[Nexo Institutional Vault] Collateral Onboarding Risk Evaluation

Legal Disclaimer: This communication is provided for information purposes only. This communication has been prepared based upon information, including market prices, data and other information, from sources believed to be reliable, but Maker has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation. This content is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. References to any digital assets and the use of finance-related terminology are for illustrative purposes only, and do not constitute any recommendation for any action or an offer to provide investment advisory services. This content is not directed at nor intended for use by the MakerDAO community (“MakerDAO”), and may not under any circumstances be relied upon when making a decision to purchase any other digital asset referenced herein. The digital assets referenced herein currently face an uncertain regulatory landscape in not only the United States but also in many foreign jurisdictions, including but not limited to the UK, European Union, Singapore, Korea, Japan and China. The legal and regulatory risks inherent in referenced digital assets are not the subject of this content. For guidance regarding the possibility of said risks, one should consult with his or her own appropriate legal and/or regulatory counsel. Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any decision. The content speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects, and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others.

  1. Summary Proposed Risk Parameters
  2. Introduction
  3. Nexo Summary
  4. Metrics and Analysis
  5. Risk Parameters

Summary Proposed Risk Parameters

INST-ETH-NEX

Starting Stability Fee: 1.5% - decreased after 400m combined debt migration (0.1% per additional 100m minted)
Origination Fee: 0% - increased to 1% after 400m combined debt migration
Liquidation Ratio: 120%
Minimum Active CR: 150% (peace under new DssCharter implementation)
Debt Ceiling: 900m
DC-IAM gap: 50m - enabled once migrated
DC-IAM ttl: 8h - enabled once migrated
Cut: 0.99
Step: 90 seconds
Buf: 1.20
Cusp: 0.4
Tail: 140 minutes
Chip: 0.1%
Tip: 300 DAI
lk.chop: 20%
Tolerance: 0.5
Ilk.hole: 50m DAI
Dust: 10.000 DAI

INST-WBTC-NEX

Starting Stability Fee: 1.5% - decreased after 400m combined debt migration (0.1% per additional 100m minted)
Origination Fee: 0% - increased to 1% after 400m combined debt migration
Liquidation Ratio: 120%
Minimum Active CR: 150% (peace under new DssCharter implementation)
Debt Ceiling: 600m
DC-IAM gap: 50m - enabled once migrated
DC-IAM ttl: 8h - enabled once migrated
Cut: 0.99
Step: 90 seconds
Buf: 1.20
Cusp: 0.4
Tail: 140 minutes
Chip: 0.1%
Tip: 300 DAI
Ilk.chop: 20%
Tolerance: 0.5
Ilk.hole: 30m DAI
Dust: 10.000 DAI

Introduction

As the DeFi ecosystem in general, and MakerDAO specifically, continues to grow at an increasing rate, new institutional participants, such as investment funds, financial institutions, and other crypto enterprises, are looking for ways to engage in, and be part of emerging DeFi opportunities. However, for many institutional actors, safe, customised, and efficient solutions need to be put in place before they can engage fully in the ecosystem.

For MakerDAO, one solution to meet institutional participants’ needs, is the newly proposed concept of Institutional Vaults. Institutional Vaults are agreed upon, permissioned, vaults between MakerDAO and institutional partners. Vault terms for the agreement are proposed and voted on by Maker Governance before the institutional partner can create the vault. The agreed-upon terms can benefit institutional partners primarily through increased capital efficiency and more stable borrowing rate predictability. This is achieved by offering favourable conditions that would otherwise not be possible with Maker’s permissionless vaults, such as lower liquidation ratios and a predetermined fixed stability fee. For MakerDAO, Institutional Vaults can help (i) increase DAI supply, (ii) generate more revenue, and (iii) incentivise long term borrowing commitment.

The Institutional Vault proposal suggests that three new mechanisms should be included in Institutional Vaults.

(i) an origination fee - an institutional partner will have to pay a fee at the time of minting DAI,

(ii) a set stability fee - a stability fee that is determined upfront and fixed for a period of time, and (iii) a vault starting size - to ensure a minimum amount of DAI is minted.

