Continuing discussions on Inflation

[b3nnn]

Hi everyone, I want to pick back up the inflation topic and provide an update.

In previous inflation conversations we set out to clarify some target parameters that would reflect the consensus for change. However, the survey and subsequent discussions revealed an underlying confusion that had not been fully identified earlier. Lowering inflation is not a competing or alternative solution to other possible improvements to the emissions system, but rather one important change we can take to start driving improvement.

Emissions increased to 24% last year. Rather than leave them to (potentially) increase further, we are planning to propose parameter changes to take control of emissions, aiming to slow them down to roughly the rate at which they accelerated last year (as a first step). We will share the draft LIP for this next week.

This will be in parallel to wider efforts to improve the emissions system over the next 6 months. I expand below on the broader emissions system, its impact on staked and unstaked LPT holders, as well as other challenges, and set out a path for multiple efforts to improve the system.

WHAT ARE TOKEN EMISSIONS FOR

Not to rehash too much, but it’s good to make sure we are aligned on core purposes of emissions.

• Bootstrapping and attracting participants like stakers and orchestrators. This has been a critical part of driving network participation so far, and while still important, is probably less of a focus right now given the current level of emissions

• Rewarding specific behaviours. Securing the network, rewarding stakers, supporting orchestrators and driving “active” participation are all behaviours that emissions can reward. This is a key focus for emissions today. In particular we focus on active participation, as we seek to ensure rewards are distributed to types of participation that provide the most value to the network

• Funding growth and creating/preserving token value. The onchain treasury is a critical component of broader network economics, ensuring funds are available to support network development, security and the growth of fees on the network. Sustainability is good token economics, and we all acknowledge that emissions should transition to fee-based revenue and a more stable, low-inflation model over time.

In short, the overall goal of emission is to bootstrap participation, reward the right behaviours and fund the treasury and growth towards sustainability. I think we are primarily through the bootstrapping phase, and need to bring focus to how emissions and inflation reward the right behaviours and help fund the actions that will bring more fees onto to the network.

WHERE ARE EMISSIONS GOING TODAY

Emissions are going to three primary recipients, each with different needs and some current challenges worth highlighting.

• Delegators: Delegators provide capital to support the project. We currently reward stakers/delegators while diluting unproductive (unstaked) capital. However, this simplification of “productive” capital belies perhaps a different reality. Much of our stake today might be considered dormant or passive, as seen in stake allocated to low performing orchestrators and the low mobility of stake moving between orchestrators. While not necessarily a problem, further discussions around how we use emissions to reward our desired or most-valuable behaviours is important

• Orchestrators: Orchestrators provide GPUs and the network’s service. They are rewarded through emissions and through fees from customers. In an ideal world fees would generate sufficient rewards such that emissions are required less for this purpose, although we are not there today. One challenge today is that emissions go to all orchestrators, and are not rewarded on a “quality” basis. In a more ideal system we would see greater rewards go to high performing orchestrators. Similarly to re-exploring participation, we should explore how emissions might help align to goals around quality and performance.

• Treasury: The onchain treasury exists to support acceleration of the ecosystem’s goals. This is a core part of the emissions system, but has been somewhat neglected with the treasury reward cut offline for so long. With treasury rewards coming back online, and the development of the ecosystem Roadmap for Livepeer (including marketing, BD and growth initiatives), it is worth refocusing on the important role of the onchain treasury in our ecosystem. Using emissions to fund the right type of public goods and ecosystem growth initiatives is critical to our success.

Emissions are about rewarding not just participation but the right behaviours. This is nuanced and requires tradeoffs. Tuning the flow and distribution of emissions to generate better overall results for the ecosystem is part of having a strategic emissions system.

HOW MUCH ARE EMISSIONS TODAY

The current rate of emissions is very high relative to a range of other cryptoeconomic systems and projects. This has a significant impact on holders of unstaked LPT, who last year saw their share of the network diluted by 24%. While not necessarily unreasonable, we should not shy from asking why it is that our emissions need be high relative to other projects.

Consider the reality of where these emissions currently go. The majority of emissions go to delegators, whether they are active (moving stake to the highest performing orchestrators) or whether they are passive (keeping stake for years on low performing or inactive orchestrators). Consider also network maturity. The project is searching for product market fit, and yet we have deployed little capital from the onchain treasury in pursuit of this (such that the treasury reward cut was left at zero for nine months).

WHAT DO WE NEED TO DO THEN?

