[MRC-29] ATOM Market Parameter Adjustments

[delphilabs]

Summary

This proposal aims to adjust two parameters related to the interest rate model of the ATOM market in the Red Bank with two objectives: 1) making the ATOM interest rate less volatile and 2) making the market more efficient. The two parameters are slope_1 and the optimal utilization rate. In the next sections these parameters will be explored in more detail, as well as the rationale for the suggested adjustments.

Motivation

As mentioned above, the current proposal has 2 objectives:

1. Making the ATOM interest rate less volatile.
2. Making the market more efficient.

Let’s explore.

With the recent Mars Farm stATOM/ATOM vault cap increase, ATOM demand has spiked, making the ATOM interest rate increasingly volatile. The reason why the interest rate has become so volatile has to do with the interest rate model used by Mars. Specifically, Mars uses a 2-slope interest rate model that targets a certain utilization rate for each market. When the utilization rate increases above that target, the interest rate will start increasing at a faster pace. This mechanism is used to incentivize the availability of liquidity within each market, which is important to be able to honor user withdrawals and to ensure the liquidations system works smoothly. The next figure summarizes how this mechanism works. As can be seen, when utilization goes above the optimal utilization, the interest rate starts increasing at a faster pace to both incentivize users to deposit and borrowers to repay their loans, which should increase liquidity within the market.

With this background, it’s easier to understand the ATOM interest rate volatility. In short, what’s been happening is that the market has tended to go above the optimal utilization, where the accelerated interest rate (slope_2) starts to kick in. As can be seen in the figure above, any change in demand that happens above optimal utilization is going to have more pronounced effects on the ATOM interest rate, which is what has been happening.

The above dynamic suggests that the ATOM equilibrium interest rate is contained within the slope_2 curve, which is causing the current volatility. As such, we suggest two adjustments:

1. Increasing slope_1 from 0.15 to 0.20.
2. Increasing the optimal utilization from 0.6 to 0.7.

The above two adjustments will have the following effects:

1. The change the interest rate at optimal utilization from 15% to 20%. As mentioned above, one of the reasons for the volatility is that ATOM demand is taking the interest rate above 15% (current rate at optimal utilization), where slope_2 starts kicking in. Increasing it to 20% will allow the interest rate to operate on slope_1 (significantly less volatile) for a broader range of interest rates. With this adjustment, we intend to capture the equilibrium interest rate within slope_1.

2. The rationale for setting it at 20% is that it has consistently been rising above 15% and we believe the equilibrium rate will be somewhere closer to the current ATOM staking APR of 21%. Previously we suggested 23% as we anticipated the staking rate of 21% would help reach an equilibrium - however this original proposal received some negative sentiment on the forum so we have reduced the increase from 23% → 21%.

3. The optimal utilization increase will allow for more ATOM to be borrowed before slope_2 starts kicking in. This helps both with the interest rate volatility and market efficiency.

Risks

The most important risk worth highlighting is related to increasing the optimal utilization threshold. As mentioned above, the optimal utilization tries to mitigate the risk of the protocol running out of liquidity in any given market. The lower the optimal utilization is, the more conservative the protocol is and vice versa. As such, an increase in the optimal utilization rate makes the ATOM market more likely to suffer from illiquidity scenarios. Having said that, we think 70% is still conservative and the rewards for the adjustment are worth the added risk.

Implementation

This is a signaling proposal, not an executable proposal.

The Mars smart contracts on the Osmosis chain are currently controlled by the Builder Multisig address. If this proposal passes, the builders will utilize their multisig to make the necessary parameter change.