(iii) a vault starting size - to ensure a minimum amount of DAI is minted.

The first implementation of Institutional Vaults will work specifically with Nexo as Maker’s institutional counterpart. In addition, ETH and WBTC are proposed as the initial collateral. Hence, this report will assess the risks related to offering ETH-A and WBTC-A Institutional Vaults to Nexo. The smart contract implementation supporting Institutional and Long Term Vaults is explained by @hexonaut and @Derek in MIP59: DssCharter.

The proposed terms to Maker’s agreement with Nexo are presented in the table below:

Institutional Vault Type Permissioned ETH and wBTC Vaults
Qualifying Vault size: 400m
Debt Ceiling 1-1.5B
Origination fee: 1% of DAI minted (not including capital migrated from an existing vault )
Origination fee behaviour: Origination fee is not restricted and can be changed through Governance via a Parameter Changes Proposal. Typically this will mean that Nexo has a 2-3 week lead-time to any Origination fee rate change.
Origination fee future commitments Maker and Nexo will work together to propose reductions to the origination fee if the vault size grows significantly.
Stability fee: 1.5%
Stability fee behavior: Fixed for 6 months. An additional 100m DAI minted will reduce the stability fee by 10 basis points (per vault) to a minimum of 0.5% for 6 months. Stability fee reduction will be manual via an executive vote. After 6 months, open community dialogue facilitated by the Growth CU will determine ongoing stability fees. Note that these can only be increased by a maximum of 1% for each subsequent 6month period.
Collateral Ratio: 200 %
Collateral Ratio Behavior: Failure to maintain 200% CR over the majority of any 7 day period will result in increasing the SF to 5% and increasing the LR to 150%. The 7 day period would be calculated by ensuring that 95% of the OSM updates show that the CR is equal to or greater than 200%.
Liquidation Ratio: 120 %

Nexo Summary

Overview of Company

Nexo is a financial service provider for cryptoassets, founded in 2017 by Antoni Thenchev and Kosta Kantchev. The company focuses on providing a suite of financial services, such as (i) crypto credit lines, (ii) earning through lending, and (iii) crypto trading. Nexo is a for-profit private company with centralised control and custodianship.

Nexo also has its own token, NEXO (NEXO). The token has several use cases. For example, Nexo token holders have previously been given a daily dividend payout as a percentage of the earnings from the Nexo company. In aggregate, this has resulted in approximately $29m in dividends paid out over the last three years. Other benefits of the token include more competitive lending and borrowing rates on Nexo and the option to participate in certain governance votes on the platform. The market capitalisation of the Nexo token is approximately $1 billion, with a circulating supply of 560 million Nexo tokens, and a total supply of 1 billion Nexo tokens.

According to the Nexo website, the platform is currently available in 200 jurisdictions and has over 2 million users. They accept over 40 fiat currencies for their fiat onboarding, trading, and borrowing services.

Overview of Key Services

Exchange

On the Nexo exchange, 16 cryptoassets can be purchased with fiat. These include Bitcoin, Ethereum, Tether, Nexo token, USDC, DAI, Bitcoin Cash, Litecoin, EOS, Stellar, PAXG, Chainlink, Tron, Polkadot, Cardano, and Dogecoin. The Nexo exchange also has over 100 cryptoasset trading pairs, many of which are coupled with BTC, ETH, or NEXO. The Nexo exchange platform offers zero fees on trades and up to $100,000 USD per trade. They collect revenue via an increased spread. Nexo does not necessarily hold all the assets traded. Instead, they utilise their so-called “Smart Routing System’’. This is a system that connects to multiple external exchanges and splits market orders depending on the price per volume.

Earn (Lending)

The Nexo “earn” service offers up to 12% interest with daily payouts. Nexo supports 24 assets for deposits. The interest rate can vary between 1%-12% depending on which asset is deposited, and if the user owns any Nexo tokens while depositing. If a user owns Nexo tokens, a preferential rate of 2% more is offered to the lender. According to the Nexo website, over $50 million has been paid out in interest so far.

