It’s Time to Grow: Phase 3 Staking Proposal

[Lucian Mincu 4]

This proposal is focused on expanding the Elrond staking economy. Please read the information below carefully in preparation for our feedback sessions.

The changes we are proposing are meant to open a new era of growth for Elrond, specifically increasing the token distribution and decentralization, strengthening the security of the network,  and enabling sustainable business models around staking-as-a-service providers. In short, by enabling more eGold to be staked, more people and validators will participate, more tokens will be locked, security will improve, potentially increasing the value of the locked assets. 

A secure environment will give confidence to more businesses to start building on Elrond, increasing the demand of eGold, in a perpetual virtuous circle.

The following changes are proposed:

1. Increasing the staking cap:

   1. increasing the maximum number of Validator nodes

   2. Increasing the amount of stake per node, and enabling compounding feature

2. Enabling open delegation via system smart contracts

 

1. Increasing the staking cap

There are 2,169 Validator nodes active in the Elrond network, each with a 2,500 eGold stake, totalling 5,422,500 eGold staked. This is 37.49% of the circulating supply.

The Delegation Queue enabled additional participants to engage with staking and increase the amount of eGold locked to more than 8 million, or 55% of the circulating supply.

We therefore propose to increase this staking cap by a) increasing the number of Validators and b) adding a compounding rewards function that enables each validator to increase the amount staked per node.

1. Increasing the number of Validator nodes

We propose increasing the maximum number of Validator nodes from 2,169 currently to 3,200.

The number of shards will remain the same: 3 shards and one metachain. Also the number of eligible nodes per epoch remains at 400. What is changing is the number of nodes per shard that are not eligible but are on the shard waiting list that will increase from a maximum of 142 to 400 (not to be confused with the queue from Phase 1). The minimum number of nodes on the waiting list will be 80, instead of 142 as it used to be, so the minimum number of nodes will also change from 2169 to 1920.

With 2,500 eGold staked per node, the total amount of eGold staked will be exactly 8,000,000 eGold, on par with the current locked amount.

The change in the maximum number of nodes is more complex and will enable a variable number of active nodes on the Elrond network:

1920 minimum, 3200 maximum. The actual number will be driven by the market demand for eGold staking. Since the amount of rewards is fixed per year (2,16 Mil eGold in the first year) the change in number of nodes will directly affect the rewards earned by validators and delegators as well.

Please find below the rewards for Validators & Delegators in the min & max scenarios:
 

Validator	Min. nodes: 1920	Current: 2169	Max. nodes: 3200
APR %	40.65%	36%	24.40%

 

Delegator	Min. nodes: 1920	Current: 2169	Max. nodes: 3200
APR %	32.75%	29%	19.65%

 

What the limits mean:

• Minimum 1920 Validators: unStake is not possible when only 1920 Validators are in the network. Elrond Foundational nodes will be added in this (unlikely) scenario, to enable community validators flexibility

• Maximum 3200 Validators: creating a new Validator is not possible when 3,200 Validators are already staked. Elrond Foundational nodes will be progressively removed in this scenario, to allow more community nodes

2. Increasing the amount of stake per node

Another important part of this proposal is enabling people to stake an amount larger than 2500 eGold per node. The minimum value will remain 2,500 eGold, but the maximum value will be uncapped. Validators will be able to add more stake to their nodes and therefore earn more.

The amount of eGold staked per Validator does not influence its chances of being selected as block proposer or consensus member, but rewards will be proportional to that amount.

A diminishing returns mechanism will be applied, so it will be more profitable to i.e. run 2 nodes with 2,500 eGold each, instead of 1 node with 5,000 eGold. Please see the rewards system details taking into consideration top-up stake at the end of this post in the Appendix.

Some simulation data can be found below. This is based on year 1 inflation (10.84%), considering 100% hit rate (no missed blocks) and the proposed values in Appendix.

