Musings on the Pond
Special thanks to those who encouraged me to publish this and those who helped with review and feedback — Patrick, Krby, SilentRhetoric, D13 and Stitch
In a pond called Algorand, quiet and still,
Where waters recede, a test of will.
Little rain to ease its thirsting land,
Just gentle ripples, a tranquil stand.
Years of waste, in shadows cast,
Toxic remnants of a troubled past.
Life struggles in its murky embrace,
A world unkind, a challenging space.
A solitary reign, one species’ creed,
Hoards the sustenance all others need.
First to feast, while others wait,
In this ecosystem, they dictate fate.
Yet, in this pond, some choose to stay,
Their home it is, come what may.
A vision of Algorand, resilient and strong,
A place of potential, where many belong.
For home is home, in joy or in sorrow,
Hoping for rain, maybe tomorrow.
Striking a Balance
A protocol’s economic sustainability hinges on the balance between costs (expenses incurred on behalf of the protocol) and revenues (txn fees collected in the fee sink). With transaction fees small and insufficient to cover expenses, Algorand’s treasury is depleting at an alarming rate. If we ever want the community to pick up these responsibilities and costs, there needs to be an incentive to do so.
The current revenue, approximately $5k per day, falls far short of the expenses, estimated at around $350k per day (total protocol spend). The worst part is, all this spending is not being reflected in the market. A sustainable model that ensures long-term viability and price appreciation is essential, especially as the distribution of all 10 billion ALGO approaches its end.
Securing the Castle
Ensuring network security is paramount. Algorand faces multiple vulnerabilities due to the concentration of ALGO staked, and the low token price. A stakeholder with 22% of the online ALGO could potentially halt the chain, and currently with ~ 1.4 billion ALGO online, this attack would require ~ 300m ALGO ($49.5M) if coming from existing stake, or ~ 395m ALGO ($65.2m) if coming from “outside / new” stake.
To achieve true decentralization and withstand nation-state-level threats, Algorand must encourage far more node operators while simultaneously reducing the reliance on any one single entity for operations. Implementing a reliable and indefinite reward mechanism for node operators is a step towards incentivizing greater, more reliable, and more resilient network security.
Substantial, Efficient and Effective Incentives
Algorand’s funding mechanism for public goods lacks efficiency and effectiveness, resulting in significant treasury waste over time.
Programs like Xgov and Algorand Ventures have severe limitations and aren’t equipped to adequately support builders, creators, and individuals who contribute public goods to the ecosystem. Excluding Ventures, Xgov is only allocated a max of 8,000,000 ALGO ($1.28M) per year. This is meant to somehow sustain an entire ecosystems’ infrastructure, builders, and creators.
Builders face immense challenges due to insufficient funding available on Algorand. ALGO is currently performing so poorly in the markets that finding new VCs to come into the space is nearly impossible. The only possible way for an Algorand company to grow is to leave Algorand.
The Xgovs often have to make decisions with limited information, relying solely on a proposal, it can be impossible to determine the technical feasibility or identify potential flaws in a project. Additionally, assessing the necessity, value, and likelihood of project completion (as evidenced by the history of past grant recipients) involves a high degree of uncertainty and personal bias.
The system simply isn’t making a difference for the builders and creators in the ecosystem. It’s time we start valuing people and projects for the impact and value delivered, not promised. We need an efficient, credibly neutral mechanism to do this and the best I’ve seen is OP’s Retroactive Public Goods Funding. We need to learn from how other communities do things, and I think giving something like RPGF a chance could be game changing for our ecosystem.
The Core Issues:
The protocol’s economics are unsustainable
The protocol’s security is vulnerable
The protocols funding mechanisms are broken
A Holistic Proposal for Change
To address these challenges, I present the following changes:
- Halt current ALGO distributions in various programs — Governance, Xgov, Defi Rewards and NFT Rewards
- Divide remaining ALGO into three buckets, and allocate future funds from the fee sink to the following three buckets:
- Node Incentives — to encourage securing the network
- Retroactive Public Goods Funding (RPGF)— to encourage the creation of public goods
- Algorand Foundation Operations — to allow AF to continue operations and ensure a safe hand-off of responsibilities to the community through RPGF (moving too fast can be equally as dangerous)
- Adjust transaction fees up from .001 -> .01
- Uncap Algorand’s supply — Add programmatic issuance of ALGO per block for validators of .25 ALGO per block (mining)
- Add Burn mechanic per transaction — Set dynamically at 50% of the txn fee (.005 without congestion), a block with 51 txns or more would be net deflationary (50 txns *.005 ALGO = .25 ALGO burnt = new issuance)
Example Block
200 txns
= 2 ALGO in total txn fees (.01*200)
50% is burned = 1 ALGO
The 1 ALGO remaining is split between
the fee sink (.5) and validator (.5).25 ALGO is issued to the validator from the protocol
The validator would also get an incentive from the Node Incentives program (ranging up to 35 ALGO in the first year)
The validator would end up making ~35.75 ALGO from processing a single block in the first year with 200 txns
The fee sink would generate .5 ALGO from the block
The protocol would issue .25 new ALGO and burn 1 ALGO making total supply net deflationary (-.75 ALGO)
All of these parameters could be adjusted over-time through governance
Other Changes
- Switch to snapshot based governance based on users online stake (must be participating in consensus to have a vote, no lockup needed, voting available inside wallets)
- Allow the RPGF collective to put forward protocol-level governance proposals in general governance, along with Algorand Foundation
- Remove the opt-in requirement, allow it as an account parameter that users can switch on (users can opt-in to opt-in) — This can be handled in the app layer but I see no reason for not putting it in the protocol.