Retail Credit Line (Borrow)

The Nexo borrowing service, or “credit line” product, allows users to make crypto-collateralized loans in fiat or stablecoins. Nexo uses Loan-to-Value (LTV) ratios to determine how much a user can borrow from a specific collateral type. For example, the LTV for both BTC and ETH is 50%. This amounts to a 200% collateralization ratio (CR). Other collateral types, such as LTC, BCH, EOS, BNB, XLM, LINK, and TRX have a 30% LTV ratio, while the NEXO token LTV is 15%. A user can either borrow the whole credit line at once or smaller amounts multiple times until the credit line limit is reached.

Nexo uses a Loyalty Programme to determine the borrowing rates of their crypto credit line. The loyalty programme rewards users that hold high amounts of NEXO tokens. For example, if a user holds at least 10% of their portfolio in NEXO tokens, the borrowing rate amounts to 6.9%. If 5-10% of a user’s portfolio is in NEXO tokens, the borrowing rate increases to 8-9%. At 1-5% NEXO tokens, the borrowing rate is 12.9%. Finally, if less than 1% of the portfolio is in NEXO tokens, the borrowing rate is 13.9%.

Security and Insurance

Liquidation Mechanism

According to Nexo’s terms and conditions, borrowers should provide additional collateral and/or make repayments to rebalance their credit line if the LTV increases above a predefined threshold. The process of adding additional collateral to the Credit Wallet can be automated. In such a case, the required amount of a specific cryptoasset will be automatically moved from the Savings Wallet to the Credit Wallet to serve as collateral.

Nexo will liquidate the necessary amount of collateral to rebalance the Nexo Credit Line in case a loan’s LTV ratio increasing above a predefined maximum permitted threshold. Nexo states in the terms and conditions that they will devote significant efforts to notify the borrower in advance of liquidation. However, they also state that due to the nature of digital asset markets, it may not be technically possible to notify the borrower prior to liquidation. Ultimately, it is the borrower’s responsibility to monitor the market conditions and maintain an adequate LTV.

The Nexo platform incorporates a “collateral exchange feature”. This feature allows borrowers to swap employed credit line collateral for other cryptoassets with an LTV ratio equal to or higher than the existing collateral. This gives borrowers the added flexibility to convert negative trending assets into more stable collateral, such as stablecoins, to manage the credit portfolio and maintain a specific LTV ratio.

Insurance

Nexo’s total insurance portfolio amounts to $375 million. Their stated plan is to increase the insurance portfolio to $1 billion in 2021. Currently, Nexo uses both Ledger Vault and BitGo as custodian partners for hot and cold wallet custody, from whom they also receive the insurance.

The partnership between BitGo and Nexo was initiated in May 2018. In March 2019, Nexo announced an insurance policy of up to $100 million for cold storage Nexo Wallets on BitGo. This was made possible through the BitGo arrangement with the insurance firm Lloyd’s of London. This insurance policy covers digital assets where the private keys are held 100% by BitGo.

The partnership between Ledger Vault and Nexo was announced in February 2021. Through this partnership, Ledger Vault agreed to insure an additional $150 million through a program with insurance firms Arch and March.

Nexo’s Intended Plans with MakerDAO

If the proposal of the Institutional Vault agreement with Nexo is approved, the first step for Nexo will be to refinance their already existing Maker Vaults. Nexo’s existing Maker Vaults have approximately $1.88 billion in assets locked, with 573.16 million DAI in total debt. At the time of writing, the asset composition of the vaults is 86.74% ETH-A (531,493 ETH locked) and 13.26% WBTC-A (5,797.65 WBTC locked). The collateral ratio for the ETH-A vault is 329.02%, while the collateral ratio for the WBTC-A vault is 328.64%. The refinancing of Nexo’s outstanding debt would result in a lower liquidation ratio (120% according to proposed terms), which, in turn, would allow Nexo to lower their collateral ratio, and ultimately borrow more DAI. The proposal also states that for the refinancing of Nexo’s existing Maker Vaults, the Origination Fee will not be charged. However, a 1% fee for all additional DAI minted will be implemented.

