Changes on Polkadot in March 2026

Ecosystem
,
1

Note: Please find a summary of the changes made to the post at the end

March 2026 marks a period of major, interconnected changes for the Polkadot network. This short post aims to recap and summarize these changes, but we urge anyone unfamiliar to read the detailed linked resources for all the information.

DOT Supply Cap and Issuance Changes

On September 14, 2025, the Polkadot community approved an on-chain Wish for Change (WFC) governance proposal that set a hard cap on the DOT token supply at 2.1 billion DOT.

As a result, the first issuance reduction will take place on March 14, 2026 (Pi Day). At that time, annual issuance will be reduced from 120 million DOT to 55 million DOT per year, and every two years thereafter it will be reduced by 13.14% of the remaining supply to be minted, gradually converging toward the 2.1 billion DOT cap.

Dynamic Allocation Pool (DAP) and Staking: Phase 1

Accompanying the new issuance model, a proposal for a Dynamic Allocation Pool (DAP) and a new approach to staking, budget allocation, and network security was put forth. On January 28, 2026, the community approved Phase 1 of the implementation of the DAP through another WFC referendum.

Phase 1 will include the following changes.

Changes pertaining to the DAP

These changes will be introduced with the 2.1.1 runtime upgrade, expected to go live on Polkadot towards the end of March 2026, 1-2 weeks after Kusama.

  • A basic version of the DAP pallet and a permanent account that can hold DOT

  • Treasury burns will stop and instead the DOT tokens will be sent to the DAP

  • Validator slashes will also be directed to the DAP

Staking changes

These changes will be introduced through another referendum that will update on-chain parameters. The timeline isn’t fixed, but they’re expected to be enacted within a month after the enactment of the 2.1.1 upgrade, by the end of April.

For validators

  • Validators will be required to have a minimum self-stake of 10,000 DOT that is slashable

  • A minimum commission of 10% will be introduced

Validators that don’t have at least 10,000 DOT in self-stake when the minimum is introduced will be at risk of being permissionlessly chilled by anyone.

So, to avoid any unpleasant surprises, we urge all validators to begin preparations now to meet the 10,000 DOT self-stake before it is enforced on-chain.

Session key management

  • With the 2.1.1 runtime upgrade, validators will be able to set session keys on Asset Hub using the `stakingRcClient` pallet.

    • In the beginning, validators will be able to set the session keys on the Relay Chain too, as it is now, but eventually this will be deprecated.
    • To set the session keys on Asset Hub a deposit of ~60 DOT will be needed (and ~3 KSM on Kusama AH)

StakingOperator Proxy Type

The introduction of the 10,000 DOT minimum self-stake may pose challenges for staking provider companies (“Operators”) that run validators for institutional stakers (“Stakers”), as they may be unable or unwilling to supply the required self-stake themselves.

To address this, Polkadot will introduce a new on-chain proxy type: the StakingOperator proxy. In this setup, the Staker controls the account that appears on-chain as the validator, which holds the self-stake. That account can have a StakingOperator proxy, which has a narrow set of permissions, controlled by the Operator (i.e. the company running the node). .

image|428x238.03101246579513
image|428x238.031012465795131426Ă—794 22.9 KB

This mechanism enables Operators to run validators on behalf of Stakers in a fully non-custodial way.

The StakingOperator proxy type will become available with the 2.1.1 runtime upgrade (2.1.0 on Kusama) and it is already available for testing on Westend. Operators and their institutional clients should take advantage of the time delay until the minimum self-stake is enforced, if they wish to set up their validators using the StakingOperator proxy.

To learn more about how to use the StakingOperator proxy type, please refer to the accompanying documentation.

For nominators

Once the vast majority of active validators have the minimum self-stake a new referendum will be proposed to apply the following changes for nominators:

  • Nominators will be unslashable

  • Unbonding time for nominators will be drastically reduced from 28 days to 24 to 48 hours, depending on the timing of the unbonding.

