Triweekly parameter committee meeting - Oct 2, 2025
These are the minutes of the PCs (Parameter Committee) triweekly meeting. This is a summary based on automatically generated notes and may therefore contain inaccuracies or errors.
The purpose of the Parameter Committee is to provide technical advice and recommendations relating to protocol parameters. It discusses all parameters relating to the Cardano protocol including network, technical, economic and governance parameters.
Summary
The meeting focused on the urgency of implementing the “min margin” parameter, with Carlos Lopez de Lara presenting an extensive five-year review of Shelly’s incentive design. The discussion included the number of active and abandoned stake pools, the cost and sustainability of running a stake pool, and the influence of large entities and pledge on the Cardano ecosystem. Key talking points included the proposal of “delegation address mobility” and “Virtual Stake Pool Operators (VPOS)” by Neil Davies to address challenges faced by small stake pools, and concerns about the “linear layoffs” design and CIP-50’s pledge leverage.
Details
- Introductions and Meeting Agenda The meeting began with introductions, including new attendee Bjarne, an SPO of Techpool and part of Tingvald, who joined to observe the process. Alexander Moser outlined the agenda, which included an update on the parameter update for “prus primitives” and a discussion on the “min margin” parameter. Kevin Hammond was requested to post a rationale file for the governance action, which he indicated he would put online in Dropbox.
- Min Margin Parameter Urgency Alexander Moser highlighted the critical need for Carlos Lopez de Lara’s insights on the min margin parameter, a new parameter that requires a decision on its urgency for Mainnet implementation. There are two proposed methods for implementation: an intra-era hard fork in Q1 for hardcoding the parameter, or a proper era-changing hard fork in Dystra (Q2/Q3 next year) to add the parameter with guardrails. The discussion focused on the urgency of implementing min margin.
- Analysis of Shelly’s Incentive Design Carlos Lopez de Lara presented findings from a five-year review of Shelly’s incentive design, focusing on on-chain data analysis to understand the behavior of the stake pool landscape. A key observation was that the number of stake pools, approximately 3,000 active SPOs, far exceeds the desired number of 500, according to the ‘K’ parameter. Neil Davies noted a discrepancy in the number of active SPO certificates, stating there were 4,500, while Carlos Lopez de Lara believed it was around 3,000.
- Abandoned Stake Pools Carlos Lopez de Lara identified a significant number of “abandoned” stake pools, estimating around 1,000, based on criteria such as no updates or block production in the last year, minimal ADA delegation, and lack of governance participation. These pools inflate the total count of small pools, though they collectively hold a tiny portion of the stake (around 30 million ADA), posing no security issue. Kevin Hammond suggested that these pools might be dormant, not worth the trouble of reclaiming the deposit, or hoping to reactivate.
- Cost of Running a Stake Pool Carlos Lopez de Lara conducted an investigation into the cost of running a stake pool, estimating it to be around $600 USD per month for three machines based on cloud services like AWS and Google Cloud. Jonathan Kelly challenged this, suggesting cheaper alternatives like Netcup could reduce costs to about $150 per month, and noted that labor is also a major cost. Carlos Lopez de Lara acknowledged that smaller SPOs reported costs between $300 and $400.
- Sustainability of Small Pools Carlos Lopez de Lara’s analysis indicated that about 300 to 400 small pools are losing money, but a good portion are breaking even. He aimed to determine the ADA needed per epoch for sustainability, estimating around 150 ADA per month, which translates to about 3 million ADA in delegation to consistently produce blocks. Carlos Lopez de Lara observed that smaller pools with less than 3 million ADA delegation tend to set higher pool costs (e.g., 340 ADA).
- Healthy vs. Struggling Stake Pools Carlos Lopez de Lara’s report categorized stake pools, identifying 741 “healthy” stake pools with over 3 million ADA in delegation, controlling 21 billion ADA. He also defined “struggling” pools as those actively producing blocks but not breaking even, and “inactive” pools as potentially abandoned. Kevin Hammond questioned the community’s desired number of supported pools, emphasizing its importance.
- Decentralization and ‘K’ Parameter Neil Davies clarified that the original ‘K’ parameter was set to achieve decentralization with around 100 pools, potentially growing to 250, suggesting that with current numbers (around 1,000 active pools), the system is “over decentralized” from a mathematical perspective. Jonathan Kelly noted that this assumes each stake pool is an independent entity, which is often not the case.
- Role of Large Entities and Stake Distribution Carlos Lopez de Lara showed that about 20% of the network stake is controlled by large entities like exchanges (e.g., Binance, eToro, Coinbase), which he referred to as “proxy stake”. Neil Davies explained that exchanges prefer to run their own nodes to ensure returns promised to customers. Carlos Lopez de Lara also presented the distribution of stake, highlighting that the vast majority of wallets (over 1 million) have less than 1,000 ADA, while only 2,200 wallets hold over 1 million ADA, controlling 14 billion ADA.
- Community Importance of Small SPOs Carlos Lopez de Lara emphasized that smaller SPOs are vital community members, actively participating in conferences, discussions, and promotion of Cardano, even if they don’t hold significant stake. He suggested that solutions might involve skewing things in favor of smaller entities or implementing a progressive tax system to redistribute income to them. Neil Davies summarized that from a security perspective, more stake pools are not needed, but there’s a need to address the economic sustainability of smaller SPOs.
- Min Margin and Min Pool Cost Relationship Carlos Lopez de Lara concluded that introducing min margin should ideally be paired with reducing min pool cost, potentially to zero, to create a more free market. Neil Davies clarified that this implies introducing min margin as a modifiable parameter, meaning it would have to wait until next year’s hard fork. Carlos Lopez de Lara stressed the risk of hardcoding a parameter level before it is modifiable, as it could compromise the entire hard fork if not well-received.
