Liqwid Risk Framework for Cardano Native Tokens

Liqwid Risk Framework

Introduction

Any borrowing and lending platform relies on the collateral factor of an asset to set the maximum amount to be borrowed against the asset or to determine the liquidation threshold (Liquidation LTV) for an asset. Overall, the collateral factor of an asset reflects the perceived collateral quality and risk of the underlying asset.

Therefore, the “Risk Framework” proposal aims to determine the relevant processes to:

  • Select the eligible assets to be added on the Liqwid platform, based on qualitative and quantitative criteria.
    Exclude the assets, which are not meeting the Liqwid minimal criteria.
  • Define the collateral factor and other market parameters for the eligible assets (e.g. borrow and supply caps).
  • Document the governance processes for any further votes on collateral factor or market parameter updates.

Methodology

Criteria to consider

In an efficient market, where the information is accessible to everyone, the price of an asset reflects the equilibrium between the offer and the demand. The price encompasses the body of knowledge used to support the decision-making of choosing a token over another.

Every price is influenced by market participants, where their actions can be measured through various data, such as:

  • Volume: It represents the total amount of transactions (buy or sell) in each market.

  • Volatility: It represents the standard deviation of a price over a period.

  • Maximum drawdown: It represents the maximum loss during a day that a coin can experiment.

Nevertheless, blockchain markets are also existing on two different ecosystems, the on-chain and the off-chain, which partially exchanged with each other. Therefore, the on-chain liquidity on decentralized exchanges is also a relevant criterion for Liqwid, to assess the risk of having collateral assets not being liquidated due to lack of on-chain liquidity causing losses via slippage.

As a qualitative factor, having an existing product and a community for a token is a strong indicator to better analyse the inherent value. Having a product implies that a competent team is behind the project, having completed, at a minimum, the work required to deploy the product on the Cardano mainnet. A diverse and engaged ecosystem of community members implies a project with a strong community of active users.

The Liqwid DAO, community of LQ delegates and voters, will analyze projects across all the metrics outlined above.

Circulating Supply (percent of tokens in circulation compared to total supply): As a general rule Liqwid will exclude any Cardano token from being supported as collateral with a circulating supply < 20%. This 20% minimum float criteria will apply as a general threshold to any token.

Scoring model

The “Primary score” seeks to capture the information known by the market participants at any point in time and used to measure a token’s fundamental and intrinsic value drivers.

Primary score: It computes the average score of the below criteria, where the lowest score between 90 and 365 days is considered for each criterion:

  • Volume: Average 24h volume during the last 90/365 days
  • Volatility: Standard deviation of the daily price return over the last 90/365 days.
  • Maximum drawdown during the last 365 days : Maximum price change intraday during the last 90/365 days.
Tier Maximum intraday drawdown over 90/365 days Volatility of the daily price return over 90/365 days Average 24h volume over 90/365 days Score
1 -10% up to 0% From 0% up to 2.5% Over USD 10 million 1
2 -25% up to -10% From 2.5% up to 5% USD 1 million - 10 million 0.7
3 -50% up to -25% From 5% up to 10% Tier USD 500’000 – 1 million 0.4
4 -100% up to -50% From 10% up to 100% USD 0 – 500’000 0.1

Final score: Based on the primary score, two other criteria are added:

On-chain liquidity: USD Amount of liquidity at the time of the analysis. This data is accessible using Taptools PRO.

DEX liquidity add-on per tier

Tier On-chain liquidity Liq. LTV add-on
1 Above $3 million 10%
2 Between $1-3 million 5%
3 Between $0-1 million 0%

Existing product: Assessment on the existence of a product for the related token.

Other considerations

Frequency for reviewing this criterion: The final score must be recalculated at least every 3 months or earlier in case of any major market / adverse event for the token listed on Liqwid.

Discretionary risk premium: Additional risks can be existing for some token/protocol under Liqwid DAO’s review for listing, and a specific haircut on the final score can be attributed to any token on a case-by-case basis.

