Should APY Scaling lean more favourably towards Long Term Stakes?

Governance
24

I did some additional thinking and found that if we assume that risk scales with the square root of time, rather than linear time, the calculation looks a lot different.

sqrt(0.5) ​≈ 0.71
sqrt(8) ​≈ 2.83

Using square root scaling, the time increase is calculated as follows:

2.83/0.71 ​≈ 4

So under square root risk scaling, the risk is only 4 times greater, rather than 16 times greater using linear scaling. The reward difference remains about 1.9x higher for 8 years.

While that looks a lot better, it’s still a significant gap. But there are other factors we can include that will make the disparity even lower. For example, age bonus and daily compounding.

With the max age bonus after 4 years and rewards compounded daily, a 6 month neuron gives 11.64% APY (not APR, since we’re compounding) and an 8 year neuron gives 23.59%. Square root time increase remains at 4, and the reward increases to ≈ 2.03x, up from 1.9x.

There are other things to consider that are difficult to put an exact number on, like the value of voting power, i.e. the ability to control changes to the network. Furthermore, keep in mind that if investors do expect the price to increase over time (why would they invest otherwise?) the compounded returns are much greater with a higher APY which could close the gap even further. Can anyone think of other factors we can include in this analysis?

2 Likes
25

You can potentially get out in time if you sell your Internet Identity on IDgeek for a discount. Say, you sell your 6 month neuron for 3% discount (or your 8 year neuron for 7%-10% discount) and ICP crashes 50% 1 month later.

As soon as IDgeek was released, the concept of risk in staking has changed quite dramatically. Therefore, as anyone with locked neurons can exit basically whenever (and re-enter whenever by rebuying lower on exchanges), current rewards might actually be greater than necessary - and 6-month neurons are obsolete imo.

2 Likes
26

You’re right, that’s a great thing to consider as well and completely throws the ability to quickly estimate risk and fairness of these rewards out of the window. In fact, it may be the wrong approach to try to calculate this at all. A better solution might be to look at the outputs rather than the inputs and make adjustments from there. In other words, is the amount of ICP flowing in to higher dissolve delay neurons increasing or decreasing? Perhaps a system where the long and short term rewards automatically vary depending on how much ICP is being locked up long term in the past X days/week/months would help accomplish the goal of encouraging longer lock-ups.

27

As far as I know you can’t use geek to sell your II if you set a recovery phrase and with current updates that can no longer be removed. So unless you plan doing that it is not really an option for the masses. Also why is nobody talking about the minimum lock up period of 6 mo. In order to get a 6 mo reward you need to start dissolving after 6 months and such you have your tokens locked for a year but only get rewards for half so you need to split the numbers.

1 Like
28

Thank you for your thoughtful insights; they provide valuable alternative perspectives on the substantial gap in staking rewards highlighted in the initial article. However, it’s important to explore these points more deeply, as they may not completely alleviate the original concern.

Square Root Risk Scaling:

The concept of square root risk scaling offers an intriguing approach to assessing risk, one that appears to temper the perceived disparity between short and long-term staking. While this perspective is noteworthy, it’s essential to underscore that it doesn’t entirely eliminate the core concern. Even with this adjusted method, there remains a discernible difference in rewards between short and long-term staking periods. The square root model may indeed soften the perceived risk gap, but it doesn’t obliterate it. Investors still encounter significant variations in reward structures based on their chosen lock-up period.

Age Bonus and Daily Compounding:

The inclusion of age bonuses and daily compounding undoubtedly provides a more intricate understanding of how staking rewards accrue over time. However, it’s crucial to emphasize that these mechanisms are integral components of the current reward system. Acknowledging their influence does narrow the gap between short and long-term staking, but it’s vital to remember that this doesn’t entirely resolve the initial concern regarding disparity. While it highlights the subtleties within the reward structure, it doesn’t eliminate the core issue of divergent rewards based on lock-up durations.

Value of Voting Power:

Recognizing the significance of voting power in governance decisions is a valid point. Nevertheless, precisely quantifying its impact on rewards and risk remains a formidable challenge. The value of voting power can fluctuate significantly, contingent upon the governance structure of the network and the extent of influence wielded by individual stakeholders. While it underscores an additional dimension of staking, it doesn’t directly address the fundamental concern of varying rewards linked to lock-up durations.

