We have repeatedly received the feedback that allowing Indexers to become overdelegated by an arbitrarily large amount creates too much uncertainty for Delegators, as each unit of delegated stake above the maximum delegation capacity dilutes the commissions received by the existing Delegators.
I would recommend capping the % by which an Indexer can become overdelegated. This threshold should be expressed as a protocol parameter that is set via governance. I suggest starting at 25%.
Would love to hear the community’s thoughts on this change. Any downsides to limiting the amount of overdelegation that I may not be thinking of?
I think capping over-delegation at protocol level would be a good way forward. Yet i also think, seeing not making it a hard limit is to allow for what’s merely a signal function, that 25% is high. On the biggest indexer we’re talking about potentially over 0.5 BILLION tokens being allocated in a diluting manner.
Why allow significant overdelegation? Why allow overdelegation at all? What was the original rationale around overdelegation?
Thinking from the delegator’s perspective, it seems to me that overdelegation is overly punitive to a delegator’s returns - especially for existing delegators who cannot be expected to project when an indexer is going to hit their cap nor trust whether an indexer is going to add more self stake. This leaves delegators in a catch-22 when an indexer is approaching their capacity.
As an indexer (with capacity) I am fine with overdelegation because I benefit from the absence of the overdelegated funds from the indexerRewards distribution, however in the interests of the overall ecosystem I would explore single-digit percentage cap on overdelegation, the lower the better, unless there is a critical reason for overdelegation to exist that I am not aware of.
Edit: I’ve just seen your reasoning around overdelegation on another thread @Brandon and think I understand the rationale better now - hands off approach, let the markets find equilibrium, minimise offchain interference. I still think this number should be conservative given the impact on delegators. Do overdelegated indexers suffer in the same way i.e. lower yield?
Allowing Indexers to become overdelegated creates uncertainty for Delegators
Allowing Indexers to become overdelegated provides an important market signal that they should either deposit more of their own stake or make their delegation parameters less attractive. This serves the purpose of a more efficient delegation market, while satisfying a design principle we had while designing the network, which was to not require off-chain coordination for mission-critical protocol functionality., This principle is intended to reduce centralization vectors and is also why the protocol has a curation market.
All that being said, we agree it would be better for Delegators to have more clarity around the rewards they should expect to earn. A relatively simple option here would be to allow Indexers to become overdelegated but only by up to a certain amount, for example by only up to 25%. More complex solutions might involve dynamic bonding curves.
I wonder if some of the problem is more to do with the perception. The idea that rewards are being diluted is scary! But - I’m not sure that people are grokking that incoming delegation is diluting their rewards no matter what and there are just degrees here.
There are 3 cases, each with a higher degree of dilution of one’s own stake → rewards ratio (assuming rewards cuts are unchanged):
Incoming delegation delegates to another Indexer
Incoming delegation delegates to the same Indexer
Incoming delegation overdelegates to the same Indexer
Looking at the network tab, in most cases the difference between case 2 and 3 has been insignificant in practice. There is maybe only one Indexer that is overdelegated to the point where this is a real problem - but they could fix this by increasing their rewards cut above the relatively low 10% it’s at now.
So before we make significant protocol changes, I’d like to see hard numbers indicating that this is a real problem and not just an overreaction to a scary idea. If it’s the latter, we might do well to communicate more clearly instead.
I think that maybe we could cap delegation rewards. But I am not convinced this is the crux of the problem. I think we need to get more solidity statistics out first. Such as:
effective cut
indexer trustworthy / efficiency scores. Such as, how often they change their rewards, historical payouts, how well they collect query fees.
Once we get these out, I think delegators can make better decisions. Right now I think this is the real problem we are seeing:
delegators appear to be willing to overdelegate because they can actually get a better return on indexers who are offering low cuts.
I think when we post those two data points this will help with this. But also, it is clear the graph is dominated by indexing rewards right now, and with only 1 subgraph in existence, it is pretty easy to be an indexer. As long as this is the scenario, I expect many Indexers will keep cuts low because their job is easy and they are happy to take low cuts and get a lot of delegation.
In summary - I think we need to get more data to make it clear which indexers are doing a good job. And then Delegators can better delegate their tokens. Also, I believe that cuts will be low until it becomes harder to run a node. If an Indexer can still make money at 15%, and get a lot of delegation, it might be economically rational for them to do this. It would be good to have the Indexers more aligned on their effective cuts, but I don’t think we can do any big protocol changes until the query market has some maturity.
An uncapped overdelegation capacity allows Delegators to keep hopping on Indexers who have low enough cuts to still achieve a higher ROI than others, even while overdelegated. I’ve also seen many people aiming for certain Indexers based on other metrics, outside of what you’re hoping to push and educate on (EG stake amount/position in the dApp, etc), so that should be considered also.
I think there’d be a wider distribution of delegations in the long-term if a cap was implemented, but something closer to 10%, IMO. Hard to say, since we haven’t seen this issue have chance to build up over a long period… yet.
As much as i understand wanting a free market, there may be merit in some parameters that act as guidance in the interest of the network and Indexers/Delegators as a whole.
Capacity limit (16x) in general is another thing to discuss in this area (and the snowball effect of centralization this could cause over time), but that can wait for another time/thread. I see the overdelegation limit as a good stepping stone towards this, but education comes first, too.
We should definitely aim to be more informative and provide extra tools for assessment. Yet we should also be pragmatic and take human nature into account. And while i understand the desire for more metrics to be collected, the longer the wait, the more difficult it is to correct, as people get set in their way and / or their existing delegation.
It’s much easier for whale indexers to press through such low cuts, whereas smaller indexers are pushed towards higher cuts to make break even. Whale indexers are also benefiting in the “economic security” criteria. So leaving overdelegation could be a threat to decentralisation.