Disclaimers/Disclosures

This proposal is being made by Delphi Labs Ltd., a British Virgin Islands limited company. Delphi Labs engages in incubation, investment, research and development relevant to multiple ecosystems and protocols, including the Mars Protocol. Delphi Labs and certain of its service providers and equity holders own MARS tokens and have financial interests related to this proposal. Additionally, Delphi Labs is one of several entities associated with one another under the “Delphi Digital” brand. Delphi Digital’s associated entities and/or equityholders or service providers of such entities may hold MARS and may have financial interests related to this proposal. All such entities, service providers, equity holders and other related persons may also have financial interests in complementary or competing projects or ecosystems, entities or tokens, including Osmosis/OSMO. These statements are intended to disclose relevant facts and to help identify potential conflicts of interest, and should not be misconstrued as a complete description of all relevant interests or conflicts of interests; nor should they be construed as a recommendation to purchase or acquire any token or security.

This proposal is also subject to and qualified by the Mars Disclaimers/Disclosures. Delphi Labs may lack access to all relevant facts or may have failed to give appropriate weighting to available facts. Delphi Labs is not making any representation, warranty or guarantee regarding the accuracy or completeness of the statements herein, and Delphi Labs shall have no liability in the event of losses or damages ensuing from approval or rejection or other handling of the proposal. Each user and voter should undertake their own research and make their own independent interpretation and analysis of all relevant facts and issues to arrive at their own personal determinations of how to vote on the proposal.

[Elster]

Modifying the curve below the optimal point feels unconnected with the rationale of stabilising the atom interest rate, more in respect of extracting profit. If you increased the current utilization max point and not modify the the curve slope, I’d argue you’d achieve the same thing. Mars is already taken a large % APY from depositors and Mars could become uncompetitive in respect of borrowing ATOM. Just my thoughts … thanks

[deif]

If the minimum deposit rewards are going to be increased to higher than statom yield, I think as part of the build the community pool should start being replenished by skimming a portion of the statom/atom farm rewards on osmosis to buyback Mars. The apy is far higher than necessary and if you’re planning on holding a significantly larger portion of circulating ATOM/STATOM the security of the protocol should be considered.

[markwatney]

@Elster

Modifying the curve below the optimal point feels unconnected with the rationale of stabilising the atom interest rate, more in respect of extracting profit

Can you elaborate on this? I’m interested in what you see as extracting profit, and the pros/cons as you see it.

FWIW I agree, I think the optimal_utilisation_rate change will extract more profit for Mars (and depositors), but not at the expense of borrowers. It does so by being more capital efficient.

I think the issue if we only modify O_I_R don’t modify slope_1 is that the market is currently happy to pay up to ~21% borrow rate (because of stAtom/Atom farm), if the kink is @15% we get very volatile interest rates.

Personally I think moving the target rate to 23% is better for borrowers as well. Currently the rate jumps from 15% to 40+% in the space of a few hours, which makes it difficult for ATOM borrowers to understand what their interest rate will be

[Elster]

HI Mark, I think my observation was that the 'optimal_utilisation_rate" change would nett more profit for Mars, I was unclear that this was not at the expense of borrowers. The deposit rate for Atom on Mars is quite low IMHO and Mars is taking around 30% of the equivalent staking yield already. My concern is that if the deposit rates remain like this and the cost to borrow increases it will make Mars a quite expensive proposition? The Atom-stAtom farm is working well, low risk and is pretty much guaranteed to be “sold out” with the next cap increase…I wonder why!
I’m jumbling up a few things here so to get back to my main observation, and I’m happy to be corrected.

1. the cost to borrow is already quite high, is there a risk of pricing mars out of the market?
2. mars profits clearly increase with the proposed change - can you explain how this is not at the expense of borrowers?

cheers

[markwatney]

Yea agree I think the atom-stAtom farm is the main driver of borrowing atom, so I think we can expect the market interest rate (for borrow) to basically balance out somewhere close to 21%.

Currently, borrow rate is about ~14.5%, but likely this is going to increase if the caps for the statom farm are raised and more people borrow atom to use in the vault. This means that the rates will get really volatile again, which happens when they move into slope_2 (>15%). Previously we saw rates changing from 15% to 40% within an hour or 2.

