[ARFC] MaticX SupplyCap Increase Polygon v3

[Llamaxyz]

---

title: [ARFC] MaticX SupplyCap Increase Polygon v3
author: @Llamaxyz
created: 2023-03-09

---

Summary

Llama proposes increasing the MaticX Supply Cap on Polygon v3 from 8.6M unit to 10.75M units.

Abstract

The SupplyCap for MaticX was increased from 6.0M units to 8.6M units. Within 24 hours of the upgrade going through, the utilisation of the MaticX reserve has reached 99.30%.

This proposal seeks to increase the SupplyCap by 25% to 10.75M units. Increasing the SupplyCap will enable users to deposit MaticX, earn SD rewards and enter the recursive MaticX/wMATIC yield strategy.

Stader Labs is currently offering SD rewards to user who deposit MaticX. This proposal seeks to encourage Stader Labs to continue distributing SD rewards by increasing the SupplyCap and enabling more MaticX to be deposited into Aave.

Motivation

Over the previous months, Llama has been working with various communities to craft favourable conditions on Aave v3 Polygon to facilitate the creation of several yield aggregation products.

The below proposals details are applicable to MaticX:

• Proposal to add MaticX to Aave v3 Polygon Market
• ARFC Aave v3 Polygon wMATIC Interest Rate Update
• ARFC SD Emission_Admin for Polygon v3 Liquidity Pool
• ARFC stMATIC & MaticX Emission_Admin for Polygon v3 Liquidity Pool
• ARFC MaticX Polygon v3 Upgrade

This proposal is a continuation of the above work.

After executing the ARFC MaticX Polygon v3 Upgrade proposal, within 24hours, the SupplyCap of 8.6M has nearly been reached.

With reference to the new ARFC Aave V3 Caps update Framework the preferred forward path is to implement several upgrades that gradually increase Aave’s MaticX exposure over time.

A 25% increase in the SupplyCap was selected as it is slightly more than 50% of Circulating Supply on Polygon, 10.58M units. This is loosely aligned with the lower bound of the aggressive Supply Cap methodology provided by Gauntlet after excluding trading volume considerations reflecting due to the nature of the asset being a Liquid Staking Token.

Llama will incorporate input from either or both risk service providers before progressing to submit and on-chain proposal.

Specification

The following risk parameters being proposed by Llama.

Ticker: MaticX

Contract: polygon: 0xfa68FB4628DFF1028CFEc22b4162FCcd0d45efb6

Parameter	Current Value	Proposed Value
SupplyCap	8.6M units	10.75M units

Copyright

Copyright and related rights waived via CC0.

[sakulstra]

What would be the max supported by risk providers(@gauntlet, @ChaosLabs)?

With 2.6M being supplied within 24h, i guess it’s safe to assume another 2.1M will be reached in a matter of day(s) as well. With:

• maticX oracle being a calculated price feed and
• maticX beeing redeemable in 2-3days

i would imagine that perhaps circulating supply(of maticX) as a metric doesn’t matter so much in that case?

On a sidenote, 50% of current active proposals on aave governance are cap adjustments and i think it would make sense for risk providers to bundle them to reduce governance overhead.

[Llamaxyz]

Hi @sakulstra,

This proposal was shared with Chaos Labs and Gauntlet prior to publication on the forum. The proposed Supply Cap figure was shared soon after the AIP was executed yesterday when we noticed how the market was responding.

We are happy to include all other pending Polygon v3 proposals within this AIP submission. With the introduction of additional rewards beyond SD to start flowing soon, it is reasonable to expect an even higher volume of parameter adjustments in the coming months.

If a higher Supply Cap than 10.75M units can supported by the risk service providers, that would be great. We would also advocate for the respective methodologies to be updated reflecting this approach as well.

This proposal was drafted based on public methodologies we strongly encourage more transparency around how risk parameters are determined.

[Pauljlei]

As of last week, below are Gauntlet’s max recommendations, following our methodology. The proposed value of 10.75M MaticX on Polygon V3 is above Gauntlet’s aggressive recommendation. I will relay this proposal to our team and come back with refreshed numbers but wanted to provide this quickly to the community in the meantime.

[MatthewGraham]

Hi @Pauljlei,

Can you please share what is the limiting factor in calculating the 9,590,000 value ?

Is the Polygon MaticX Circulating supply the limiting factor here, as this value exceeds the multiple of daily trading volume on Polygon stated in the methodology and the price manipulation consideration is not applicable given this reserve uses a calculated oracle feeds.

The increase in Supply Cap correlates to the increase in circulating supply of MaticX across both networks, most notably, MaticX did migrate from ETH to Polygon during this time.

• 2.6M units Supply Cap increase
• 1.56M Polygon Circ. Supply increase
• 0.39M ETH Circ. Supply increase

Most of the increase in deposits into Aave was from a single address, 1.63M units.

Liquidity appears to have not notably changed during February…

[ChaosLabs]

Using our supply cap methodology, Chaos Labs supports an increase of 2X the current cap.
To follow best community practices, we do recommend waiting for Gauntlet’s updated analysis and recommendations before moving forward.