Model limitation: Market data are still subject to a lot of uncertainty due to the inherent risks of such young projects and ecosystems. As we are collecting more data, we could revise this section in the future and update this model following a successful governance vote.

Core Team: Currently, the Liqwid Core team is responsible to performing the risk assessment and proposing appropriate risk parameters, and anyone can comment on the Liqwid governance forum. The Core Team is responsible for periodically initiating on-chain votes to launch new markets or to modify the market parameters for every token listed on Liqwid.

Score calculation

Step 1)

Primary score: Average (Min(Volume90;Volume365) + Min(Volatility90;Volatility365) + Min(MaxDrawdown90; MaxDrawdown365))

Step 2)

Primary score compared to a strong token (e.g. ADA Liquidation LTV) to assess the Liquidation LTV:

If PrimaryScoreTOKEN Z > PrimaryScoreADA then 80%; otherwise PrimaryScoreTOKEN Z * 80%

Step 3)

An add-on for on-chain DEX liquidity included in the Liquidation LTV calculation. The data are coming from TapTools, which sums all the on-chain liquidity of all the decentralized exchanges on Cardano.

Other risk parameters

The Liqwid protocol is also using other risk mitigating measures embedded in its code such as:

  • Supply cap, which is a limitation for a supplier to add a specific token after a threshold is reached.
  • Borrow cap, which is a limitation for a user to borrow a specific token after a threshold is reached.
  • Interest rate model, which determines the borrow and supply rates given the utilization rate for a specific token.

The cap thresholds and interest rate parameters are configurated on a case-by-case basis per token and through an on-chain governance vote.

We note that centralized exchange (CEX) liquidity is an important factor to determine the supply and borrow caps and can be consulted for each token on Coingecko (->Token-> Market-> +/-2% depth).

Liquidation factor and Liquidation discount

Definition

The liquidation threshold is the ratio where a borrowing position is liquidated.

Example: A user has borrowed USD 750 worth of DJED after supplying USD 1’000 worth of ADA tokens.

In this example, the Loan-to-Value is 80% (800 / 1’000). If the liquidation threshold is also 80%, then the loan is liquidated.

The liquidation penalty is the percentage taken from the seized collateral of a defaulting loan.

Example: A user is getting liquidated, when its LtV is reaching 75%. At this moment, considering a liquidation discount of 10%, USD 833.3 (= USD 750 / (1-10%)) of the collateral assets are seized. The defaulting user keeps the rest of the collateral (e.g. USD 166.7 worth of ADA), and USD 833.3 worth of ADA are being exchanged against USD 750 of DJED paid by a liquidator.

Risk factors per risk category

We summarize here below the different asset tiers and their related risk variables.

TIER Assets Liq. LTV Liq. Penalty Comments
ADA, Stablecoins 65- 80% 10%
1 AGIX 55.1% - 65% 13%
2 WMT,MIN,SHEN 45.1% - 55% 15% SHEN 12%
3 35.1% - 45% 17%
4 25.1% - 35% 20%
5 20%.1 - 25% 25%

Conclusion

The Liqwid Risk Framework is critical to ensure that borrowing and lending markets are safe for all participants on the Liqwid platform. By standardizing the way we analyze the quality of the assets, we can compare their inherent risks consistently and transparently arrive at a consensus amongst DAO members before submitting on-chain votes.

The Risk Framework can review any Cardano native token and determine appropriate risk parameters. For example, riskier assets should have lower liquidation thresholds with tighter borrow and supply caps. In this approach, it allows the protocol to support riskier assets without jeopardizing the overall health of Liqwid and its lenders.

Appendix

Scoring template

A separate Excel file has been created to summarize token market data and assist with quantitative analysis.

  • I support the implementation of the Liqwid Risk Framework as presented here above.
  • I do not support the implementation of the Liqwid Risk Framework as presented here above.

0 voters

2 Likes

I support this proposal.

1 Like