Important Note: These are proposed changes. Changes to the protocol would require the communities approval, then the new code to be written, and then a 90% supermajority of node runners (by stake) to approve.
Fee Sink Distribution
The fee sink should be programmatically set to distribute as such:
45% Node Incentives Pool
45% Retroactive Public Goods Funding Pool
10% Algorand Foundation Operations Pool
This acts to rejuvenate the pools as they’re being spent from, allowing additional epochs to occur for node incentives and RPGF after the scheduled ones end in 5 years.
Node Incentives
Proposed ALGO Total: 750M (7.5%)
Years Emitted: 5
Emission Type: Linearly descending fixed amount per year
- The single best way to ensure network security is to incentivize it
- The incentives (on top of mining rewards & txn fees) will create lucrative opportunities for individuals and companies to run nodes
- Services can offer staking services, paid nodes, etc. to individuals
- Yields in defi will be less choppy and more predictable, and underpinned by real protocol yield + scheduled incentives
- Staked tokens can act as the basis for other defi opportunities
Retroactive Public Goods Funding (RPGF)
Proposed ALGO Total: 750M (7.5%)
Years Emitted: 5, with 3 month Epochs
Emission Type: Linear
ALGO Amount Per Epoch: 37.5M
RPGF rests on the assumption that it’s easier to agree when something was useful compared to if something will be useful.
Proposed Process for RPGF:
Create an initial ‘Results Oracle’ (a DAO) consisting of
- ~30 highly engaged community members
- ~70 highly engaged builders / creators
- ~15 Algorand Foundation members
Give each member a token with clawback and freeze — this is their membership to the Results Oracle and what gives them the ability to vote in each RPGF Epoch.
Every round, allow current RPGF members to elect at least 3 new community members and at least 7 new builders / creators to the Results Oracle. People will be free to leave / not vote at anytime, and their spot will be forfeited. The spot will be open to fill in the following round. Allow RPGF members to vote to kick members. Reasons for being kicked could be— dishonest voting, corruption, abuse of power, etc.
Anyone can signal their desire to be a member on-chain — possibly just by calling a smart contract to toggle interest on/off, or utilizing a user defined metadata field in their NFD. RPGF members query the interest endpoint to get their list of potential nominees for the next session with all their social media / Algorand accounts connected and easily verifiable.
Members assemble a list of active projects / individuals to vote on. Anyone can be nominated (on-chain) to receive RPGF by sending a note transaction to the DAO wallet with the nominees NFD or wallet address and name — Nominations should be nearly effortless.
Each Results Oracle member goes down the nominations list, scoring each nominee 0–100 on the impact that project / individual has had in the ecosystem in the previous 3 months. The first RPGF epoch shouldn’t be limited to the previous 3 months and should take into account the entirety of the projects impact in the ecosystem. However, if the current project isn’t live on Algorand mainnet, it isn’t eligible to receive RPGF.
Results Oracle members can self-divide into specialists groups and delegate between each other (gamers might know the most about gaming projects, while traders might know the most about defi). Members can optionally self-divide into specialties to help be more effective in distributing funds. Members and the broader community are responsible for holding active members accountable for not abusing their powers. For this reason, all voting needs to be made publicly auditable.
Tally totals for each nominee and average the totals to find their allocation of the epoch's funding.
Cap the total any one project / individual can receive to 5% per epoch. If the same individual is contributing to multiple different projects, then it’s okay for the combined total of their projects to be greater than 5%, as long as each project is making a distinct impact and delivering value. This percentage is another parameter that can be adjusted iteratively.
Use a smart contract to tally the votes, enforce the caps, and send the grant funding payments (remove the need to trust anyone for administration).
The only responsibility of the Results Oracle is to produce an averaged impact score for each eligible nominee every epoch. Everything else should be handled programmatically. RPGF fundamentally relies on the assumption that voting on what was useful is easier than voting on what will be useful. By using an averaged impact score we can try to further remove bias and complexity from the process. The larger the Results Oracle group becomes the more accurate the judgement of impact should be (given honest actors are nominated). Additionally, the Results Oracle can come up with criteria to help quantify impact based on a certain projects category, to ensure impact and ‘real’ metrics are not way out of balance. This should be an iterative process, looking to improve in efficiency and effectiveness each epoch.
