Adjusting the current inflation model to sustain Treasury inflow

[rich]

@raffael

Thank you for initiating this discussion, I think how we collectively think about and manage the Treasury will only grow in importance.

https://twitter.com/monsieurbulb/status/1676864993260720128?s=20

[Ricosworks]

Firstly, I concur with your proposal’s merit. It’s vital to deliberate treasury matters, fund allocation, and the system’s future sustainability. I am wholeheartedly in agreement that consistent funding of the treasury is essential.

Undeniably, the current inflation model could lead to a scenario where the treasury receives no funding, a situation that unquestionably needs addressing. Over the project’s lifespan, the treasury’s funding should increase. In comparison to other blockchain ecosystems I’ve studied, Polkadot’s governance and treasury system is unsurpassed.

However, a comprehensive analysis is warranted before amending the current model, to discern any undesirable consequences. Furthermore, we must ascertain whether this is an immediate priority or whether we have a 12-24 month window for consideration. For instance, come October, 100 million DOT exiting parachain slots may likely be restaked and 5 parachains will stop existing ? Or will they continue existing, ideal staked amount will necessarily diverge and hence, income of treasury will increase… Consideration must also be given to Polkadot 2.0, where not all parachains will be re-elected, enabling us to begin forecasting future staking ratios and potential treasury funding.

My emphasis is on the necessity of examining near-future network staking dynamics before deciding on your proposal.

Additionally, if this proposal progressively boosts treasury revenues, it should account for the significant number of tokens already being burnt by the treasury. Thus, predicting the volume of tokens the treasury might burn would be insightful.

It’s crucial to understand why tokens are being burnt. Given that the treasury and governance system is the most effective means of resource allocation for the Polkadot ecosystem, any burnt resources are those not invested in Polkadot’s development. And so on…

[alice_und_bob]

To inform the discussion, I aggregated the Polkadot Treasury spend of previous years in USD and DOT

year		       USD		      DOT
2020	   565,629 USD	  118,466 DOT
2021	 1,805,639 USD	   70,585 DOT
2022	16,359,027 USD	2,301,333 DOT
2023	 9,106,917 USD	1,464,989 DOT

Polkadot Treasury annual spending in USD & DOT | 2020 - 2023 June

I wanted to understand the total USD & DOT amount the Polkadot Treasury has been spending so far per year. I used the data from DoTreasury as base, which seems to use the daily rate of treasury grants around the time they were awarded - so in-time USD values, which should provide an accurate view on the value spent. To solve the task, I… fetched all Polkadot Treasury proposals with the status awarded from DoTreasury manually copied the values over to this spreadsheet processed the data with t…

[alice_und_bob]

@jonas If I read your proposal correctly, it suggests a 2.2m DOT monthly inflow to the treasury, resulting in 26.4m DOT inflow, or ~132m USD inflow at a rate of 5 DOTUSD.

If we add together the spend of Kusama and Polkadot, we arrive at 31m USD in 2022.

We can see in the previous arguments of the thread, that people believe the market could change up or down and I have my personal opinions, so we cannot really answer the question.

We have to assume that any value flowing into the treasury will eventually be spent, so we have to ask the question of how much the Treasury should reasonably spend. Personally I believe that 4x-ing the potential of the Treasury to spend right now is too much. Polkadot just transitioned to OpenGov and OpenGov should be bound to start with a similar budget like the previous governance mode.

Thus I suggest that we aim for a fixed inflow that would amount to ~30m USD (this assumes that Kusama governance will become more conservative in spending and push more costs to Polkadot governance. The amount could cover the spend of Polkadot and Kusama on the scale of 2022 spending).

Translating this into the parameters of your proposal, I propose that we set the fixed yearly inflation into the treasury to 0.5% instead of 2% (6.6m DOT/33m USD per year). This is a more conservative approach and could be changed by governance later as we gather more data.

[Ricosworks]

doTreasury | Empower treasury with credit building 465K latest burn, increases as treasury size increases apparently.

[Ricosworks]

Adjusting the current inflation model to sustain Treasury Inflow – A wrap-up

(below is my personal understanding of where the discussion stands as well as my personal view on how to continue the debate)

Recipients: @jonas @alice_und_bob @alistairStewart @joepetrowski @taqtiqa-mark @Olanod @lolmcshizz @rich @guinnessstache @raffael

Summary:

Issue Identified by Jonas

Jonas identified a potential risk with the existing Treasury funding mechanism, which could result in zero inflow to the Treasury.