With the conditions in the proposed Institutional Vault agreement, Nexo has committed to mint a further 100-200 million DAI initially, targeting 600 million DAI. Over time, depending on market conditions, the expectation is to increase the amount of DAI minted to 1 billion and above.

Metrics and Analysis

Solvency Risk

One of Nexo’s main incentives for minting DAI from Maker Vaults is to profit from the rate arbitrage between Maker Vaults and their own product offering. Currently, both ETH-A and WBTC-A vaults have a 2% stability fee, and a 145% LR. Nexo, on the other hand, offers a minimum of 6.9% borrowing interest rate, which can reach up to 13.9%, depending on how many NEXO tokens a user holds in their Nexo portfolio.

Nexo’s Current and Historical Collateral Ratios in Maker

Nexo has been able to maintain a relatively healthy collateral ratio over time. The two charts below illustrate the historical collateral ratio for both Nexo’s WBTC-A vault and ETH-A vault. The WBTC-A chart covers the period May 20, 2020, to August 10, 2021, while the ETH-A chart covers the period April 29, 2020, to August 10, 2021.

The average CR for the WBTC-A vault during the measured period was 286%. The CR dropped below 200% on 31 unique days within the 447 day period. The most notable period was between July 14, 2020, to July 30, 2020. Nexo’s CR was below 200% for the whole period. On July 16, Nexo experienced its lowest CR to date of 165.09%. It took 16 days to bring the CR back up above 200%. Another notable period was between October 1, 2020, to October 8, 2020, when the CR was below 200% for 8 days. However, it only reached 191.7%. For 2021, the CR dropped below 200% on two dates, May 19, and May 23, reaching 195% and 194% respectively, but recovering within one hour on both occasions.

Historical Collateral Ratio for Nexo’s WBTC-A Vault at MakerDAO


Source: Block Analitica

The ETH-A vault has in general held a higher CR compared to the WBTC-A vault. This is probably due to ETH-A making up the majority of Nexo’s vault composition. At the time of writing, 86.74% of all Nexo’s vault collateral is locked in the ETH-A vault (531,493 ETH). The average CR for ETH-A over the period measured was approximately 363%. Just like the WBTC-A vault, the most notable period for the ETH-A vault was in July 2020. From July 19 to July 29 the CR dropped below 200% on several occasions. It took about 10 days to stabilise the CR above 200% again. The lowest CR during this time, and for the whole period measured, was 178.83%. Other periods below 200% CR include September 5 to September 6, when the CR reached a low of 186.76%. It is also worth noting that Nexo’s ETH-A average CR for 2021 has increased to 443.7%, and has not dropped below 200% for the whole year.

Historical Collateral Ratio for Nexo’s ETH-A Vault at MakerDAO


Source: Block Analitica

Competitive Risk

Competition between Nexo and MakerDAO

Nexo and MakerDAO are both lending platforms that offer overcollateralized cryptoassets loans. There is however a big difference in both the underlying system, product offering, and organisational structure. As Nexo and MakerDAO are both lending providers, they can be viewed as competitors. However, one of MakerDAO’s key objectives is to issue DAI. Nexo does not issue stablecoins. Hence, in this important sense, they are not competitors. Furthermore, Nexo already makes use of Maker Vaults to generate DAI, which directly benefits Maker. By having access to Institutional Vaults, Nexo can either earn a higher margin on the DAI rate arbitrage or offer its customers more competitive borrowing rates on their platform.

Currently, there are two key reasons why Nexo borrowing services may be preferred to Maker Vaults: (i) Smaller retail customers may choose Nexo borrowing services, despite the higher interest rates, because they do not have the requisite funds to meet the Dust requirement levels for Maker Vaults. (ii) Larger retail and institutional customers may choose Nexo borrowing services because they prefer to borrow from a centralised legal entity with appropriate loan insurance.

Nexo is already using Maker Vaults. Hence, by offering more favourable terms to Nexo, through the use of Institutional Vaults, both parties may benefit. Maker will benefit from: (i) an increase in DAI supply, (ii) more diversified collateral composition from ETH-A and WBTC-A vaults, meaning less USDC dependence, (iii) greater revenue generation, and (iv) more advantageous incentives for long term borrowing commitment.

