During the extensive debate around Prop 848, many NO voters argued that the adaptive inflation model was essential to the Hub security model and targeting 67% was essential. I would characterizeYES voters as strongly feeling that Hub has spent 2022 and 2023 inflating too many ATOM while trying to reach the target staking rate
The following chart of the historic staked ATOM percentage vs inflation can be interpreted as supporting this assertion.
In light of empirically observed staking behavior with the current InflationRateChange parameter, increasing the rate of change will hopefully both achieve the target staking rate more quickly. The author also speculates that it might lead to less net inflation as a whole.
Support. Higher responsiveness of monetary policy to bonding ratio changes will incentivize participation in staking. At present, whether you bond or unbond, the result of your actions takes a year to become a reality at which point you are facing a different economic situation. The slow rate of change disincentivizes staking participation. Desired inflation effects have to come to fruition faster.
It’s been almost 2 years since the inflation rate has been at the min or max threshold. Inflation rate change is also based on the distance of the bonding ratio from the target. Higher distance, higher inflation rate change. At 67% bonding ratio, inflation will hit 7% in like 25 years.
ah ok… The 1 year reference through me off as It felt like a reference to the original 13% max change so replied in kind to the same simplified model.
With distance factored in changing rate change to 3 wont make much of a difference…(8 years instead of 25 :P)
A rework of the NextInflationRate function is needed instead.
In reflecting on the inflation reform debate, we share the sentiment that the focus has been disproportionately placed on individual factors rather than the holistic evaluation of the inflation formula. A comprehensive public debate on the formula, with due consideration for the interplay of min and max caps within this framework, would provide a more nuanced understanding of the dynamics at play.
While adjusting the InflationRateChange parameter may offer a solution, we advocate for pausing the debate on min and max caps and redirecting the discourse towards a thorough examination of the entire inflation formula. By addressing the formula as a primary driver, we can foster a more informed and nuanced discussion that ensures responsiveness and alignment with the broader goals of the Cosmos Hub.
We look forward to an inclusive and comprehensive dialogue that considers all facets of the inflation model, confronting competent parties with different visions and experience in economic modeling. We will be happy to be part of these discussions as well as existing persons who contributed valuable content to this debate such as @effortcapital and Blockworks with their inflation reform proposition. At Govmos we suggest to extend the scope of the formula by including the ratio of LSTs in the parameters to increase responsiveness in certain situations requiring faster slopes. The idea is to create a model that increases the rate at which the formula reduces inflation toward the minimum. We would characterize economic activity by using the LSTs ratio as a primary demand driver signaling the need to lower inflation with greater amplitude. We are in the process of building such model to highlight the benefits it could bring.
This inflation reform should take the necessary time for proper multi-party agreement and avoid repeating the mistakes of 848 by going prematurely on-chain without proper pre-proposal consensus.
I agree. At Umee we came to the same conclusion: the current inflation rate is too slow. Especially when the current inflation is not at the min/max.
A month ago we changed the formula to speed up the inflation rate:
I wonder what is the correlation between a the desire of an actor to sell, i.e. asset owner’s needs and the amount of an asset issued per day to a given market. My guess it roughly = 0
I have to throw this out here: the inflation rate change parameter should be set to -3.0
Why?
I think the incentive scheme right now is entirely wrong. The algorithm is incentivizing unstaking and for that reason we can’t achieve the bonding target. What do corporations do with their stock? If there is a bull market and their stock is expensive, they sell stock in the euphoria. If there is a bear market and their stock is cheap, they buy it back into the fear and reduce the amount of stock out there. This makes sense - buy low and sell high. Cosmos Hub should be doing the same thing.
What is the ATOM algorithm doing now? Let’s establish a relationship between staking ratio and bull/bear market conditions. If we are in a bull market, what will people do? They will unstake their ATOMs and sell them to lock in gains. If we are in bear market, what will people do? They will buy ATOM and stake it to get yield and thus lower their entry point over time if they made a mistake timing the entry well. As such what should the algorithm do? It should go against the market. When people are unstaking and selling ATOMs, it should decrease the inflation, punish the stakers for unstaking and make the ATOM scarcer as people are dumping. When people are buying and staking ATOM, it should increase the inflation and incentivize them to keep staking it. Let’s say, we are in a bear market and the price is tanking hard. What do people do then? They start to unstake and sell their ATOM. Well that is when the inflation should go to zero and make ATOM scarce thus stabilizing the price. If people are buying and staking ATOM, clearly there is demand for it so that is when you want to be selling more ATOMs into the demand.