[shanmug]

Thank you to all the stakeholders (@sakulstra @Llamaxyz @MatthewGraham @ChaosLabs @Pauljlei) for their comments above.

MaticX holders on the ETH network are keen to deposit on Aave and we would love for the supply capacity to be high enough to accomodate the inflow. ~25M MaticX lies on the ETH network and this is a significant opportunity to bring it over the Polygon network and contribute to the DeFi ecosystem

Stader has already deployed SD rewards to incentivize MaticX deposits into Aave but the limited supply cap leads to sub-optimal utilization of the rewards.

It would be great if the risk managers can take this into consideration and have a meaningful increase in supply cap so that we can continue to nurture this mutually beneficial partnership.

Thank you

[MarcZeller]

Hello,

The ACI considers this ARFC too conservative and is supportive, as @ChaosLabs said, of a 100% increase of the current supply cap.

“shanghai” equivalent is already live on Polygon. Any MaticX can be redeemed at fair value on the L1 after a 3-day delay.

It makes no sense to us to consider it less liquid than MATIC itself, the worst case secondary liquidity scenario (excluding architectural & Smart contract risks) would result in a 3 days “temporary virtual excess debt” event in Aave (virtual because maticX is a calculated price feed) while MM would make a killing arbitrating PoS network secondary liquidity inefficiency.

every MATICX is a potential MATIC in 3 days that earn staking revenue. That’s it.

Let’s not slow growth, LSDs are a real revenue vector for the Aave Protocol.

[MatthewGraham]

Hi @ChaosLabs,

Can you please walk through how this 2x increase figure was calculated ?

Liquidity has not meaningfully changed and the methodology for LSTs does not include underlying native token liquidity. If it did, then the wMATIC Supply Cap is notably higher again. To be specific, what I am seeking is the quantitative analysis to be shared, ideally so I can calculate a similar value. As mentioned in the original post, Llama followed the public methodologies as best one can based upon the detail provided on the forum.

If we have reached the rule of thumb 2x increase number, it would be good to know what this has been scaled down from.

As Stader Labs has highlighted, there is notable MaticX supply on other networks. It is reasonable to expect that a significant portion of the 8.6M units increase will come from a select few whales.

[ChaosLabs]

We used our supply cap methodology here.

We first find the amount eligible for liquidation given an extreme drop in the asset price using Chaos Labs’ Risk Explorer . We looked at a 15% price drop for MaticX, as a depeg scenario.

We then look for the maximum amount that can be liquidated while keeping the slippage below the Liquidation Penalty, using the 1% LP for the mode category.

Inserting this data into the formula below (taken from the publicly shared methodology) yields the recommended cap.

1. Evaluate the ExtremeProfitableLiquidationRatio:

This ratio represents how many times the ExtremeLiquidationAmount can be profitably liquidated had it all been liquidated at once. Requiring 90% of an asset’s liquidated amount will be liquidated at once is a defensive approach and can be set differently given changing risk appetites.

2. Set R = min(ExtremeProfitableLiquidationRatio, 2)

3. RecommendedSupplyCap = CurrentSupplyCap * R

---

It is important to note that in every recommendation we make regarding caps, we also look at qualitative data regarding the specific assets (such as redeemability in this case) and do not rely solely on the above methodology.

[MatthewGraham]

Thank you for the feedback. This is really good, because now we can discuss and it will lead to improved proposals on the forum. I genuinely appreciate the time and effort that has gone into replying to my comment. I hope the below reflects my keen interest in learning more in effort to improve the quality of the proposals coming to the forum.

@ChaosLabs

We looked at a 15% price drop for MaticX, as a depeg scenario.

As Aave uses a calculated oracle feed, how would this depeg practically occur ?

The MaticX oracle, like GHO’s, is decoupled from spot pricing. This is a key point of difference for LSTs with an oracle derived from base asset oracle plus network rewards. The model as presented is more applicable to stable coins with floating spot price derived oracles.

Within the remainder of the math, we have an arbitrarily defined 0.9 parameter on the denominator of the ExtremeProfitLiquidationRatio which if set lower, ie: lower percentage of assets liquidated amount will be liquidated, leads to a higher overall value. I would really like to see the analysis of where this 0.9 figure comes from given how sensitive the model is to this input.

The forumla in simpler terms:

The rule of thumb, the supply cap can not be increased more than 100% is hard coded by limiting the multiplier of the existing cap to be min (ratio,2).

With little historical data in general and zero data for when Liquidity Mining is ongoing, these arbitrarily selected parameters are essentially driving how risk parameters are defined.

If we exclude the rule of thumb, then:

The 2,070,000 value is from Chaos Labs dashboard. I would like to suggest we over look the rule of thumb here and push for Supply Cap of 21.56M units. However, I don’t want to be drawn into a debate, or upset anyone, and would prefer to move quickly. 17.2M units v 21.56M units is a notable difference.

I’ll let Chaos and/or Gauntlet come back on there preference.