In terms of Maker’s key objectives, the benefits of Nexo adopting Institutional Vaults is likely to outweigh potential downside competitive risks and does not present any additional competitive risks over and above the existing status quo.

Precedent implications of providing Institutional Vaults to Nexo

Nexo would be the first institutional partner for MakerDAO Institutional Vaults. The Nexo onboarding would set a precedent for future institutional partners. This implies that the rationale, qualifications, and requirements for institutional onboarding should be well understood. How should an institutional partner be defined, and what criteria would form the basis of this definition? For example, variables such as (i) company size, (ii) company category, (iii) balance sheet assets, or (iv) minimum requirement for DAI minting might be considered.

The benefits of favourable terms and conditions mean that there is likely to be a high demand from many different actors in the ecosystem to access Maker Institutional Vaults. The key question is how should these partners be selected? A clear definition of qualifying criteria for an institutional partner may be important to select and distinguish future institutional partners. On the other hand, each potential partner may be considered on a case-by-case basis and ultimately agreed upon by the Maker Governance process. There may be a combination of both, as the process iterates going forward. However, a consensus on qualifying criteria might be helpful in terms of scaling and decision making.

MakerDAO is currently a permissionless platform, anyone who wishes to open a vault can do so. If conditions and qualifications are created for institutional vaults, what impact will this have on MakerDAO’s core values? Permissionlessness, openness, and censorship resistance are key attributes of MakerDAO and DAI. The way that qualifying criteria for institutional vaults impacts these values may also have to be considered.

Answering these questions is outside the scope of this report. However, discussion and resolution of these matters may be important in order to build a sustainable and scalable Institutional Vault service offering in the future.

Asset Liability Duration Matching

Normally, MakerDAO DAI liabilities should be matched with DAI assets in its duration. However, there is a risk of a matching discrepancy. We define this as an “asset liability duration matching risk”.

Currently, stablecoins make up circa 60% of all DAI backing. If DAI demand outpaces DAI supply, asset liability duration matching risk decreases. In turn, the utilization rate of the Peg Stability Module (PSM) increases. In such a scenario, the debt ceiling of a specific PSM is the key limiting variable preventing an excess of PSM minting.

On the other hand, if DAI supply exceeds DAI demand, PSM utilization decreases. An example of an extreme case is where a PSM is emptied of all deployed collateral. Increasing rates then becomes one key method to mitigate the imbalance and the subsequent downward effect on the DAI price. Other methods for disincentivizing minting include (i) setting higher collateralization requirements, (ii) higher dust levels, and (iii) higher penalty fees. If and when MakerDAO has to increase rates to mitigate this situation, variable-rate vault borrowers will be disproportionately affected in comparison to vault borrowers who enjoy a predefined fixed rate.

A severe DAI demand shock, potentially combined with a DAI supply increase would create an even more extreme scenario. If this happens, MakerDAO may have to consider increasing rates on fixed-rate vaults to protect the integrity of the DAI peg. The consequences of deploying such a solution would be highly undesirable. Increasing rates on a fixed-rate product negates the purpose and intent of the product. As a result, MakerDAO’s reputation may be damaged and lead to negative publicity.

These types of extreme scenarios can be alleviated by constraining the debt ceiling on fixed-rate vaults as a ratio of existing stablecoin backing. In our opinion, MakerDAO’s exposure to fixed-rate lending, with an average term of 6 months, should never exceed a ratio of 50%. Depending on the trajectory of DAI demand, even 50% may be too high. However, demand projections for stablecoins going forward are promising.

The proposed maximum exposure for the Nexo Institutional Vault loan is 1.5 billion DAI. In the near term, there is no evidence that Nexo would make use of the full 1.5 billion debt ceiling. However, if they do, this would result in approximately 40% of the current total stablecoin collateral at Maker. In our opinion, this limit should not be exceeded as other fixed-rate products are planned going forward.