For example, right now when we are below the staking ratio target, the inflation should start decreasing towards the min inflation. People are not staking and as such they get less rewards. How do people get higher rewards? They need to start staking. When the bonding ratio target is reached, then inflation is starting to increase towards the max inflation target. We should be rewarding with inflation when the staking ratio target is achieved. And we should be disincentivizing with deflation when the target is not achieved. We should be doing the opposite of what we are doing now.
Think about it. What is your reward for staking right now? Lower rewards. Who the hell wants to do that? So big ATOM holders are incentivized to keep some of their ATOM unstaked in order to keep the bonding ratio just below 2/3rds. If going over the target gave them the highest inflation, people will stake all their ATOMs happily and be well over the target. But even worse, just as people are unstaking and dumping, we are putting more ATOM out there to be dumped further hurting the price. We shouldn’t be doing that.
We have been doing this for 2-3 years now, ATOM has a very unique monetary policy compared to all other coins who mostly have fixed inflation schedules. In the real world, ATOM’s monetary policy is not working at all as it was designed prompting this reevaluation we have right now. I think a big part of the problem (in addition to the 20% max inflation target which was ridiculously high) is that inflation rate change parameter is positive instead of negative.
Repeatedly, we’ve stressed the importance of approaching the reform as a broad discussion covering all four parameters: InflationRateChange, InflationMax, InflationMin, and GoalBonded. To illustrate the necessity of this holistic perspective, we’ve created a dedicated research topic to the subject, presenting a straightforward inflation model: The x/mint Module and understanding the Inflation Rate Calculation.
Making adjustments to the parameters individually, without considering the entire set, lacks coherence. These parameters are interdependent, and their evaluation should be done comprehensively. Our approach would have been to discuss the model first and then subject the conclusions to a vote. We express regret for the current trajectory of the debate, which has focused on individual parameters while overlooking the overarching architecture of the inflation formula.
The ongoing discussion serves as a vivid example of this oversight. Setting the “InflationRateChange” to an optimal level hinges on defining appropriate levels for the other three parameters.
Utilizing our model (source: Google Sheets), we observe that adjusting “InflationRateChange” to “3” with the proposed “InflationMax” at “0.1,” “InflationMin” at “0,” and “GoalBonded” unchanged would lead to 0% inflation in about six months if the bonded ratio converges toward 75%. This scenario would necessitate a prompt adjustment of the “GoalBonded” parameter to avoid a sudden shift to 0% inflation, altering the economic model drastically and likely damaging the chain’s reputation durably.
We insist, it is crucial that we engage in a comprehensive debate on all four parameters, constructing a unified vision that garners consensus from the majority of participants, rather than blindly modifying parameters one by one.
Alas, i dont think anyone that put these props up cares. Im serious. I.E asked a question:
I wonder what is the correlation between a the desire of an actor to sell, i.e. asset owner’s needs and the amount of an asset issued per day to a given market. My guess it roughly = 0
No one will try to take this seriously, as there is some powerful obsession with believing that spending 1-2 years trading crypto makes a person suddenly understand economics, twitching inflation, etc
I doubt that there has been a single extensive research or mathematical modelling (even as simple as an a/b scenario analysis) done prior to the submission of any of these props.
I really appreciate the fact someone is posting some numbers and charts, alas, it seems a dead end. End of the day, its about whose whistle is bigger. Which is sad, but thats precisely whats happening now.
PS. answer was for @Govmos sorry for the confusion
Even though I’m in favor for a change in the current inflation model, I’ll vote NO on this prop for the same reasons expressed on the two latest replies above me.
We need to consider the entire set of parameters and try different models/simulations. Then, show the different results to the community so we can do an informed vote. Let’s have a better consensus and overall solution before putting props on-chain.
I could be in favor of all types of changes, even the one already changed recently or exotic new ideas, as long as it is part of a complete solution that make sense as a whole.