A timeline for these is difficult to estimate as they depend on the behaviour of the majority of validators (and governance’s voting period on a new set of parameters with enough approval from DOT holders), but a proposal to alter these parameters should be reasonable expected to happen in Q2 2026.

Beyond Phase 1

For an overview of the next steps that the community will be asked to ratify, we urge validators, nominators and all stakeholders to read the DAP roadmap and the DAP proposal. They involve some breaking changes to how staking will work, and below are the key points for Phase 2:

Phase 2 (ETA Q2-Q3 2026)

  • Full DAP implementation

  • Separate budget for validators and nominators

  • Validators make three types of rewards:

    • DOT rewards linearly vested over a year, dependent on the self-stake amount

    • Stablecoin reward for operational costs

    • Self-stake is part of the total stake and also receives nominator rewards on top of the vested rewards

  • Proposed optimal self-stake T = 30,000 DOT and cap C = 100,000 DOT (i.e. max self-stake that generates rewards). The suggested target APY is 30% if all validators bond 30k DOT, and 9% if everyone bonds 100k DOT.

  • Proposed reward for operations is $2000 per month per node, paid in stablecoins.

  • Both these rewards are proportional to era points accumulated

  • There is no more validator commission

The proposals and details of Phase 2 will need to be ratified by OpenGov and may change before Phase 2 is implemented, especially the numbers.

Timeline

In short the expected timeline of events is as follows:

March 14, 2026

  • Issuance reduction

End of March, 2026

  • Runtime upgrade to 2.1.1 that includes:

    • Basic version of DAP

    • Treasury burns and slashes directed to DAP

    • StakingOperator proxy type

    • Session key management enabled on Asset Hub

    • ~60 DOT deposit to set session keys on Asset Hub (and ~3 KSM on Kusama AH)

Q2, 2026

  • First Stage of changes proposed to OpenGov (Estimated by end of April 2026):

    • Validator minimum self-stake of 10,000 DOT
    • Validator minimum commission of 10%

  • Afterwards:

    • Nominators become unslashable
    • Nominators can unbond in 24 to 48 hours

The details of these implementations and the dates may change. We’ll update this post as these become more crystallized.
———————–

Notable changes to the post on March 6, 2026

  1. The biggest change is to the timeline:
    a. The 2.1.0 runtime upgrade will most likely happen between 23-27 March instead of March 12.
    b. The governance proposal to introduce the minimum self-stake and minimum commission will likely happen towards the end of April
    c. The governance proposal to make nominators unslashable and reduce the unbonding period will come after that
  2. A deposit of 10 DOT is introduced in order to set session keys on Asset Hub (and similarly 1 KSM for Kusama AH)
  3. The StakingOperator proxy is already available on Westend for testing

———————–

Notable changes to the post on March 10, 2026

  1. Updated the runtime version of Polkadot that will have the changes to 2.1.1 (from 2.1.0).
  2. Updated the expected enactment on Polkadot to “end of March” from 23-27 March.
  3. Updated the values of the deposit to set session keys on Asset Hub to ~60 DOT (and ~3 KSM for Kusama AH).
2

PSA to all Polkadot validators:

Once this phase happens and if the current issues still remain, you should relocate your machines to Europe.

Given recent changes and the current issues (1)(2) with validators’ points away from the Europe supercluster (even validators in the US east start to be impacted, US West Coast with slight impact Oceania, Asia, Central/South America and Africa most impacted), it is recommended to relocate your validators to Europe once these changes have taken place at risk of losing era points.