Implementation Risk

This section outlines the new collateralization ratio mitigation feature implemented by the DssCharter module. The module protects Maker by restricting borrowers’ ability to increase their debt exposure or withdraw collateral while their loan is less than the agreed collateralization ratio, termed peace. In Nexo’s case, this collateralization ratio limit would be set to 150%. Nexo will not get liquidated until they reach the liquidation ratio of 120%. The tool is very useful as it protects Maker against unsafe borrowing behaviour. However, the inability to withdraw collateral presents a certain risk that is addressed below.

Large borrowers may sometimes protect against liquidations by first withdrawing collateral and then converting it to DAI through a manual transaction, with the objective of repaying the debt. With the DssCharter, this strategy will not be possible if the borrower’s collateralization ratio is already below the predefined peace at 150%. However, other strategies to increase collateralization ratios are still available, such as (i) top-up of additional collateral, (ii) loan repayment by using external sources, or (iii) flash loan DAI - withdraw collateral - convert collateral to DAI on DEX - repay flash loan.

It is crucial that borrowers are aware that certain types of transactions cannot be used to repay loans in the event that the collateralization ratio is below the peace. We have communicated this to Nexo and they have assured us that their vault risk mitigation strategies will rely on top-ups or repays from external sources.

Concentration Risk

From a liquidation risk perspective, it is not entirely clear which of the following is most beneficial to MakerDAO. (i) Many small vaults or (ii) a few large vaults.

A large number of small vaults creates a risk of many liquidations at the same time. A limited number of Keepers, and a large number of liquidations, means that the Keepers may not be able to liquidate all vaults effectively, due to large transaction activity, coordination issues and the gas expenditure requirements.

In contrast, a few large vaults that are liquidated simultaneously may present a greater tail event risk. The Liquidation 2.0 system would struggle to effectively liquidate such a large amount of capital at once. An inadequate amount of DEX liquidity combined with cascading liquidation effects would have negative consequences on, and limit, the auction throughput.

The key to assessing the trade-off between a few large or many small borrowers depends on knowing that large borrowers have good vault mitigation strategies in place. MakerDAO must have confidence that the large borrowers are consistent and reliable with their vault management. This is a preferable situation to not knowing how many small borrowers are managing their vaults, even though there is a low risk of them being liquidated all at once.

Nevertheless, if MakerDAO starts engaging with more large institutional borrowers, we will have to (i) quantify vault concentration risk metrics in more depth, and (ii) evaluate large borrowers more frequently. As a point of reference, Nexo and one other entity, that utilises 7 vaults, already represent about 40% of debt exposure minted from crypto collateral. Nexo alone currently represents 22% debt exposure from crypto collateral, which could potentially increase to a 36% share, if they mint the full available 1.5 billion debt ceiling.

Legal and Regulatory Risks - Outside the Scope of This Report

This report does not assess or account for legal and regulatory risks. However, these issues are relevant, important, and merit further discussion in another setting.

Currently, any agreement with Nexo on offering Maker Institutional Vaults would not be legally binding nor fall under any legal jurisdiction. The implications of this state of affairs mean that Nexo, or any other institutional partner, who cannot execute on the terms and conditions of the partnership agreement, may not face the binding penalties and consequences that are normally included in an agreement with jurisdictional enforcement.

The consequence of this is that MakerDAO will have to place a high degree of trust in Nexo to fulfil the promises, terms and conditions in the agreement. Nexo is a centralised corporate entity structured to operate within a defined legal jurisdiction. MakerDAO is currently not on an equal legal footing in this regard. Hence, any failure to execute on the agreement, for whatever reason, may place MakerDAO at a disadvantage in any dispute that needs to be resolved through a jurisdictional court.

Crypto lending has recently attracted increased regulatory attention from the SEC. Service providers such as BlockFi, Celsius, and Coinbase, are either the subject of enforcement action or have been threatened with enforcement action if they continue offering lending services. If the SEC makes a successful case that such services fall within the existing securities regulation, Nexo may also face the risk of an SEC action and enforcement. It is currently unclear how Crypto regulation may develop. There may also be legal challenges to any SEC rulings in the future. Be that as it may, regulatory risk assessments do not fall under the Risk Core Unit’s mandate. We are not equipped to assess these risks which are primarily legal in nature.