So regarding point 1)

• Personally I think the market will find the equilibrium rate based on perceived risk of the vault, not much we can do about that. If its cheaper people will go elsewhere causing rates to go down. Likely the vaults will drive the price to be close (but lower) than the stAtom rate.

regarding point 2:

• I think slope_1 provides more security to borrowers as we should be less likely to get the large spikes to 40+% interest.
• It does not mean borrowers pay more per se, it just means the borrow rate changes less rapidly while under 23% (instead of 15).
• increasing optimal_utilisation_rate should impact depositors positively, borrow rate should not be impacted much. For example, if 70% of the deposited atom is borrowed, paying 15% interest, ( 0.15 * 0.7 = effective rate of 10.5% before mars fee) that is a greater return for depositors than if only 60% is lent out ( 0.15 * 0.6 = 9% before mars fee).
• Mars takes 20% of total interest payments, so by increasing the optimal_utilisation_rate we should see an increase in the amount of total amount borrowed, which means greater profit for mars at the same rate

Let me know if this doesnt make sense, I’m kinda just thinking out loud here . These changes are subtle but can be surprisingly complex when you think about the different impacts and the game theory of it all

[0xphilipp]

Mars is already one of the most expensive places to borrow ATOM and this proposal further increases borrow rates by 30% and alienates borrowers. Borrowers are the income source for MARS and they should not be milked as a cash cow.

To make the market more stable, an increase in the slope of the borrow rate is not required. A simple increase in the optimal utilization already makes the borrow rate more stable due to a higher max borrow rate at the utilization goal.

Targeting a higher than staking rate for provisioning ATOM makes no sense, as liquidity benefits are ignored. Currently the usage of the leverage vaults already costs 30-50% of the APY for lenders / mars and this is a high number when looking at the additional risks borrowers take. Any kind of increase in the borrow price will negatively impact borrow volume. A higher borrow rate will not make for a higher “income” for mars, as capital will outflow from mars. Directly depositing into the apollo vaults make 23%, just now looking at my position it also only made 23% APY even with a high leverage. That means the additional risk in taking a leverage is not worth it (as is today without the increase)

So my proposal would be to set the slope_1 only to 0.175 and the optimal utilization to 0.7. The proposed 0.23 is an increase in the borrow price of ~30% which is unacceptable.

Any borrow rate adjustment to the upper side that involves assets in locked LPs must require a minimum 14 day delay after passing the proposal, so that people have the choice to remove liquidity. Otherwise this is a clear change of signed contracts by users and should be voted NWV.

[markwatney]

@0xphilipp

Any borrow rate adjustment to the upper side that involves assets in locked LPs must require a minimum 14 day delay after passing the proposal, so that people have the choice to remove liquidity. Otherwise this is a clear change of signed contracts by users and should be voted NWV.

This is a good point - its difficult for people to react given the 14 day unbonding and only 3 days on chain voting period.

I also agree with your sentiment that borrowers are a crucial part of Mars - they are the paying users, and we need to attract as many as possible with attractive rates.

@0xphilipp

Mars is already one of the most expensive places to borrow ATOM and this proposal further increases borrow rates by 30% and alienates borrowers

I think this is where I differ - From my perspective, borrowers set the market rate. If the rate is too high, borrowers repay and utilisation and rates go down. If the rates are deemed ‘low’, people borrow and the rate goes up.

ATOM is ‘expensive’ to borrow because of the Apollo vaults - there is a low risk high yield opportunity with a good UX that is driving demand for ATOM.

IMO, the borrow rate is being suppressed by the vault caps and if caps were to be raised (likely) we will see borrow rate increase. I think we have seen enough to suggest the borrow rate the market is willing to pay on mars is >15%, which is why I am for raising slope_1.

Having a slope_1 below the natural market rate leads to a negative user experience for borrowers as the rates are very volatile once they move onto slope_2.

