TradeHub Security & Decentralization Improvement Proposal"
On the technical side, a strong TradeHub ecosystem starts with healthy validators and decentralization of voting power. This responsibility rests in part with validators themselves, and in part with delegators distributing voting power towards the validators.
One can not assume that validators or delegators will act from a place of altruism, and therefore the basic framework regulating voting power must disincentivize centralization of power by design. The current framework lacks such measures and must be optimized to ensure a strong future TradeHub.
The more voting power is concentrated among a handful of validators, the more vulnerable TradeHub is in the event of server malfunctions, rogue validators or validators who may in time simply be in dereliction of their duty.
Currently, several TradeHub validators are running on zero or close to zero commission. We believe that this is a direct reason for TradeHub's lack of ability to attract a full set of validators. Out of the current maximum of 23 validators there are only 22 active validators, and there is no competition for available spots. This is a security risk. A new validator launching on TradeHub is highly unlikely to attract delegators if they set their fees at a level that is likely to cover actual server costs, much less turn a profit to incentivize active participation and additional effort.
For the above reason, we propose to disincentivize delegators who delegate their stake to validators above a certain voting power threshold. This proposed measure follows a stepped approach which gradually increases the disincentive as risk increasing behaviour escalates.
This added commission will be distributed to the Switcheo Development Fund. Later proposals may suggest distributing these funds evenly across delegators as increased block rewards.
We propose to increase the commission for all delegators who delegate their stake to validators with more than 7.5% voting power, increasing by 0.5% per 1% voting power (rounded up) above the threshold, minus validators usual commission.
A validator with 7.42% voting power will incur no additional commission, he's near the limit. If his usual commission 1%. His adjusted commission will be 1% (1% for validator and 0% for the Switcheo Development Fund.)
A validator with 7.77% voting power will incur an 0.5% additional commission. If his usual commission 2.5%. His adjusted commission will be 3% (2.5% for validator and 0.5% for the Switcheo Development Fund.)
A validator with 10.15% voting power will incur an 1.5% additional commission tax. If his usual commission 3%. His adjusted commission will be 4.5% (3% for validator and 1.5% for the Switcheo Development Fund.)
A validator with 12.88% voting power will incur an adjusted 3% additional commission. If his usual commission 1.2%. His adjusted commission will be 4.2% (1.2% for validator and 3% for the Switcheo Development Fund.)
A validator with 17.66% voting power will incur an adjusted 5.5% additional commission. If his usual commission 2.5%. His adjusted commission will be 8% (2.5% for validator and 5.5% for the Switcheo Development Fund.)
One could argue that this proposal, if implemented, would penalize a delegator who delegated their stake to a validator while it had a low voting power. To this we would respond that the delegator is free to move at any time and this change could also be broadcast to take effect in 30 days, so re-delegations aren’t affected. TradeHub should encourage active delegator participation and monitoring of the project in shaping the ecosystem.
Rationale for the defined threshold:
TradeHub has a current maximum of 23 validators.
The proposed threshold is slightly lower than double an evenly distributed average voting weight of 4.3%, which provides scope for validators to innovate without adjustment.
Setting a higher threshold of would incentivises a concentration of power among a handful of validators (the current status quo), not providing sufficient incentive to launch a new validator on TradeHub.
A lower threshold of for example 5% would incentivize a near even distribution of voting power among validators, potentially removing the necessary incentive for a validator to provide higher quality services to TradeHub.
From MaiTePora DevHub Commununity - for Switcheo - 2021 MAY
I strongly agree, I will vote yes
We are fine with implementing this if the majority agrees.
I agree in princple that the current situation lacks a mechanism to discourage/prevent a validator from growing too big.
I agree with the proposed approach above as well.
I would just want to point out that, even if BH reduces to 7.5%, the difference is 9.79% - if you spread it out, it's still <0.5% per validator, which might not save a good budding validator, like S3.
To achieve the second goal of this proposal, which is to incentivise new validators and help smaller validators, I would suggest for a performance bonus as well. A smaller validator can still earn SWTH if they maintain a good setup (hence good uptime) and vote often. That can be discussed at a later stage.
Great for decentralization
I think this is a fantastic idea and we should proceed. We cannot risk peoples investments and the security of the chain on a few validators. At the end of the day we most likely want 50 or more validators anyway. It would be a yes vote from my end.
I agree it's a fantasitic Idea... We all aim decentralization, and it's a perfect proposal. And, it's good for new validators who wants to get stakers without intense maketing. We can imagine a "negative commissions" which gonna helps, new validators.
I also agree that the existing commission only mechanism is too blunt an instrument and does not adequately address the inherent conflict between validators maximising their voting power and chain health via a wide vote distribution, ensuring the safety and continuity of operation of the chain.
I would additionally propose the following:
- The introduction of the adjusted commission have a moratorium period (of say 30 days), to allow stakers to re-delegate without incurring any reward penalty.
- That the proposal incorporate the option to distribute the adjusted commission pro rata across stakes on < 8% voting power validators. This will make the proposal a redistribution of rewards and eliminate the criticism that it is a further tax.
@kang - A kick starter program of some kind for new validators is an interesting idea. Given that stakers (including validators), would be voting to increase the number of validators, it would also be reasonable that they would support them via a stake.
I like the idea in its essence. Some open questions:
- How does it scale to > 23 validators? (e.g. How does this look like with 100 validators?) Ideally it should work nicely even if max validator num is voted to be larger or smaller by using scaled values, rather than using hard-coded numbers.
- In terms of implementation - what is needed? Has the cosmos-sdk codebase that is relevant to this (see x/distribution module) been looked at? Core devs only has time to do so after Stargate. If we have consensus on the idea, ideally the patch can be funded by SDF too.