Recommended Solution by Jonas

Jonas proposes to adjust the current model to divert 20% of inflation to the Treasury, replacing the current floating parameter. This proposal was informed by the situation in Kusama, where the Treasury now receives no inflow. The aim of this proposed change is to ensure a steady revenue for the Treasury, not necessarily to augment it.

Observations and Suggestions Regarding Treasury Funding Mechanism Adjustment by discussion participants

1. It is advisable to determine the actual treasury run rate before proceeding with changes to the funding mechanism
2. Polkadot 2.0 might address the identified issue through its ‘pay for block’ mechanism.
3. Consider a comprehensive re-evaluation of the mechanism: is funding via inflation necessary
4. Inflation ought to be directed towards expanding the validator set to ensure additional parallel cores.
5. Recommendation to trial the proposed analysis on Kusama prior to any implementation on Polkadot.
6. How many tokens are being incinerated? Detailed information about treasury activities can be accessed here: doTreasury | Empower treasury with credit building
7. Strive for equilibrium between inflow and outflow, whilst retaining the current token-burning procedure.
8. Coretime income is likely to increase gradually; the rate of increase is uncertain. In contrast, income from inflation is more predictable.
9. The aim is for the Treasury to have a more consistent income source.
10. The stated problem is not guaranteed. The true challenge lies in determining desirable inflow and outflow during a market cycle, prior to any system parameter alterations.
11. The Treasury currently has a surplus of DOTs and a dearth of projects.
12. Maybe the question should be how to allocate treasury resources more efficiently.

Annual Analysis of DOT/USD Expenditure by the Treasury and Proposal

The Treasury’s outputs, income, and value have seen consistent growth, with the income declining and the output increasing as correctly pointed out by Jonas. In USD terms, however, except for spending, they have been falling since November 2021 when the DOT price peaked at $54. According to dottreasury.com, there were four proposals in November 2021 amounting to a $184K spend, which was negligible compared to available Treasury funds. By June 2023, nine proposals amounted to $444K in spending. This looks indeed like a really bad situation to be in for a treasury.

In USD terms, the Treasury’s spending has increased tenfold since November 2021. Back then, it spent 0.02% of its reserves in a month, while in June 2023 it spent 0.2%. There were in-between to important spending in January 2022 and July 2022, but we disregarded these for the purpose of the exercise, as the spending was mostly bounties, not proposals. Based on a calculated annual growth rate of 667% on spending, by December 2023 the Treasury should be spending around 0.65% of its total available funds monthly, on proposals, rising to about 3.5% by the end of 2024, assuming current trends continue.

Jonas has a point and I would add that in my opinion, relying solely on a fluctuating currency for Treasury funds is risky. The Treasury should diversify its holdings to include ETH, BTC (up to 5% total), and stablecoins determined by the projected 3 years of necessary spending for example. This would ensure the Treasury is diversified and can cover real-world costs, such as developer salaries, without depleting its DOT reserves. Decisions regarding liquidity creation and asset rebalancing should be made logically, considering not just the Treasury’s needs but also market cycles.

Treasury funds are tokens with intrinsic value and are subject to supply-demand dynamics. Increasing or stabilizing Treasury income unnecessarily could lead to wastage and distract the community from focusing on real-world use cases and not necessarily bring additional real funds in the treasury.

Polkadot’s governance mechanism outpaces others such as Ethereum, Avax, Cosmos, and Polygon. The key is prudent Treasury management, including clear guidelines for the use of funds and the appointment of responsible proposal managers. Mismanagement could lead to the unnecessary burning of valuable resources, impeding the growth of the Polkadot ecosystem.

[jonas]

The 2% of total inflation is debatable but I chose it because it is the average of the whole history of Polkadot since it’s inception. It comes from the fact that the Treasury right now appears healthy and that it was able to fund projects properly through all the previous market cycles. In a sense, it’s therefore the only “guess” that is backed up by something real. Adjusting it based on price predictions for the future is adding more uncertainty.

An overflow of the Treasury is prevented by having the burn mechanism, which exactly counters potentially too much balance. What I haven’t seen much in the discussion is that burning is actually beneficial for everybody in the ecosystem, because it “pays back” to each token holder by making each DOT more valuable. So, I think it’s better to have too much DOT in the Treasury than too little (especially since staker rewards seem decent in this configuration). But a bigger balance in the Treasury should not urge people to spend it for unnecessary things. I feel, however, that managing the outflow is something that has become a spotlight for the community recently and that there are many good ideas around to make the spending more meaningful.