 

Scenario for 1920 Validator nodes in the network (minimum)

Avg stake per node	2,500.25 eGLD	3,000 eGLD	4,000 eGLD	5,000 eGLD
Total stake on the network	4,800,480 eGLD	5,760,000 eGLD	7,680,000 eGLD	9,600,000 eGLD
APR base stake	40.67%	38.65%	35.43%	33.47%
APR top-up stake	10.36%	10.12%	8.73%	7.20%
Average APR per node	40.66%	33.89%	25.41%	20.33%

 

Scenario for 2169 Validator nodes in the network (current)

Avg stake per node	2,500.25 eGLD	3,000 eGLD	4,000 eGLD	5,000 eGLD
Total stake on the network	5,423,042.25 eGLD	6,507,000 eGLD	8,676,000 eGLD	10,084,500 eGLD
APR base stake	36.02%	33.99%	30.78%	28.81%
APR top-up stake	10.36%	10.12%	8.73%	7.20%
Average APR per node	36.01%	30.01%	22.51%	18.00%

 

Scenario for 3200 Validator nodes in the network (maximum)

Avg stake per node	2,500.25 eGLD	3,000 eGLD	4,000 eGLD	5,000 eGLD
Total stake on the network	8,000,800 eGLD	9,600,000 eGLD	12,800,000 eGLD	16,000,000 eGLD
Average APR base stake	24.40%	22.38%	19.16%	17.20%
APR top-up stake	10.36%	10.12%	8.73%	7.20%
Average APR per node	24.39%	20.33%	15.25%	12.20%

 

This will be a powerful mechanism for Validators to compound their rewards on a daily basis, while further increasing the value of assets locked for staking and therefore the network security, with all the benefits detailed above.

Increasing the stake of a node will be possible using eGold from any source, not just Validator rewards. 

2. Enabling open delegation

Presently, everyone delegating eGold is doing so towards the Elrond Foundational Nodes. This interim solution was instrumental to securely bootstrap the Elrond blockchain. Our common goal is to decentralize the network and decrease the number of Elrond Foundational Nodes, in favor of more community nodes. Open delegation enables us to achieve that.

Open delegation will be enabled at protocol level through system smart contracts. This enables Staking Providers to build a great business model around their services, creating very secure staking products without the need to invest in smart contracts programming resources such as developers and audits. Staking Providers will be able to tailor their staking terms according to the interests of their respective customer base.

Staking through regular (non-system) smart contracts will be possible as well, and we look forward to these options gaining in popularity and complementing the system smart contracts with innovative staking products.

The Delegation System Smart Contracts will enable Staking Providers to register their nodes and define parameters related to their individual staking products, such as the maximum amounts to be delegated, and service fees.

End users looking to delegate their eGold will be able to select their preferred Staking Provider and delegate their eGold towards them.

Most importantly, every staking as a service provider will be able to start marketing Elrond, offering compelling and differentiated services to the Elrond investors and community, increasing their eGold amount delegated and growing their revenue, while at the same time significantly increasing the Elrond network exposure and overall security.

The specific details about the Delegation System Smart Contracts can be found here: https://github.com/ElrondNetwork/elrond-specs/blob/main/sc-delegation-specs.md 

Outlook on Phase 4

The Phase 4 of the Elrond Staking mechanism will cover a soft-auction mechanism, but this will only come into play at a later stage, once the maximum number of 3,200 Validators is reached and most of them staked more than 2,500 eGold. A separate post will detail the mechanism and begin community discussion once components are ready.

Appendix

A copy of the proposal and the detailed formula for calculating the top-up rewards can be found in this document.

[Flying Stone 0]

Two suggestions for improving the clarity of this proposal:

- explain whether becoming a Staking Provider who can use the Delegation System Smart Contracts is going to be purely an on chain process which is open to any existing validator, or if it requires some kind of whitelisting through establishing a partnership  with Elrond.

- clarify this sentence: 'Minimum 1920 Validators: unStake is not possible when only 1920 Validators are in the network. Elrond Foundational nodes will be added in this (unlikely) scenario, to enable community validators flexibility'. Is this a promise by Elrond, or something enforced by the protocol based on some reserve collateral locked for the purpose.

[josefcoap 9]

About Delegation System Smart Contracts:

I suggest by default to have some variables that helps to define:
standard fee (without lock commitment)
3 months fee
6 months fee
12 months fee

If a provider only "fill" the variables standard fee and 12 months fee, only this 2 options will be available in the SC

This change will cover a lot of scenarios for staking providers

[Trust Staking 1]

On 11/21/2020 at 8:02 AM, josefcoap said:

About Delegation System Smart Contracts:

I suggest by default to have some variables that helps to define:
standard fee (without lock commitment)
3 months fee
6 months fee
12 months fee

If a provider only "fill" the variables standard fee and 12 months fee, only this 2 options will be available in the SC

This change will cover a lot of scenarios for staking providers

Should we also have a flow in the main DSSC where people will choose lock commitment for a higher APR comparing with no lock commitment ?