Concluding, your insights offer alternative angles and enrich the conversation, but they don’t fully mitigate the initial apprehension regarding the substantial disparity in reward rates.

The central question persists: does the existing reward structure align with incentivizing long term staking?

29

Dear @Accumulating.icp,

What would be your own answer to your question ? Would you be in favor of reducing short term staking incentivization while letting the long term staking one as it is ? Reducing both ? In both cases, which ratio would you recommend ? Something like this ?

Regards

PS : given the other topic you initiated and Kyle’s statement there saying the selling pressure comes mainly from neurons’ rewards and dissolving rather than from NP, a new ratio could be a way to reduce this part of the inflation, I mean if short term stakers tend to sell more than long term ones.

1 Like
30

I genuinely didn’t plan on saying, as whenever I give my perspective, it gets framed as though I am “building a narrative” or a “bad actor” - so I was content simply starting the conversation & chiming in where I found interest.

However, given you’ve asked, I’d be in favour of a couple changes;

  1. Redefining the scaling ratio of APY to significantly incentivize long term staking over short term staking

  2. Lowering the maximum APY to create more sustainable tokenomics (as burn rate is extremely low in comparison to what we print)

2 Likes
31

Of course ! And I hope I did not sound as if I was wanting to challenge you. Thanks for the answer and your involvement in these questions among others.

PS : I agree with you. Especially as we are dealing with a reverse gas model.

2 Likes
32

No worries, I didn’t take it that way at all!

I appreciate your participation in the discussion - we’ve been desperately lacking serious governance contributors as of late !

1 Like
33

Going to share a tweet by @dominicwilliams as I feel as though it has some relevance to this thread. Hoping to receive further clarification upon his second point - is this implying DFINITY is internally discussing redefining APY scaling?

Going to tag a few other relevant parties as well to continue this conversation.

@diegop @bjoernek @Kyle_Langham

4 Likes
34

Hello everyone ! I would like to give an opinion with relevant points if you can take them into consideration.

  • It is very important to motivate and encourage people to block the neuron for longer.
  • I think that currently the benefit of blocking for 8 years is little in relation to the amount of time one supports the blockchain.
  • We currently have a reward base of 16.8% per 8-year lock and a reward base of 9.5% per 1-year lock. The reward should be reduced in neurons that block 1 year and increased in those that block 8 years.
  • The idea that occurs to me is to reduce the rewards (neurons less than 4 years of blockage) and add them from neurons with blockages of more than 4 years.

Example:

Days || APR

1 - 365 - > %5 (now 9.5%) | Lose 4.5 %
365 - 730 → % 7 (now 10.5%) | Lose 3.5 %
730 - 1095 → % 9 (now 11.5%) | Lose 2.5 %
1095 - 1460 → % 11.6 (now 12.6 % ) | Lose 1 %

1460 -1825 - → 14.7 (now 13.7 % ) | Gain 1 %
1825- 2190 → 16.7 (now 14.7 % ) | Gain 2 %
2190- 2555 → 18.8 (now 15.8 % ) | Gain 3 %
2555 -2919 → %20.8 (now 16.8% ) | Gain 4 %
2920 (8year) → %22.3 (now 16.8% ) | Gain 5.5 % ( Plus 1.5 %)

Added to this is the bonus for age, which would work the same as now.

I know it’s a simple idea. But it is better to make minor and simple changes, since making them more complex I think would harm rather than help.

Thx !

3 Likes
35

Thank you for this.

Look forward to knowing more about this.

I am open to the idea of attracting more investors by lowering the min. lock. and make them trying the staking/governing, but if the rewarding ratio is not also revised, we will just increase the minting of ICP and selling pressure, because the shorter the term of staking is, the more the staker tend to mint and sell at the market price at the time of minting.