3

Thanks for the great summary, @michalis . Couple of small notes / corrections:

  • the session key deposit on Asset Hub will be around 60 DOT on Polkadot Asset Hub and ~3 KSM on Kusama Asset Hub (10DOT / 1 KSM mentioned above still refers to the initial hard-coded implementation then replaced by a proper per-key/per-byte deposit calculation, apologies for not communicating the change earlier).
  • Kusama referendum for 2.1.0 is out see Runtime Upgrade 2.1.0 with enactment ~17th March
  • Polkadot will be updated straight to 2.1.1 a couple of weeks later (date still TBC but expected by end of March). 2.1.1 will contain the fix for this issue
4

Please note that as of today, March 23, the runtime version 2.1.1 is live, which means the StakingOperator proxy is available on Polkadot. As a reminder please check these resources:

  1. StakingOperator proxy documentation

  2. Guide on how to use the StakingOperator proxy

Furthermore, with this runtime version session keys can be set on Asset Hub, which requires a ~60 DOT deposit.

5

Can we please not relocate machines to Europe?

If there is a rewards problem, then we need to slow down enough to keep the geographic diversity, or maybe somehow restore the de facto geographic affirmative action that DN/1KV provided.

Is there a bug on this? If not, can someone please file one with suitable evidence? Feel free to cite me as requesting it.

Also, please check out https://github.com/polkadot-fellows/RFCs/pull/119 and comment if you think it causes geographic problems. That’s still out target method for validator rewards. We farmed it out to an external team, and I probably did not provide enough guidance, so it probably fell by the wayside, but it’s still the least incorrect algorithm for rewards.

6

Hi @burdges AFAIK after the sunset of DN the majority of individual validators who were in geographical locations away from the European and North American clusters are either out of the active set or relocated to Europe or NA. The fixed costs of operating from South America or Africa are simply too high to be sustainable, especially with the price of DOT that keeps falling.

During the time my Polkadot validators in LATAM were in the active set I have reported the increased issues with increased missed votes, especially with elastic scaling parachains, same for Kusama, both are linked above in @SAXEMBERG’s post.

Unfortunately there is also little support from community nominations, geographical decentralization is not a motivation for most nominators and decreased performance due to latency means less rewards for validators and nominators alike. I decided to keep paying my server in LATAM with the hope that on pi day the min self-stake of 10K would come into effect but with several delays, bugs, etc this just came into effect today, over 3 months later than expected. During this time I have done nothing but loose nominations, same as other validators. At this point most ex-DN validators have stopped paying for their servers and chilled their nodes.

@dozenodes has a geographic saturation tool that approximates the active set’s current composition: https://www.dozenodes.com/polkadot/location-saturation there are still people operating some nodes in South America and Oceania and many have moved to Asia, which is a good thing (maybe they can comment here or on the GitHub issues). However, Europe and NA combined still represent over 75% of the active set.

7

For now it’s going to be the market talking so it only makes sense to move all machines back to Europe for lower operation costs as well as a mitigation for lost points on average along with the paltry rewards thanks to the DAP upgrade (yes, it is lower than before as previously anticipated and pointed out during the DAP discussion and without the promised base 5000 USD / stable rewards which was originally intended and keeps lowering every time it’s inquired about). If you provide a validator with uncorrelated geography, lower rewards and lower points then the staked delegation will be lost to other Europe machines while keeping higher operating costs so it only makes sense to move back to Europe once delegator unstaking starts picking up (it already has to some extent like Florentina has already pointed out). So for now it’s better to play it conservatively and wait for future breaking changes like the planned further reduction of the validator set to implement geo decentralization again.

Not so sure if calling that “affirmative action” is the right approach as simply not implementing breaking changes to validation would have sufficed to maintain the former decentralization so overall our strategy and recommendation is still to move back to Europe unfortunately.

There is already issue 10921

It started after elastic, somewhat fixed after inclusion of ZK (so around the time of 1.22 with BLS batch proof verification / optimizations made it better maybe or was it something else?, not sure) and the consensus by that issue was that it was caused by latency regardless of the heavy variation shown pre-post elastic and pre-post zk :man_shrugging: .

This was checked a while ago but since it never picked up steam it felt it was abandoned. Good to see it coming back but at this point it might be one fix within many others requiring it namely DAP, uncertainty about future rewards, uncertainty with the new JAM token, uncertainty about future validator reductions etc. We might have a look at it later.