[Sever Moldovean 23]

On 11/20/2020 at 2:56 PM, Flying Stone said:

Two suggestions for improving the clarity of this proposal:

- explain whether becoming a Staking Provider who can use the Delegation System Smart Contracts is going to be purely an on chain process which is open to any existing validator, or if it requires some kind of whitelisting through establishing a partnership  with Elrond.

- clarify this sentence: 'Minimum 1920 Validators: unStake is not possible when only 1920 Validators are in the network. Elrond Foundational nodes will be added in this (unlikely) scenario, to enable community validators flexibility'. Is this a promise by Elrond, or something enforced by the protocol based on some reserve collateral locked for the purpose.

Hey Flying Stone,

 

1. Anyone can apply to become a Staking Provider - this is 100% onchain and with zero involvement from Elrond the company.

2. This is a promise by Elrond, in the (unlikely) scenario that we drop to those low numbers.

 

Thank you for engaging with this proposal.

[Sever Moldovean 23]

On 11/21/2020 at 10:02 AM, josefcoap said:

About Delegation System Smart Contracts:

I suggest by default to have some variables that helps to define:
standard fee (without lock commitment)
3 months fee
6 months fee
12 months fee

If a provider only "fill" the variables standard fee and 12 months fee, only this 2 options will be available in the SC

This change will cover a lot of scenarios for staking providers

Hey Jose,

Acknowledged. Lucian already mentioned he is leading such efforts for the next version of the delegation SYSTEM smart contract. Just to be clear, anyone can make their own delegation smart contract with more creative features than currently included in the system one. But timelocks are great to have overall and as such is already part of the research work for the next versions.

[Sever Moldovean 23]

On 11/22/2020 at 8:32 PM, Trust Staking said:

Should we also have a flow in the main DSSC where people will choose lock commitment for a higher APR comparing with no lock commitment ?

Hey Truststaking,

That makes sense sense to be considered in the research work mentioned above. Thank you!

[ken crowley 1]

small fish

hi all, well im a small fish here and i only run one validator node, at the moment i dont want to go into becoming  staking as a service provider untill i could become more comfortable with the smartcontracts etc so untill then im quite happy running my node, if i could i would had run more nodes but i do not have the funds for this, so im just going to put this out their, i understand the proposal to increase the stake per node but putting the amount up too much may price even more people out and a few people/comapanies could end up running a majority of the nodes, this to me is not decentaliaztion as you might as well stay with the Elrond bootnodes, this is just me speaking up for the small single node operators, i would also like  to say that any increase should not be more than 500egld per year and this figure is purely based on the fact i only have 3000 egld :) so if the price does go above i hope one of you staking providers will give me a good deal to delegate hehe!, as for josefcoap proposal i also agree that their needs to be a secure lock in period like he mentioned to create continuity for the validators and staking providers.

all the best

ken.

[acelasiyo 2]

It's a good idea to remove the fixed number of Elrond needed for a node. We can compound our rewards. 

[Adrian Dobrita 12]

On 11/30/2020 at 9:42 PM, ken crowley said:

small fish

hi all, well im a small fish here and i only run one validator node, at the moment i dont want to go into becoming  staking as a service provider untill i could become more comfortable with the smartcontracts etc so untill then im quite happy running my node, if i could i would had run more nodes but i do not have the funds for this, so im just going to put this out their, i understand the proposal to increase the stake per node but putting the amount up too much may price even more people out and a few people/comapanies could end up running a majority of the nodes, this to me is not decentaliaztion as you might as well stay with the Elrond bootnodes, this is just me speaking up for the small single node operators, i would also like  to say that any increase should not be more than 500egld per year and this figure is purely based on the fact i only have 3000 egld 🙂 so if the price does go above i hope one of you staking providers will give me a good deal to delegate hehe!, as for josefcoap proposal i also agree that their needs to be a secure lock in period like he mentioned to create continuity for the validators and staking providers.

all the best

ken.

For the phase 3 staking proposal there is no change in the minimum node price, it is still 2500 eGLD

Also no validator can be priced out, as long as they maintain at least 2500 eGLD per node.

So although we are increasing the possible amount to stake per node, this is more of an increase in acceptance conditions rather than a change in rejection conditions.

Plus, for the extra 500 eGLD you mentioned you have, you can now earn some rewards, even if you do not have enough to register a new node.

Edited December 3, 2020 by Adrian Dobrita

[Kevin Lydon 1]

This all really great stuff guys! Excited for this next phase. Few questions / comments from me: 

1. Is there an auto compounding feature available?

2. When a node reaches 5000 EGLD is there a mechanism  to stake an observer automatically while putting next in line in the validator queue? 