This process gives funding to good operators who create impact and value without them needing to ask for it. This allows productive teams to focus on building, growing, and continuing to create value.
With a process like this, no one will be able to show up and receive free money, and those that deserve the most will receive the most. It creates an incentive for builders to put products on mainnet as fast as possible and to start creating value / impact. There will be no more ‘if I get a grant I’ll build it’ mindset. Builders will need to be so confident in their ideas they either bootstrap it (take personal risk, OMG), or if the idea has commercial viability they find someone else's capital to risk (VCs).
To those who say we don’t have evidence of this working, I present you the OP token chart, an ecosystem that has used RPGF since inception.
More details on retroactive public goods funding:
https://otherinter.net/research/positive-sum-worlds/
https://medium.com/ethereum-optimism/retroactive-public-goods-funding-33c9b7d00f0c
https://app.optimism.io/retropgf
Algorand Foundation Operations Budget
Proposed ALGO Total: 500M (5%)
The AF needs a set budget for operations that they’re currently responsible for on behalf of the protocol. This includes salaries, business development, ecosystem support, marketing, R&D, education, platform infrastructure and tooling. It would be irresponsible to propose taking all of the AF holdings for the community, although I think the balance needs to be a lot more fair. These expenditures can slowly be handed from the AF to the community through RPGF and governance.
This SHOULD be seen as an inefficiency tax on the protocol. We (the community) are paying operators to do work we could be doing ourselves. Or even better, we could design credibly neutral, maximally decentralized mechanisms to do them for us.
In the meantime, AF will have 500M ALGO, $85M USD, and a 10% stream from the fee sink. This should allow them the time needed to handoff all responsibilities to RPGF and general governance.
Assumptions Made
Algorand Foundation Treasury as of Mar 31, 2023
-$40.5M USD & -202M ALGO (202m ALGO * .23 = $46.5M USD)
Total Net Outflow over Previous 6 Months = -87M ($43.5 / qtr)
Algorand Foundation Treasury as of Jun. 30, 2023
-$14.1M USD & -249M ALGO (249m ALGO * $.12 = $29.9M USD)
Total Net Outflows over Previous 3 Months = -$44M
Algorand Foundation Treasury as of Sep. 30, 2023
-$10.9M USD & -433.5M ALGO (433.5m ALGO * $.10 = $43.3M USD)
Total Net Outflow over Previous 3 Months = -$54.2M
We should get the next transparency report soon (1 month late now) — but for the purposes of this proposal I will be using my best guess as to the current treasury balance and current quarters flows. I can update my proposal with current numbers when they’re released.
I’m guessing (and hoping) we see a big cut in spending this quarter. While I think seeing a 50% drop is too optimistic, I will go with 40% as a ‘realistic’ guess. With this assumption, we’re looking at projected net outflows of ~$32m for the Oct — Dec quarter, leaving the balance sheet something like:
2,000,000,000 ALGO (assuming 108m ALGO outflow) ($17.8m USD)
$84.9M USD (assuming $14.2m outflow)
$32M burn per quarter * 4 quarters = $128M / year
$128M burn / 365 days = ~$350k / day burn 🤯
I hope I am way off on my assumptions which would make the AF’s financial situation far less dire. It is also unclear how much USD the AF treasury holds as they only report ALGO and Investments (in USD) on the balance sheet. With an incomplete view of treasury assets, I can’t judge runway.
Transparency reports: https://www.algorand.foundation/transparency
Conclusion
Algorand must change its monetary policy to ensure sustainable funding for both the network and public goods well into the future. These two things are where I argue 90% of the remaining ALGO should go, and where 90% of future fees should go.
Proposed changes, such as increasing transaction fees, removing the cap, introducing a burn mechanism, and reallocating fees between the fee sink and validators are steps toward ensuring we have an economically sound and dependable network. These changes also introduce a deflationary aspect while providing a sustainable yield for validators, which is vital for token appreciation, and the network’s economic security and longevity.
Large ALGO distributions like Governance, Xgov, Targeted DeFi Rewards and NFT Rewards have not met their goals of attracting new users or increasing TVL / volume. I believe these initiatives were misguided from the start — and it’s time to put all our focus into what matters most.
The goal is to be a decentralized network that’s secure, censorship resistant and permissionless, with substantial amounts of funding, and sound underlying economic principles that lead to increased trust, adoption, and price appreciation.
With these changes Algorand will share many similarities to Ethereum, but will be far more efficient and scalable. I think we need to lean into that as our sales pitch to new users and users who don’t like the current Algorand. If capped-supply gold is sound, and deflationary eth is ultra-sound, then highly-efficient deflationary algo is highly-ultra-sound.🦇🔊
Written with love,
towards peace,
fisherman.algo🎣
All images were generated with DALL E