1 Like
36

Long-term staking is already significantly incentivized over short-term staking. That is why such a large amount of ICP is locked up in 8 year neurons.
To provide additional incentives to long-term staking, there would need to be some proof that the stated aim of getting the long-term stake to 90% is actually a good goal. I think the opposite is the case, a 90% long-term lock-in would be very bad indeed. But it is just taken for granted by the 8 year gang that the more long-term staking there is, the better.
It has always been clear to me that people making such proposals are not thinking of incentivizing new investors to stake for long periods. They are motivated solely by a desire to boost the APY they receive for neurons that are already staked. They staked for 8 years knowing the inflation trajectory and want to change rules after the game has started. Well, the NNS is built in a way that all rules can be changed. So, I don’t blame 8 year stakers for wanting to grab a bigger share of the pie. Everyone wants to maximise their personal gains, nothing wrong in that. By all means leverage that extra voting power to get more rewards. But let’s not pretend it is to set right some injustice or for the betterment of the ecosystem.

1 Like
37

Hello Denis! I had a feeling someone would grossly miscategorize everything I’ve said, so I’m going to correct a few things😅 Thanks for taking the time to participate in the conversation regardless.

This article has shown that the reward rate is not proportionate or linear at all. I have provided the data to showcase this, although if you have contradicting data, I would appreciate if you provided it. However, simply saying, “no trust me, it’s fair - believe me” doesn’t make it so.

I’m not sure where you’ve seen that I think APY should be raised. That is actually the opposite of my standpoint - we have an issue with inflation, and a lack of tokens being burnt to compensate.

I’ve stated I believe APY should be re-scaled, incentivizing long term stakes over short term (as short term stakes are much more lucrative, given the lack of lockup for similar reward rates) - however this does not equate to increasing inflation or APY.

If you continued reading past where you quoted, you would see where I said this:

Furthermore, this is not indicated as a proposal, but rather a discussion - there’s not even a proposed change to come of this, the goal is to simply read the room.

It makes no sense to portray it as though this topic has been raised for personal gain.

Utilizing your own logic, the point could be made that you’re so aggressively defending short term staking APY due to the fact that you intend to maximize personal gains - recognizing the over incentivization associated with them. But I won’t.

So, I’d appreciate if you laid off on the unfounded accusations, and let me get back to fixing shit🤠

1 Like
39

Thanks @wbp, for your defence. Greatly appreciated.

I’m just going to leave a few quotes from the conversation here and people can judge if the issue is solely about reducing the APY for short term stakers or increasing it for long-term stakers.

Of course I do. I do not claim to be an investor for charitable purposes. The difference is that in my case I invested based on certain promises made. Of course, the NNS allows those promises to be changed, which I also knew, and played into my decision not to stake for the longest term.

1 Like
40

I got a notification you quoted me so I’ll just say that I stand by my statement but I respect your right to disagree.

1 Like
41

Given he quoted me, rather accusatorially at that, it is completely reasonable that I respond.

On the other hand, you had no part to play in this conversation, and quite clearly have interjected yourself to assert your moral high ground & frame a narrative of character. Very transparent - not a good look.

Surely you can appreciate the irony, right?

I responded to him because he asserted what my opinions are, yet when I showcase the irony in his logic you try to use it as some type of “gotcha” moment.

You quite clearly have a bias within this conversation, that I’m not going to entertain further. Stop trying to be a moderator.

You’re quoting other people, and portraying it as though it is my opinion, or as though there is some type of collusion.

This is beyond intellectually dishonest - I have started a conversation, and people have responded to it, with their own opinions.

Additionally, just because people agree that it is not proportional does not equate to increasing inflation (I’m not sure why I have to keep repeating this part).

42

Going to tag moderators so that the community doesn’t attempt to assert moderation over this thread, there’s probably a few comments that shouldn’t be here by now: @Ang @diegop

44

I quoted most who had posted, you included, to prove my point. Never suggested collusion, never thought there was any collusion going on, nothing dishonest about it.

I’ve made the point I wanted to make, take it or leave it, makes little difference to me.

1 Like
45

Your “point” was quite clearly that I raised this topic for personal gain, in an attempt to raise my own APY.

You then provided quotes from other individuals within this conversation to “prove” something you’re claiming about me.

It makes no sense to make these claims - so again, please lay off the unfounded accusations & let me get back to fixing shit🤠