3. "Increasing the stake of a node will be possible using eGold from any source, not just Validator rewards". - Love this. No barriers to daily cheap compounding / growth.  

4. How much EGLD should we hold in our delegation contract to ensure we are always validating? 

5. With delegation can we automate the process of staking new nodes?  

6. Will we be able to set delegation contracts up to include time based rewards for our delegators. For example someone who lock for 90 days pays less commission?

7. Does the delegation contract include a variable to remove the server cost from the gross rewards before the main distribution?

Cheers!

Edited December 3, 2020 by Kevin Lydon

[Merzouk 1]

For people who want to delegate I suggest a variable reward:

Example:

Free Delegation:  he can undelegate anytime (the actual system) :X % reward amount

3 month Lock delegation: he can't undelegate before 3 month : XX% reward amount 

6 month Lock delegate : he can't undelegate before  month : XXX % reward amount 

1 year Lock delegation :  he can't undelegate before  a year : XXXX % reward amount 

 

Cordially 

[Merzouk 1]

my point is to reward more people who are committed  to the project, and  want to secure the system for the long term. and it's a way to keep gambler away

[Marius Grigoras 3]

Just one question please: the SC for mainnet will be enabled starting with phase 3?

[Adrian Dobrita 12]

On 12/4/2020 at 5:44 PM, Marius Grigoras said:

Just one question please: the SC for mainnet will be enabled starting with phase 3?

If you mean the delegation System SC then yes, it will be available and enabled with phase 3 staking.

[Trust Staking 1]

@Lucian MincuWould be possible for others to add their nodes under our service ? 

For example, X has one node and wants to add it under my service.

Could we have in SC one new endpoint where I can allow X wallet to be able to add nodes ?

Right now I thinkaddNodes([](BLSPublicKey, SignedMessage)) will work just from SP wallet right ?

[Adrian Dobrita 12]

On 12/10/2020 at 2:32 PM, Trust Staking said:

@Lucian MincuWould be possible for others to add their nodes under our service ? 

For example, X has one node and wants to add it under my service.

Could we have in SC one new endpoint where I can allow X wallet to be able to add nodes ?

Right now I thinkaddNodes([](BLSPublicKey, SignedMessage)) will work just from SP wallet right ?

Currently only the owner (SP wallet) can add or remove nodes.

If there really is a need for this, we can improve and add some options in future iterations, but to keep things simple and the edge cases manageable I would keep it like this for the first release.

[Marius Grigoras 3]

Hi folks! I have some questions please. My understanding is that (soon) starting with the phase 3, the max no of validators will be increased up to 3200.
Also, some of the Elround Foundation nodes will be removed letting some space for some external nodes. So far so good.

My questions:
1. How are selected the nodes to be removed? Just random?
2. What's happening exactly with the tokens delegated to those Elround Foundation nodes? Will be moved transparently to some staking partners? 
3. The affected delegators will be announced and can decide for themselves?

Thank you,
Marius

[Adrian Dobrita 12]

11 hours ago, Marius Grigoras said:

Hi folks! I have some questions please. My understanding is that (soon) starting with the phase 3, the max no of validators will be increased up to 3200.
Also, some of the Elround Foundation nodes will be removed letting some space for some external nodes. So far so good.

My questions:
1. How are selected the nodes to be removed? Just random?
2. What's happening exactly with the tokens delegated to those Elround Foundation nodes? Will be moved transparently to some staking partners? 
3. The affected delegators will be announced and can decide for themselves?

Thank you,
Marius

Hi Marius,

The removal of Elrond foundation nodes is not yet in discussion for the phase 3 release, this will be further down the line and will be announced beforehand.

But just for the sake of discussion, the mechanism should be fairly simple, one option (which in my opinion would also be elegant) would be to keep the extra stake from the removed keys(nodes) as TopUp stake on the remaining nodes. This will indeed cause the return per staked token to decrease on average a bit, since the topUp stake APR is always below the base stake APR, but will also allow SaaS providers to become a more competitive alternative the more nodes are removed.

This mechanism will make the migration process from elrond foundation nodes to SaaS more organic, so if delegators see a better option out there with some SaaS they will migrate.

For your other questions:

1. It doesn't really matter which keys are unregistered.

2. Only delegators themselves can move their funds to other staking partners if they see a better option, and if they move or not depends probably on the advantages offered by the staking partners, but as explained above, since the APR will decrease a bit by unregistering elrond nodes, this will make the SaaS options look more competitive.

