (or even NFT projects) on-board with rewarding miners with some of their tokens as thanks for keeping the network healthy. All of these could boost the total miner reward, and thus increase the Bitcoin APY.

[muneeb]

@jude

We both lived through the Bitcoin Block Size Wars and the DAO Disaster, and we both know how bad it can get.

I think there is consensus emerging here that increase in supply is likely hard to get support. I can accept that and see the reasoning behind it. So let’s put increase in 2050 supply aside.

That still means that the (a) total security budget of the network, and (b) miner decentralization (measured as no. of independent miners) remain a challenge.

Let’s focus only on the total security budget here. There are ways to address that issue without increasing the supply e.g., by adjusting the emission schedule so that the year 2050 supply remains the same or even decreases.

There are three variables here: 1) block rewards, 2) gas fees, and 3) STX price. Block rewards is the most predictable and obvious thing to tune, gas fees will likely only become significant down the road and not in the next 2 years, and STX price is beyond anyone’s control (the market decides it). One could argue however that increased BTC rewards can lead to more STX locked up (for the higher yield rates) which is a positive factor for the network health and potentially even price (I’m not an economist).

I’d like to focus more on practical and simple things vs complicated/theoretical stuff. There is no SIP-10 token that currently even exists in the ecosystem that can make a meaningful dent on mining incentives (the total marketcap of all SIP-10 tokens is < $25M, xBTC doesn’t count)–a new SIP-10 token with a marketcap of a few millions won’t make a dent. So that idea is not very practical.

[jude]

@muneeb

That still means that the (a) total security budget of the network, and (b) miner decentralization (measured as no. of independent miners) remain a challenge.

What do you want the total security budget of the network to be (in USD), why is that the right number, and how will we know that we’ve achieved it? Also, what’s the minimum number of miners you want to see, and why is that the right number? I think we need to get concrete answers to these questions before we can reason about which strategies and tactics are the best suited to get the job done.

[incite.btc]

@muneeb

An increase in supply does effectively charge the inactive accounts (the folks who are not actively stacking).

It also charges the accounts that are actually using blockchain applications rather than parking their tokens in the stacking system. You’d be actively penalizing the people who matter most to increasing the value of the network. Lots of defi applications become less interesting in an environment with 10% inflation and 18% stacking yield. Why risk my tokens in an AlexGo farm pool or buying the latest NFT when I can just stack them for a riskless return?

Stacking is the application of last resort. It’s the safe thing to do when you don’t want to do anything else, and it generates no external real-world value. It’s only purpose is to act as the token sink for the PoX consensus mechanism. It’s APY should be low to reflect that.

Absent imminent threats to the security of the chain, I don’t see why miners and stackers should be given more network value at the expense of application users which is the net effect of this proposal.

@muneeb

STX price is beyond anyone’s control

In the short term this is true, sure. But in the long term STX price will correspond to the real world value that the blockchain provides. Hyperchains and Native BTC functionality are how we build the STX security budget and its token value. Once we get that, 1000STX per block will be plenty.

[jude]

@incite.btc

You’d be actively penalizing the people who matter most to increasing the value of the network. Lots of defi applications become less interesting in an environment with 10% inflation and 18% stacking yield.

This is a good insight, @incite.btc. I think that if there were to be an Accepted SIP that attempts to increase the PoX APY, we would want a professional economist to model out the impact on existing and upcoming Stacks defi projects.

[xan]

I’m going to close the current PR, and open another one with a redrafted SIP. The main reasons being I feel this SIP has gotten dragged into the conversation about how to get more miners on the network. That was never the purpose. The purpose was to create a better bootstrapped environment for the Stacks network during the early days, or put another way to raise APY and security budget of the network.

In addition to clarifying what this SIP is meant to accomplish in the redraft, I also want to address some of the great feedback and include a number of the suggestions that have surfaced in this conversation with the community. One big example is I’m going to remodel the SIP to create 0% inflation in the Stacks token supply, and rather it will just accelerate the coinbase reward schedule to be more front loaded.

[jpan]

I want to give a retailer’s point of view, I have STX on PoX, I read what everyone wrote on github and here. The current volume of STX is 22M USD per day, I consider that the bearish price of the miners is insignificant, but a news such as the change in the amount or time of emission, would scare me as well as many others. I trust the community and its decisions for the common welfare. @larry is who I identify with the most.

[xan]

If you look at the data, including average daily trading volume, from when coinbase reward was 2000 per block, that fear doesn’t play out. Miners aren’t and have never been a meaningful source of sell pressure.

[yurymaksimenko]

Hi, stacks community.
I am involved since stacks 2.0 launch, doing some mining, invested in STX and a big believer in success of this platform.
It’s a great pleasure to join the discussion.

Let me point out some pros and cons that appear if this SIP is implemented:

pros:

• more miners
• higher yield for stackers

cons:

• Stacks platform reputation risks

More miners:
Current reward budget per block is ~ 1800к sats
With this SIP implemented it goes up to ~ 2880к sats which is 1080к sats increase. How many new miners can really join, considering that current miners (200к,400к,400k,400k,400k,20k) will definitely increase bids? max 3.
Does it make a difference in perception of security and decentralization if number of miners go up from 6 to 9? I doubt it.