We have this chart with historical data which indicates what would happen, if the oracle was upgraded to a spot price based feed, and MaticX decoupled from wMATIC by 50%. We can tweak this to 15% and it will help with future proposals. That is a clean take away for me to improve the quality of the proposals.

Beyond this, pulling 1inch or Cowswap data is not quick to setup, it requires placing frequent queries and archiving the data to build a historical data library, but ultimately it would be great to have a SupplyCap v time chart. The data fetching is the only bit we need to build out to generate a SupplyCap v time chart dashboard. I’ll see what can be done on this front as it is a great tool to have. We can even share the feed for others to use.

The arbitrary parameters built into the formulas should undergo sensitivity analysis and in my opinion, I think we should consider conversation v aggressive risk tolerances. This lower and upper bound approach is a good way of indicating the influence of the subjective assumptions feeding into the model.

[ChaosLabs]

We do not recommend increasing by more than 2X the current supply cap.
As stated in the methodology, we assume that additional supply and borrow amounts will be distributed similarly to the current supply and borrowing activity. Therefore, we wish to limit the increase to 2X the current cap to accommodate any unexpected borrower and market behavior deviating from the above assumption.

In the methodology, we also discuss the 0.9 figure:

@ChaosLabs

Requiring 90% of an asset’s liquidated amount will be liquidated at once is a defensive approach and can be set differently given changing risk appetites.

It was chosen as a conservative limit, not completely arbitrarily, as you suggest in your comment. The community can decide on a less conservative approach, which may lead to higher caps. However, we do not think that is necessary as even with the current model, we are not limiting cap increases across the board.

[Llamaxyz]

Hi Everyone,

We have been in touch with Chaos Labs and now proceeding to prepare an AIP to increase the Supply Cap from 8.6M units to 17.2M units in line with earlier feedback above.

After this proposal goes through, we intend to encourage proactive management of the Supply Cap and in doing so, work with other contributors (Marc, risk service providers) to consolidate the number or proposals being submitted to governance.

[Pauljlei]

Gauntlet Recommendation

After evaluating MaticX’s potential for price manipulation, DEX liquidity, and circulating supply, we recommend a conservative supply cap of 6.6M and an aggressive cap of 11.1M.

Gauntlet agrees to use the discounted MATIC liquidity and volume for cap calculations because MaticX is redeemable. The limiting factor for the cap recommendation is the MaticX circulating supply on the Polygon network. We recommend against having more than 50% of the circulating supply being deposited to Aave, even though MaticX can be migrated from Ethereum. If the community believes that using the 50% circulating supply as the upper bound for the LSD supply cap is too conservative, they may vote so, but this would signal a preference towards growth over risk.

[sakulstra]

After evaluating MaticX’s potential for price manipulation, DEX liquidity, and circulating supply, we recommend a conservative supply cap of 6.6M and an aggressive cap of 11.1M.

How would a maticX price manipulation currently affect aave?
As I understand the oracle is calculated as MATIC/USD * exchangeRate, so even if 2ndary market price would be volatile aave should not be affected as the oracle assumes 1:1 peg for MATICx claimable to MATIC?
So price manipulation risk comes from Matic price, not maticx price, right?

If the community believes that using the 50% circulating supply as the upper bound for the LSD supply cap is too conservative, they may vote so, but this would signal a preference towards growth over risk.

I think this is not such a binary question. With e.g. (w)stETH where price is based on stETH 2ndary market prices I think it’s perfectly reasonable to base assumptions on potential stETH price market manipulation. With maticX perhaps a more aggressive stand is reasonable?

[Pauljlei]

Yes, the price manipulation risk comes from MATIC liquidity. Given MATIC liquidity, the cost of manipulation is not a limiting factor. Circulating supply is the limiting factor.

The circulating supply threshold is another criterion that’s independent of price manipulation calculation. We do want to be transparent about the assumptions and tradeoffs here, though (increasing the threshold would be more aggressive).

[shanmug]

Thank you for your inputs, @MatthewGraham @ChaosLabs @Pauljlei @sakulstra @MarcZeller

Given the unique nature of MaticX being ‘redeemable for Matic in under 3 days’ coupled with the safety of using a calculated price feed (which cannot be manipulated by market forces) we believe that a higher supply cap for MaticX is justified.

With respect to the concern on circulating supply cap, there is an increase in the circulating supply by 50% over the last couple of days, reaching currently 30M. With an increase in the supply cap on Aave, we believe more MaticX holders on the ETH network can move over to the Polygon network and participate in DeFi via Aave.

We hope this goes some way towards assuaging any risk concerns the community may have.

[Llamaxyz]

Hi Everyone,

To summarise, with the additional MaticX now migrated from Ethereum to Polygon the new 50% of circulating supply figure raises from 11.10M to 15.12M units.

This proposal presents a Supply Cap of 17.2M units which was extracted from Chaos Lab’s earlier comment. As we can see the difference between the two methodologies has reduced.

The AIP is currently live for voting and closes soon.

Aave - Open Source Liquidity Protocol