Maybe 0.23 is too high and 0.175 is a better value - I would argue this is probably still a little low as we want to prevent slope 2 being hit at the natural demand as this curbs usage

I think it is fair to expect the yields on the atom farms to compress until you are only capturing risk premium.

[0xphilipp]

@markwatney

IMO, the borrow rate is being suppressed by the vault caps and if caps were to be raised (likely) we will see borrow rate increase. I think we have seen enough to suggest the borrow rate the market is willing to pay on mars is >15%, which is why I am for raising slope_1.

I think people are not willing to pay > 15% to borrow ATOM and the high rate has been happening due to these factors:

• raise of stATOM-ATOM caps led to people apeing borrowing through the farm GUI without them seeing the real borrow price increase.
• low MARS price, reduced the incentives on the lend side, so people started withdrawing ATOM, reducing the max utilization point, increasing the borrow APY. And I expect that the total borrowed atom has been reduced since the statom-atom pool has been filled.
• people can borrow ATOM for 14% on nitron and 10% on umee
• stride is reducing incentives by 10% in june

I don’t have the data available but my thesis should be verifyable through the data delphi is tracking.

I understand that the borrow rate for the max utilization should be higher than the price the market is willing to pay, but I think this is reached with 15%, as the current utilization is horvering just 2-3% below max utilization with 14-15% borrow rate and your thesis would mean people would move that closer to the max utilization.

[markwatney]

• people can borrow ATOM for 14% on nitron and 10% on umee

Yea but they don’t have a Credit Manger

But I think you make some good points, I tend to agree 0.23 might be too high. In it’s current form I would still vote yes for this proposal but think I would prefer to see slope_1 a little lower.

[dancreee]

I think we should go ahead with the Increase of the optimal utilization from 0.6 to 0.7, while re-assessing what the rate/slope_1 at optimal utilization should be, and perhaps proposing for something smaller since there is some differing opinions on it’s impact.

My thoughts on optimal utilization are that increasing the rate gives users better borrow rates at higher utilisation rates, at the expense of the protocol taking on more risk that a market could reach around 100% utilisation which would block things such as user withdrawals and liquidations. Overall it seems like everyone is supportive of this change.

---

My thoughts on slope_1 is that ultimately it is the borrowers who decide the borrow rate, not the slope_1 or optimal_utilization rate params, and that this change has both pros and cons for users.

Pro - the borrower can expect less volatility on their variable borrow rate if the slope_1 is set higher above the expected market equilibrium rate. This is because there is more wiggle room for the market to recognise they can get better rates elsewhere and potentially even arbing those rates, before hitting the more aggressive slope 2 rates.

Con - Although increasing slope_1 does not necessarily increase the borrow rate, as stated an efficient market should find equilibrium - the market overall could be less attractive to borrow against (but more attractive to lend against) when a utilisation rate is high.
So for an in demand asset, or a whale, borrowers could find the rates unattractive and potentially go elsewhere - but in an efficient market lenders should find these situations more attractive also, in which case they would lend at greater than market rates until again the market reaches equilibrium.

---

Final thoughts, if anything, I think increasing slope_1, assuming an efficient market that finds market equilibrium, is actually net beneficial to the borrower, while at the same time being disadvantageous to the lenders and protocols revenue.

This is because the market should find/arb the same borrow rates, meaning borrowers experience less volatility by rarely falling inside of slope 2.
Meanwhile the lenders and protocol will be borrowing out less of their market, (less market utilisation at same borrow rate) meaning those markets are less efficient/will earn less in borrowing fees. The counter to this would be that borrow rates in Cosmos are not particularly efficient at arbing rates right now.

[delphilabs]

Taking the community feedback on board (while acknowledging we are still consistently breaking into slope_2) we have updated the proposal to change slope_1 from 0.15 → 0.2 (previously 0.23).

The proposal will be left on forum for at least an additional 48 hours to collect any additional feedback in relation to the updates, before assessing if sentiment is positive and an on chain vote is likely to succeed.