[alice_und_bob]

If I counted correctly, the total Treasury outflow in a bit over 4 years has been a bit under 4m DOT. The Treasury still has 46m DOT / 230m USD in reserves. That is 20 years of reserves at the 2022-outflow rate of 2.3m DOT per year.

Sure, the Treasury can increase the spending, but looking at the current discussions shows that accountability is a big issue atm and there is a human/social bottleneck to efficiently allocate Treasury outflow. So I just don’t see us in running into the problem of the Treasury getting drained of all DOT in the near term.

And we still have a degree of uncertainty on how inflow will change from other factors like coretime sales.

Setting up the params for fixed Treasury inflow is important as a general lever for governance, but we should not pump 26.4m DOT (that is inflation, mind you) into the Treasury per year if the total spend last year was 2.3m DOT.

[Ricosworks]

@jonas

it for unnecessary things. I feel, however, that managing the outflow is something that has become a spotlight for the community recently and that there are many good ideas around to make the spending more meaningfu

I agree and I would say that it’s early to start putting time into solving this “future problem” - maybe review it in end 2024 would be my take.

[Ricosworks]

Totally agreed and aligned, and in my opinion, it’s crucial to consider the growth factor in our calculations, given the historical evidence of expansion. Assuming continued adoption, this growth is expected to accelerate, not merely progress linearly.

[lolmcshizz]

We should wait until July has passed before deciding on what’s the current outflow rate ser - I wouldn’t be surprised to see July surpass the YTD spend.

[rich]

Plan for the worst, hope for the best.

I think it would be worth outlining a bunch of assumptions and modelling a bunch of scenarios rather than just postponing until a potential problem becomes an actual problem, which is then when fate dictates proceedings.

Some factors to consider on the skeptics side

1. Coretime income Everyone is excited about coretime and Polkadot 2.0. There is no guarantee this new positioning, mental model and narrative will result in the demand people hope for. As @AlistairStewart says, the income even if burned may not scale fast.

2. Coretime / supercomputer narrative is not imo a defensible narrative when considering that it isn’t that differentiated to the sort of “blockchain computers” and blockspace product narratives that have been around for a couple of years. This is not saying the positioning doesn’t make sense, it’s just going to be an uphill battle making it cut through when it can more easily accrue to ETH etc - in the end the best marketing is adoption.

3. Regulations - they are going to squeeze crypto for the next few years, the frenzies of BTC forks, ICOs, L1 wars, defi summers and NFT / celebrity ponzis aren’t coming back in a hurry. This is not just another market cycle, conditions have changed so lots of assumptions about this being the bottom of the market can well be wrong.

4. Spending power Treasuries are magicked into existence and their spending power is fickle - you cannot simply look at this DOT and say “there’s $230m which at current run rate gives us 20 years”, it doesn’t work like that.

5. Funding developments The quandary is that to actually bootstrap the growth needed to sustain the treasury value, and it’s spending power, requires spending down the treasury - it’s a devils pact, that if managed poorly will deliver whiplash corrections that will have huge psychological impacts on holders, voters and proposers, which then further impacts the ability to plan.

6. Time - it might feel like there’s lots of time to consider this stuff and maybe come back in a year or so, but the whole point of planning is to be ahead of the curve, not behind it. There is $25Bn of (perceived) value locked up in DAOs - much of which will never spent (arguments, regulation) or spent without any strategic foundations on fluffing token prices and getting general awareness without any real onchain utility

[jonas]

You should also include the outflow from burns to your calculation. Doing so means that the Polkadot Treasury had an outflow of 16m DOT in total up until now. In addition, I don’t think the early time (even 1-2y?) are very representative for the outflow.

I am currently arguing from an angle of keeping total inflation at 10%. That makes sense to me, because a fixed Treasury income does not need to lead to a change in total inflation. And I’d rather not mix the two, to make it hard for us as community to find a consensus. That being said, we might want to have another discussion in the future about lowering total inflation.

With the parameters that I suggested, staker APY would be around 18%. That is (especially for nominators) plenty (and as I said, probably too high). Diverting less to Treasury would mean, sticking to 10% inflation, giving it to stakers. That seems unnecessary to me.

I’d rather see inflation diverted to Treasury and eventually burned than potentially having an underfunded Treasury and boosting staking APY even further.