3. There will be no specific delegator greatly affected, this affects all delegators by a small degree, but this is already to be expected with the V2 proposal. Until keys are unregistered from the elrond foundation nodes, the APR for the elrond foundation nodes will be the highest possible (fees aside), so up to a certain number of unregistered keys keeping the funds delegated to elrond foundation nodes will still be one of the better options, out there, as long as the ratio of (topUp stake)/(base stake) for elrond foundation nodes is better than the same ratio for SaaS providers (again not considering fees). As soon as this ratio falls below what staking partners have, there will be an incentive to move.

Hope this clarifies a few things, this is not yet decided to be the way to go, but I think it is one of the better options.

If anyone has other ideas we can discuss.

Edited December 16, 2020 by Adrian Dobrita

[Marius Grigoras 3]

9 hours ago, Adrian Dobrita said:

Hi Marius,

The removal of Elrond foundation nodes is not yet in discussion for the phase 3 release, this will be further down the line and will be announced beforehand.

But just for the sake of discussion, the mechanism should be fairly simple, one option (which in my opinion would also be elegant) would be to keep the extra stake from the removed keys(nodes) as TopUp stake on the remaining nodes. This will indeed cause the return per staked token to decrease on average a bit, since the topUp stake APR is always below the base stake APR, but will also allow SaaS providers to become a more competitive alternative the more nodes are removed.

This mechanism will make the migration process from elrond foundation nodes to SaaS more organic, so if delegators see a better option out there with some SaaS they will migrate.

For your other questions:

1. It doesn't really matter which keys are unregistered.

2. Only delegators themselves can move their funds to other staking partners if they see a better option, and if they move or not depends probably on the advantages offered by the staking partners, but as explained above, since the APR will decrease a bit by unregistering elrond nodes, this will make the SaaS options look more competitive.

3. There will be no specific delegator greatly affected, this affects all delegators by a small degree, but this is already to be expected with the V2 proposal. Until keys are unregistered from the elrond foundation nodes, the APR for the elrond foundation nodes will be the highest possible (fees aside), so up to a certain number of unregistered keys keeping the funds delegated to elrond foundation nodes will still be one of the better options, out there, as long as the ratio of (topUp stake)/(base stake) for elrond foundation nodes is better than the same ratio for SaaS providers (again not considering fees). As soon as this ratio falls below what staking partners have, there will be an incentive to move.

Hope this clarifies a few things, this is not yet decided to be the way to go, but I think it is one of the better options.

If anyone has other ideas we can discuss.

Thank you Adrian, this sounds pretty good and indeed is an elegant solution that will accelerate as well the level of decentralization. 

[Marius Grigoras 3]

Another question please - do you have a time frame in mind for how many months the max no of validators will be set to 3200? 

Thank you,
Marius

[Adrian Dobrita 12]

17 minutes ago, Marius Grigoras said:

Another question please - do you have a time frame in mind for how many months the max no of validators will be set to 3200? 

Thank you,
Marius

As long as we don't change the number of shards, this should stay.

Increasing the number of shards will only be triggered when we reach a high enough block fill ratio and implicitly TPS (one block has a capacity of 1.5 *10^9 gas).

I would not expect to have the change before we reach ~40% average fill ratio of the block capacity, which would mean 10k-12k regular transactions/block (~6k TPS for the current network setting)

It would be amazing to reach this in the next one year, but very difficult.

Also unregistering elrond foundation nodes will not affect the maximum number of nodes.

[Adrian Dobrita 12]

Actually there is one change, where we plan to remove the maximum number of nodes limit, this is for when we introduce phase 4 staking which was briefly touched in the proposal, but the effect would be that we can have no limit for registering nodes, but in every epoch you would also only have 1600 nodes eligible, out of all registered ones, with a maximum of 20% swap ratio for shuffling, the nodes shuffled into eligible list being the best qualified nodes (considering stake value and possibly rating)

Edited December 17, 2020 by Adrian Dobrita

[EGod 0]

When does phase 3 begin?

[Milan Kraus 0]

1. Can also the base 2500 EGLD come from delegators? Or does it always need to be provided by validator?

2. If it can come from delegators and they undelegate so a node has less than 2500 EGLD stake, will the node be automatically deactivated?

3. I'm not sure whether slashing is already enabled, but once it will be, will it affect also delegators? If yes, it will be quite risky to lock delegation for a period of time - delegator will not be able to react to validator that starts to misbehave.