Proposal: Inflation Update 3.0

Overview

Our second inflation update was submitted on 7/2/2021 and went live on 7/16/2021, which set max inflation to 40%, minimum inflation to 25%, and inflation rate of change to 100%. This resulted in a decrease of actual inflation down to 33.94% as of 12/13/2021. We are, however, trending over our projected supply by 10mm AKT or 6.2%. With the dynamic inflation function ready for our next testnet in early 2022, we are submitting one more proposal in 2021 to lower inflation and track more closely against our originally modeled supply between now and the implementation of mainnet 3, which will feature the dynamic inflation function.

Proposal of Changes

For our third proposal for inflation change (prior to testnet 3), we are proposing the following changes:

Inflation Rate Change: from 100% to 500% (increase rate of reduction)
Inflation Max: from 40% to 30%
Inflation Min: from 25% to 20%

These parameter changes are expected to take place in mid January, accounting for 1 week of discussions and 2 weeks for the proposal voting period.

Hey, would it be possible to view a visual comparison of the actual curve and the expected/desired curve?

Also, what does it mean to change the inflation rate change from 100% to 500%?

4 Likes

It allows a 500% yearly rate change. Without it it’d take a long time for the rate to decrease.

We have implemented an exponential decay curve for inflation in the next upgrade (per whitepaper), this proposal would be to get us in line with previous statements/expectations (the whitepaper).

1 Like

Maybe my math is wrong here, but this is what I have. Given the current supply is 122,326,907 (coingecko) and the current daily inflation decay is .0404% (6.06/150) then the new decay would be .202% (.0404*5). From there we can take the yearly inflation rate of 33.94% and subtract .00202 everyday and multiple the supply by ((1.3394/365)-.00202) for the new daily rate change. After 30 days the total supply is 136,208,183.6 with a standing inflation rate of 28.08%. Replicating the curve with the current decay ((1.3394/365)-.000404) brings us 136,469,818.2 with a standing inflation of 32.77%. Therefore this change would only reduce the potential supply by 261,634, far from the 10 million targeted. The alternate calculations modeled off of the base inflation instead of 1.x returns a similar number ~ 235,000. I think the mathematical issue is that comparing the 6% new decay should be in comparison to the would be current 1.2% decay. The alternate decay model brings us the expected 1.2% (.3394-.327684-.000404). Also, it isn’t in relation to absolute supply, but the inflationary curve. If you look at the alternate charts you can see the match vs a constant inflation rate (Vs. Baseline) and the compounding difference is around 1.8% and 9.8%, but the difference from these effects is in the 200k range, as expected.

3 Likes

Ok I get it, thanks!

1 Like

Hi Amrosa,

I agree with your assessment here. I also think we should be more aggressive. Specifically:
Inflation Rate Change: from 100% to 500% (increase rate of reduction)
Inflation Max: from 40% to 25%
Inflation Min: from 25% to 15%

This would immediately tamp down inflation from where we are now to 25% and decrease the min so that it buys more time for decreases between testnet and mainnet going live, which will likely occur in March (Feb 2022 testnet means mid/end of Mar 2022 for mainnet 3 to be live).

I understand that we won’t be able to make up the entire difference quickly, but the aim is to stop the difference from growing and slowly chip away over the next several months/year. Secondly, this initial adjustment will help soften the adjustments the dynamic inflation curve will produce once it goes live.

1 Like

EDIT
For posterity, here’s my final update prior to submission for voting:

Our second inflation update was submitted on 7/2/2021 and went live on 7/16/2021, which set max inflation to 40%, minimum inflation to 25%, and inflation rate of change to 100%. This resulted in a decrease of actual inflation down to 33.94% as of 12/13/2021. We are, however, trending over our projected supply by 10mm AKT or 6.2%. With the dynamic inflation function ready for our next testnet in early 2022, we are submitting one more proposal in 2022 to lower inflation and track more closely against our originally modeled supply between now and the implementation of mainnet 3, which will feature the dynamic inflation function. The dynamic inflation function that will be introduced in mainnet 3 will ensure that future inflation adjustments will not require additional governance proposals. Late Q2 2022 is the estimated go live time frame for mainnet 3.

Proposal of Changes
For our third proposal for inflation change (prior to testnet 3), we are proposing the following changes:

Inflation Rate Change: Keep at 100% since it looks like 100% is the max
Inflation Max: from 40% to 25%
Inflation Min: from 25% to 15%

These parameter changes are expected to take place in mid-late January.