[asteeber]

This is an intriguing idea, however there is one core topic that is lacking in this discussion: security. I find it very ironic that when I suggested staking Treasury funds the ecosystem cried about it undermining the underlying assumptions of proof-of-stake, yet no one here seems to understand that this effectively does the exact same thing as staking Treasury funds:

• By staking Treasury funds the Treasury isn’t penalized in the case of a slash because it is the beneficiary of slashes - so staking Treasury funds gives the Treasury a slash free reward.
• By allocating a fixed percentage of inflation to the Treasury gives it a slash free reward.
• In both cases the Treasury is taxing the entire proof-of-stake ecosystem, the only difference is that staking Treasury funds gives the Treasury a revenue proportional to how much it bonds whereas a flat allocation give the Treasury an arbitrary revenue.

You mention,

@jonas

Therefore, adopting these parameters should have a negligible impact on the ecosystem’s actors, as it merely formalizes the historical divergence between the Treasury and stakers.

This isn’t true because the divergence will still exist and be stacked on top of this ‘fomalization.’ The only difference is that it’ll be a divergence that splits 80% of the inflation instead of 100%. Stakers are still penalized for not achieving the ideal staking ratio and, on top of that, are taxed so the Treasury can receive its arbitrarily ordained revenue.

Let’s explore how ‘negligible’ this tax really is using recent figures in Polkadot:

• Last era reward (R) = 313,899 DOT
• Total validators (V) = 297
• Average staked per validator (S) = 1,951,200 DOT
• Eras per year = 365
• Assume 1% commission
• APY without 20% inflation allocation = (R / V * 0.99 * 365) / S = 19.573%
• APY with 20% inflation allocation = (R * 0.8 / V * 0.99 * 365) / S = 15.817%

As expected, this will decrease everyone’s staking yield by 20%, or 3.7564 percent points. To say this huge tax wouldn’t impact actors in the ecosystem is frankly outrageous. This decrease will certainly cause some actors to unbond their DOT to find better staking opportunities, thus undermining the security of the network while also exasperating the divergence from the ideal staking rate.

Don’t get me wrong, I am not against the idea of using inflation to boost Treasury Revenue. I mean, I was the one who suggested staking Treasury funds. All I’m trying to point out is that this undermines network security too, perhaps in a worse way because it is such a heavy tax.

My suggestions

I think that we either:

• simply stake Treasury funds via governance managed nomination pools or via a number of staking-proxy multisigs managed by community councils. There is no reason we should subject the staking ecosystem to an arbitrary tax. Instead, I think it’s only fair that the amount of DOT the Treasury bonds becomes the governing variable to how much revenue it can extract from inflation. The downside is that this will not produce as much revenue for the Treasury. OR,
• increase the inflation rate from 10% to 12% and do a 83.3-16.7 split. This will give the Treasury the same revenue you mentioned while leaving nominal staking returns unaffected. This obviously has the effect of quickening the dilution of DOT, but that effect is spread out to the entire ecosystem rather than it being a tax for staking actors only. If we are going to create revenue for the Treasury we should make speculators pay for it too, not just the staking ecosystem.

Respectfully,

Adam

P.S. if you’re wondering why we came to different figures for return on stake it’s because your yield curve doesn’t take into account that inflation will still be 10% not 8%. Plus your change fundamentally changes the formulas; you cannot decouple maximum staker inflation from the total inflation, give the difference to the treasury, and calculate returns as if the difference doesn’t count toward inflation. This decoupling completely changes the tokenomics model. IMO, the tokenomics model probably shouldn’t use ideal interest rate as a variable since it is an implicit variable that depends on the model as a whole. When calculating staking returns in a complex inflation model, it’s best to go back to first principles.

[Ricosworks]

As we all know, treasury funds, which are exclusively in DOT, can depreciate rapidly. If, for example, someone utilizes 1M DOT from the treasury and sells it, it will cause the DOT’s price to fluctuate, subsequently influencing the treasury’s worth. Thus, an expenditure of 1M DOT at $5 not only signifies an allocation of $5M but also an indirect dilution of the treasury through percentage points times the number of DOTs in the treasury. Consider a 1% decrease in the DOT price after a 1M DOT sale. The total expenditure would then be $5M plus 1% of 230M DOTs, equaling $2.7M and $5.7M total expenditure. You can observe DAO treasuries and their USD equivalents/other tokens on Defillama’s treasury dashboard https://defillama.com/treasuries. Would anyone like to share their views on this? Is this issue being discussed and handled already ?