- Report on Intercept Milestone Carlos Lopez de Lara is under pressure to finalize a report by today or tomorrow, as it is an intercept milestone and his manager will be bothering him until it’s ready. He mentioned using ChatGPT only for styling and orthography, not for generating content, due to its tendency to make wrong conclusions. Carlos Lopez de Lara found it very useful to discuss the report with Neil Davies and Alexander Moser, as it provided inspiration.
- Plutus Script Closure Kevin Hammond inquired about the closure of the Plutus script, which Neil Davies and Alexander Moser confirmed was completed before Kevin Hammond’s arrival.
- Rationale Review Alexander Moser requested Bjarne to review the rationale document, which Kevin Hammond posted in Dropbox, by the 8th, to provide an early preview and flag any issues. Bjarne agreed to share it in the CC chat.
- Pre-Holiday Sync-up Kevin Hammond stated they need to sync up with Alexander Moser possibly tomorrow to finalize things before Kevin Hammond goes on a two-week holiday.
- Cardano Performance Analysis Initiative Neil Davies discussed an initiative to build a sideline in analyzing Cardano properly, potentially bidding for Treasury money to run a sustained performance and economic analysis. This analysis aims to benchmark the kind of work Carlos Lopez de Lara has been doing from scratch, making it a continuous process.
- ADA to Dollar Conversion Rate Modeling Neil Davies mentioned that his employee, Nikos, a statistician, is working on producing a forward pricing model to determine a reasonable value for ADA against the dollar for the next month, given past history. This effort is intended to address the current lack of a published, standardized method for converting ADA to dollars and vice-versa.
- Game Theory in Cardano Carlos Lopez de Lara noted that game theory appears to be working out, albeit slowly, within the Cardano system, contrary to some opinions that it was useless. He also shared that they drew inspiration from distribution patterns, acknowledging that the game theory ideal of fully saturated pools is not what is currently observed.
- Influence of Pledge on Rewards Carlos Lopez de Lara discussed the influence of pledge, observing that even large players like Coinbase are not fully utilizing it, which he considers a positive aspect. However, he noted that very large entities like Binance fully pledge their pools, earning significantly more rewards (5-7%), indicating their understanding of the game’s rules.
- Liquidity and Rates of Return Neil Davies explained that pledge reduces liquidity, and there isn’t enough liquidity for DApps to work effectively. He suggested that changing the rate of return might incentivize larger entities like Binance to stake with exchanges, as current rates might not be optimal for them in the mid-ranges.
- Wealth Distribution in Cardano Carlos Lopez de Lara and Neil Davies discussed the concentration of ADA wealth, noting that only 20% is locked up by big players, which is a significant fraction. Carlos Lopez de Lara noted that a quarter of the block production is controlled by the top 20% of wealth holders, indicating that the wealth distribution is not as skewed as it could be.
- Concerns about Linear Layoffs Neil Davies expressed strong concerns about the “linear layoffs” design, stating that it is a dead end and destroys scalability. He believes the design fundamentally wastes resources by performing work that is later discarded, leading to instability under load and a performance collapse.
- State Pool Viability and Sustainability Carlos Lopez de Lara and Neil Davies discussed the viability of stake pools, noting that a pool needs to attract at least 3 million ADA to be sustainable. Neil Davies suggested that while 3 million ADA may not seem like a large sum in the entire system, it represents a significant barrier to entry for new stake pool operators.
- Problem of Inactive Stake Pools and Centralization Carlos Lopez de Lara presented data showing 46 million ADA locked up in inactive stake pools, which is equivalent to 15 viable stake pools. They discussed the issue of stake pools running multiple instances on the same physical machine, leading to delayed block production, and the lack of a monitoring system to identify and publicly shame such operators.
- Proposal for Delegation Address Mobility Neil Davies proposed a “delegation address mobility” mechanism, similar to mobile phone number portability, that would allow stake pools to transfer their delegations without delegators having to take action. This would create asset value for stake pools, as their customer base could be sold, and reduce the “stickiness” problem for delegators using cold wallets.
- Virtual Stake Pool Operators (VPOS) Neil Davies suggested the creation of Virtual Stake Pool Operators (VPOS), similar to Mobile Virtual Network Operators, where small, struggling stake pools could aggregate their stake to form a single, viable pool. This would reduce barriers to entry and allow entities like charities to collectively run stake pools without owning physical equipment.
-
Critique of Pledge Leverage (CIP-50) Carlos Lopez de Lara expressed concerns about CIP-50’s pledge leverage, which aims to help those pledging and punish those who don’t, but in practice, disproportionately hurts smaller stake pools. He argued that larger entities are more likely to have sufficient funds to pledge, thus consolidating power rather than helping smaller participants.
-
System Engineering and Design Flaws Neil Davies expressed concern that certain systems were not properly designed, contrasting them with Cardano, which they asserted was successful due to proper design by Duncan and their team. Neil Davies indicated their plan to discuss these risks with their colleague Peter, utilizing Peter’s facial expressions to gauge understanding given their 40 years of collaboration. Neil Davies emphasized that these design issues are fundamental systems engineering principles that should be known and understood by professionals.
-
Virtual SPOs as a Solution Neil Davies and Carlos Lopez de Lara discussed the concept of virtualizing SPOs as a potential solution to existing problems. Neil Davies suggested that this approach would make the problem “soluble” because a virtual SPO would not require significant hardware, potentially being “literally a man and his dog”. Carlos Lopez de Lara agreed, noting that this could mitigate issues like “sticky st” and “small pools,” suggesting it as an avenue to explore.