2 Likes

I support this. I am curious what effect it will have on staking, presumably less tokens will secure the network in the short term. We will have significantly lower inflationary rewards, but it is also too soon to turn on the take rewards from usage. Would see more AKT on Osmosis and elsewhere, and less securing the network, presumably, but what about liquid staking…

A better way to solve the overshoot would be to implement a small progressive reward aspect. Where a certain threshold zone (lets say around 50k AKT, just random) has a declining reward percentage in a continuous manner (so that staking above it still results in more reward in absolute terms but relatively less than below it). Even if all large stakers would undelegate above the threshold and restake with a new wallet, it would reduce the overshoot due to the 3 weeks unbonding. Additionally, this would create more decentralisation by allowing smaller holders to relatively gain on large holders. This might not be possible due to the preset rules of the whitepaper though, not to mention the vote being dominated by the large stakers.

Do I understand correctly that this will decrease the amount of AKT received from staking from the current 40% to 25%?

If so, I really don’t like it.

I bought into Akash because I believe in it long term, and saw the high staking rewards as a good reason to just stake-and-forget. I thought it was a fair way to balance rewards vs the considerable amount of tokens vested by the team and early investors. A way for early, retail, buyers to sort of level the playing field.

I think it’s unfair and unwise to change these rewards after the fact. I also think, if we are worried about inflation, and its effect on the price of AKT, why not burn the tokens held by the team and early investors (sorry Greg)? Correct me if I’m wrong, but they are released every March and September.
What happened last September? A batch of tokens was released and the prices started falling very rapidly.

I think if you change the staking rewards, it is only logical that holders will unstake. Less incentive to keep it staked. So there will be more AKT free to sell starting in January, having a suppressing effect on the price. Then in March new tokens will be released; even more sell presure.

Conclusively, I think it’s unfair to early buyers to change the rules of the game. Second I think if the worry is that inflation has a detrimental effect on price, decreasing staking rewards is not the way to go.

The person posting above me had a luminous idea I think; declining rewards percentage. Smaller holders gain more tokens relative to large holders. I think that’s more fair.

Big fat feel-free-to-correct-me-if-I-misunderstand.

Cheers.

I appreciate that the lower inflation rate will result in us receiving a smaller number of tokens, but there’s more to it than that. The inflationary tokens also go to the whales that have staked, so you’re not really growing faster relative to them. You mention price, and I don’t want to really get into price action, but my personal assumption is that if fewer tokens are issued, they will each be more valuable.

Anyway, the proposal isn’t going to “change these rewards after the fact.” As @chenghiskhan pointed out, the change in the inflation rate will “track more closely against our originally modeled supply.” The inflation rate has always been temporary, part of this project involves inflation decay, so we shouldn’t have the impression that today’s inflation rate will remain long-term. We can see this outlined as early as the white paper, which discusses the maximum number of tokens and the inflation decay methods. Government proposals in the past have ensured that inflation is reined in and the project keeps on target.

So, I don’t think this proposal “change[s] the rules of the game.” Even if it did, everyone else has their own staked coins and are also getting proportional rewards. Also, the effect of altering inflation is more complex than just “I get less rewards, therefore I have less money each day”. But again, in conclusion, inflation decay has been an ongoing feature discussed in the white paper and in other public government proposals that have passed.

EDIT: As for burning other people’s bags… I would oppose that. If someone wants to voluntarily burn their own bags, okay, I guess that results in our bags being that much more valuable. But to burn an investor’s bag just seems unwise; surely there was some agreement between the Akash Network and the initial investors that controls what happens, and I doubt anyone is just going take it on the chin if we burn their assets. There are better ways to ensure that token unlocks go smoothly, such as reaching out to investors and talking to them about their intentions, etc. But that is unrelated to this particular point, inflation decay, because what is being done now is to keep our AKT supply and reward rate in line with the ideal model, not pump the bags/avoid a potential crash.

1 Like

Is there plan for a tokenomics post/medium article?

Initial supply distribution graph, vesting cliffs, release curves going forward and such?

Thanks

Greetings, I ve listened to the discord discussion, intresting that it was.
So with regards to the proposal and excuse me for being wary, does it have to do anything with the imminent 25M token unlock during Feb/Mar2022? I mean the timing of this proposal seems a bit off.
What @rowan290 proposed seems less stressful for us delegators.

Hi @Iherfel, welcome.

You can see the tokenomics in the white paper here:

Also the supply graph can be found here:

kb

1 Like

Hi @geostef,

The proposal doesn’t have any relation to any coming unlock. While I can’t speak for the people who created the proposal, I can explain the dynamics of the departure from the intended curve.

Basically, rewards have accidentally been too high recently (above the intended curve in the white paper, resulting in too many AKT being in circulation). This means that everyone who is already staking has benefitted equally (proportional to the length of time that they have been staking for) from the increased rewards.

To resolve the oversupply, the rewards need to be reduced (again equally to all) and the stakeholders who will be worst affected are the ones who are just starting staking now because they haven’t benefitted from the recent enhanced rewards.

Having said that, reduced rewards also have a number of benefits including reducing downwards pressure on the price and with options such as LPing and so much in the pipeline for 2022 I don’t think any newcomer is going to be disappointed anyway (NFA).

Hope that helps.

kb

1 Like