The emissions system is not just the rate of inflation and rewards to delegators and orchestrators. It is the system that rewards the right behaviours and provides additional capital to participants who can drive the network forward. I think the following 4 items should be tackled in H1 of 2026:

1. *Pass a proposal to take control of emissions, with parameters that target total emissions at around the same amount as last year (as a first step)
   *
   Under the current parameters, emissions for 2026 could easily be higher than in 2025, even though the target bonding rate (50%) is below the current rate (52.2%). Emissions could well be higher or roughly the same in H1 2026 as in H2 2025, and both are quite a bit higher than H1 2025. Should emissions be higher this year than last year? I can’t think of a lucid argument why that is necessary. By not taking action, we leave the answer to this question to chance

   Passing a proposal to change the parameters and target a rate of total emissions would take control of inflation rather than leaving inflation as a function of “whatever happens”. By first setting an H1 objective of a rate of emissions in line with the first half of last year - rather than potentially higher than the second, we can also simplify the discussion - stakers would be receiving tokens at a rate they have already experienced recently.

   The modeling done by Shtuka Research (@awma) shows that we can target this over a safe and observable time horizon of 6 months. There is no historical evidence that parameter changes would risk any large undelegation events. We will share a draft LIP expanding on this, but the practical path to choosing the target parameters was:

   1. Run a randomised set of simulations under a variety of tunings
   2. For each tuning, estimate the probability of achieving the objective using the simulations.
   3. Accept only tunings whose estimated probability of achieving the objective is >95%
   4. Among the acceptable tunings, choose the ones closest to the current parameters

   We will give some time and space to digest the entirety of this post, and then bring forward the Draft LIP for discussion (most likely next week)

   2. More effectively allocate the treasury to core network development, ecosystem growth AND active participation in the network.

   The ecosystem treasury will start to repopulate from the treasury reward cut. The Livepeer Roadmap consolidates a list of higher-value projects which require talent and funding to help us accelerate around BD, marketing, growth and other core services, and complement the efforts of Inc and other gateways in driving fees to the network.

   Governance should direct funds to these projects in priority to other projects. The treasury exists to fund the highest impact projects and deliver real ROI. It is not there to fund speculative proposals with little real connection to the ecosystem. As seen in discussions here about turning the treasury reward cut back on, there is a strong desire for greater accountability and transparency around the ROI on funds deployed. I recommend that we have an open discussion and start to set ecosystem norms for increased accountability and transparency, using existing governance mechanisms (and any new actions the community deems appropriate).

   The treasury can also be used to target “high quality participation”. Rewarding higher performing orchestrators is one such goal, as it increases the quality of the network and creates a (greater) incentive to adopt best practices that benefit users. We can explore different types of programs that will help to further incentivise the right actions and behaviours that increase the output, efficiency and growth of the Livepeer Network. This can be run in a separate discussion thread or as workshops to align on how we might make these changes, and where relevant put them put up for a vote. Of course many of these are nuanced and might be considered “subjective”, so open discussions and further truth seeking and consensus building is needed before any changes might be proposed or adopted.

   3. *Identify useful actions on passive, inactive or extractive participation.

   A range of edge cases and known issues exist as it relates to different types of participation that don’t add much value to the projects. “Vampire” nodes, inactive orchestrators, dormant stake and other types of passive participation are all being rewarded today. These also require consensus building and a nuanced and thoughtful response, but - for the most part - they do require a response, as these are not behaviours that necessarily move us forward, and rewarding them becomes a tax all other participants pay for our inaction. I would recommend that we roll these discussions into the item above about “high quality participation”.

   4. *Lower overall inflation.

   By increasing efficacy of how emissions are distributed and deployed across this range of efforts, we can start to further target lower overall inflation. In 6 months we will see progress across these items, plus an increased understanding of how network fees are growing, and the results from taking control of inflation, such that we can revisit and further adapt the emissions system or target lower target inflation.

I appreciate this is quite a long post after the holiday break, but I wanted to give a holistic view of the emissions system and actions we can take hand-in-hand to move us forward. By taking control of total emissions and using the parameters to get us to target total emissions at the same level as last year, we can simplify the math for everyone and ensure that emissions do not continue to grow indefinitely. Taking this action is not the only action available to us, and we aim to increase the impact of our emissions system by further discussing system design and the way emissions are distributed. We can revisit and (potentially) retarget lower inflation again with 6 months of additional evidence and improvements.

I’ll be on the watercooler to discuss this in the coming weeks, and we will share a draft LIP to show how the proposed parameter changes take control of inflation and the very low risks of taking action. It will also include the observation system so everyone knows how we actively manage for any scenario. I welcome any and all comments and am excited to get conversations moving on all of the above.