Higher yield for stackers:
60% increase in yield is definitely a good thing for everyone holding STX, but
Amount of STX mined per day goes up from ~140к to ~ 224к which is 84к increase.
In order to keep turnover capital, miners are forced to sell all mined coins in 1-7 days (depends on the amount of capital at disposal), so we can safely assume that there is an immediate selling preassure from miners.
from my observations STX/BTC trading volume per day on binance ~250к, on okcoin ~150к (i couldn’t get precise trading volume stats from any of this platforms)
let’s assume we are having an equilibrium (50%-50%) of sellers (200к) and buyers (200к) now. Mining 84к more per day, increases trading volume to 484к on ~18% (84к/484к) , and shifting an equilibrium towards sellers (58%-42%)(284к-200к).
This can create selling preassure on STX price and drive it lower.
The key factor here is that newly mined coins constitute 35% of trading volume, it looks substantial.
Please correct me if i am missing any of trading volumes.
What if STX price goes down ~35%? it will “eat” 60% increase in yield in absolute terms.
Current sentiment is fear, if the price start to go down it can create a negative dynamic, which can drive price even lower as it happened with MIA.
If STX price goes down it will decrease reward budget per block which decrease number of miners. And we can easily get back to a point where we started in terms of number of miners, and there will be bunch of unhappy miners that lost capital during price downturn and not willing to mine again.

Stacks platform reputation risks:
As we all know blockchain community is very sensitive to protocol changes especially token emission.
This kind of emission changes most certainly will look like large STX holders trying to maximize profit from stacking their coins.
For outside investors/developers who will consider joining the community later it will just look suspicious and they will tend to avoid Stacks.
Bitcoin maxis will use it as an additional argument to bash the platform.

It looks like pros from implementing this SIP are fragile and could become cons. So we could end up creating a mess here with no benefits at all.

[muneeb]

@jude

What do you want the total security budget of the network to be (in USD), why is that the right number, and how will we know that we’ve achieved it? Also, what’s the minimum number of miners you want to see, and why is that the right number? I think we need to get concrete answers to these questions before we can reason about which strategies and tactics are the best suited to get the job done.

I don’t think it’s about picking a USD number. The total security budget of the network should be as high as possible, especially in the early days when the total security budget is less than 1 billion USD per year. Let’s call that phase the “bootstrapping phase” of the network where the budget is “low” in absolute terms. Once the network crosses 1 billion USD per year of security budget then we’re in a different phase. So the question here is that how much STX rewards per block can come out to maximize the security budget without leading to negative behavior like too much sell pressure.

The proposed modifications e.g., go to 1,400 STX vs 1,000 STX are within a “safe zone” because it leads to a total of $10M USD sell pressure at current price of $0.50 over an entire year. That is less than $1M in total sell pressure per month (assuming these miners are selling everything), this is not even a blip on the trading market which does $300M-$500M of trading activity per month.

I personally wouldn’t be concerned about even a 2,000 STX per block. As @xan mentioned above that miners have historically been a tiny fraction of sell/trading market of STX, even when we had 2,000 STX rewards. So a much smaller number than 2,000 e.g., 1,400 is well within that “safe range”. The type of concerns people are raising here become material only at much much higher values e.g., if you go to 5,000 STX reward or 10,000 STX rewards or more. The proposals here are no where close to those numbers.

RE total number of miners, we can pick a concrete numbers like at least 100 unique on-chain miners. I have a separate SIP that I’m working on which will propose a solution to that.

@xan

I feel this SIP has gotten dragged into the conversation about how to get more miners on the network. That was never the purpose.

Agreed.

Also, folks commenting on this thread should note that since this SIP is being closed there is no longer any proposal to increase the year 2050 supply of Stacks. So most of the comments here about network reputation etc are well received and those folks need to not worry about an increase in year 2050 supply. This SIP is closed

[yurymaksimenko]

@muneeb

The proposed modifications e.g., go to 1,400 STX vs 1,000 STX are within a “safe zone” because it leads to a total of $10M USD sell pressure at current price of $0.50 over an entire year. That is less than $1M in total sell pressure per month (assuming these miners are selling everything), this is not even a blip on the trading market which does $300M-$500M of trading activity per month.

Muneeb, thank you for clarifying on trading volumes, would you mind share where this amount of trading activity takes place and on which pair?

[muneeb]

You can find some data on CMC. It shows approx $20M trading volume yesterday, so can project for average monthly using more data points.

The +2% depth and -2% depth data there is also interesting and relevant for this discussion. The depth of these pairs basically means that a 60K STX sell order (average additional per day yield with 1,400 STX) at current price (meaning approx $30K sell order) would just be a blip on the trading market, and just the listed orders would absorb it. Obviously, market conditions change, the depths on these pairs change, crypto markets are volatile etc. But my broader point is that the magnitude of change discussed here is just a blip on the active trading market and we can study historical data to confirm this.