[DSP]

At face value, this proposal seems to incentivize Treasury Funding over Network Security, which might be a slippery slope path to approach. Impacting nominators/validators might result in the depreciation of Treasury funding via USD price correlation (stakers selling and leaving the network), leaving both the Treasury underfunded and the network technically less secure (via less total amount staked, in both DOT and $$ value)

I think more data needs to be analyzed before being able to move forward with this proposal. If the concern is the depletion of the Treasury, does anyone have a reasonable forecast of Treasury spend for the next X months?

Based on the last 6 months of data, we have an approx 5M DOT being ‘removed’ from Treasury (‘removed’ = spent + burnt), with a total remaining of 46M DOT. Based on what dataset are we expecting depletion to occur within the next 12-to-24 months (timeframe which would allow to understand network behavior within the Coretime model)?

[rich]

@Ricosworks

As we all know, treasury funds, which are exclusively in DOT, can depreciate rapidly. If, for example, someone utilizes 1M DOT from the treasury and sells it, it will cause the DOT’s price to fluctuate, subsequently influencing the treasury’s worth.

Yup. This is one of the primary reasons for this work Introducing an alternative financing strategy, structure and partnership for maximising the potency of Kusama's treasury

[lucasvo]

@alice_und_bob

We have to assume that any value flowing into the treasury will eventually be spent, so we have to ask the question of how much the Treasury should reasonably spend. Personally I believe that 4x-ing the potential of the Treasury to spend right now is too much. Polkadot just transitioned to OpenGov and OpenGov should be bound to start with a similar budget like the previous governance mode.

I don’t agree with this. Here’s how you could frame treasury spending: it’s basically Polkadot’s R&D, ecosystem development and marketing budget. Parity and Web3.0 Foundation of course have budgets of their own but we don’t want Polkadot to be entirely dependent on two centralized entities controlling and funding all development.

From that perspective I think a 2% DOT issuance per year that becomes available to the treasury is not a bad idea. Polkadot is still an early stage technology and needs significant R&D. Otherwise it will just get left behind by other projects willing to spend on development.

It will be easier to spend once it’s there so we can expect to spend more but in the end DOT in the Treasury != DOT spent and this will always depend on voters.

DOT staking is still significantly more attractive than most PoS networks and as such reducing the ideal staking rate (an expense) and allocating to investing in assets (technology, ecosystem etc.) I think is a good idea.

I agree with what @jonas says here:

With the parameters that I suggested, staker APY would be around 18%. That is (especially for nominators) plenty (and as I said, probably too high). Diverting less to Treasury would mean, sticking to 10% inflation, giving it to stakers. That seems unnecessary to me.

[rich]

@lucasvo

Here’s how you could frame treasury spending: it’s basically Polkadot’s R&D, ecosystem development and marketing budget.

Perhaps even simpler than that, the treasury exists to bootstrap demand for coretime and all spending can be assessed through that lens. At some point spending on R&D that extends say the scalability of the systems, in the absence of demand, becomes wasteful.

[Ricosworks]

In two words @Rich, “I agree”.

In the blockchain landscape, we often gravitate towards the promise of decentralised freedom - Bitcoin exemplifies this. Despite its seemingly archaic tech, its immutability has proven to be a strength, establishing a gold standard of trust.

However, the blockchain ecosystem thrives on diversity, and Polkadot is a contender with its potential for interoperability and scalability, thanks to its unique architecture. We’re tasked with assessing whether Polkadot can provide the same degree of freedom without succumbing to societal challenges seen on other platforms.

Another crucial aspect to consider is the potential for rapid ROI in this sector. In this sense, Polkadot aims to democratise blockchain usage, providing a user-friendly ecosystem. Ethereum demonstrated this with tools like MetaMask, Uniswap, and DeFi platforms, despite the proliferation of scam tokens.

At this stage, Polkadot strikes me as the ‘Apple’ of crypto, Ethereum the ‘Microsoft’, while Bitcoin looms as an unpredictable but sturdy ‘elder’. To secure its place, Polkadot needs to determine the unique demands it fulfills better than others, making those services accessible to the broadest audience possible. This could pave the way for Polkadot potentially surpassing Ethereum and even Bitcoin.

To draw a Star Wars analogy, we’re on Dr. Wood’s Millennium Falcon, heading for an uncontrolled journey. In a sophisticated tech environment, missteps could lead to crashes, ending the adventure prematurely. The Polkadot community is robust, filled with brilliant minds, not surpassing Ethereum but remarkable nonetheless. However, we may be missing a ‘Han Solo’, a leader to guide us. The challenge now is to find this figure, navigating us towards a prosperous future for Polkadot.