A long-lasting and scalable institutional vault product offering may need to create the infrastructure to analyse and assess these risks in more depth.

Summary of Notable Risks

  • The solvency risk is limited by the following factors. (i) A healthy CR maintained by Nexo throughout the last few years. (ii) The DssCharter implementation. (iii) The supposedly large amount of treasury assets held by Nexo, which enables Nexo to perform top-ups or repays in tail event scenarios. Nevertheless, if a vault of 500m debt size gets liquidated, the 120% liquidation ratio may not sufficiently protect Maker since it would potentially take many hours to liquidate it. Hence, we recommend that governance should react, in a timely manner, whenever CR drops under the proposed soft CR limit of 200%, as described in the terms above. In the past, Nexo has demonstrated a high CR for both their current ETH-A and WBTC-A vaults. However, this does not guarantee or confirm that Nexo’s protection automation will continue to work well under all conditions. If the overcollateralization buffer is lowered in the future, protection automation becomes more important. The Risk Core Unit will closely monitor Nexo’s CR maintenance and report back to the community.
  • As and when permissioned vaults are implemented, additional governance and PR risks can be expected. To date, Maker has not yet had to address this category of risk. In addition, as institutional vaults are rolled out, a number of other considerations need to be addressed. These include (i) competitive risks with other institutions, (ii) enabling similar products for multiple clients, and (iii) partially moving away from permissionless vaults. In order to address these issues, we suggest that Maker Governance implements a dedicated framework for these special vault types.
  • Fixed-rate vaults increase risks of asset liability duration mismatch, as described above. We suggest implementing a formula whereby governance limits the growth of fixed-rate products in relation to stablecoin PSM exposure.
  • There are certain implementation risks related to the migration of Nexo’s existing vaults to a new permissioned vault type. We believe the PE team will address and help manage these issues. Once Nexo has successfully migrated, they will need to be aware of the unwinding technique limitations imposed by the DssCharter. These are outlined above.
  • From both a business and risk perspective, concentration risks create challenges. Nevertheless, if we are assured that vault protection management is sound, we believe, from a liquidation risk perspective, that a high concentration of exposure towards a small number of large borrowers can be more optimal.
  • The regulatory and legal risks, related to enabling Institutional Vaults for Nexo, are outside the scope of this report. We are, however, aware of the developing regulatory environment in the US, which is currently focused on regulating centralized lenders and their yield products.

Proposed Risk Parameters

The proposed parameters are in line with the terms and conditions confirmed by the community in the recent poll. We also split the 1.5 billion debt ceiling exposure into two separate ETH and WBTC vault types according to feedback from Nexo.

We are using similar auction parameters as the ones currently being updated for ETH-A and WBTC-A vault types. The only difference is in higher penalty fee (ilk.chop) and higher pending auction throughput (ilk.hole). Because this vault type has a low LR of 120%, we believe Maker should be aggressive both on the collection of maximum penalty fees and faster liquidation throughput. This should maximally assure the borrower does not risk being liquidated.

INST-ETH-NEXO

Starting Stability Fee: 1.5% - decreased after 400m combined debt migration (0.1% per additional 100m minted)
Origination Fee: 0% - increased to 1% after 400m combined debt migration
Liquidation Ratio: 120%
Minimum Active CR: 150% (peace under new DssCharter implementation)
Debt Ceiling: 900m
DC-IAM gap: 50m - enabled once migrated
DC-IAM ttl: 8h - enabled once migrated
Cut: 0.99
Step: 90 seconds
Buf: 1.20
Cusp: 0.4
Tail: 140 minutes
Chip: 0.1%
Tip: 300 DAI
Ilk.chop: 20%
Tolerance: 0.5
Ilk.hole: 50m DAI
Dust: 10.000 DAI

INST-WBTC-NEXO

Starting Stability Fee: 1.5% - decreased after 400m combined debt migration (0.1% per additional 100m minted)
Origination Fee: 0% - increased to 1% after 400m combined debt migration
Liquidation Ratio: 120%
Minimum Active CR: 150% (peace under new DssCharter implementation)
Debt Ceiling: 600m
DC-IAM gap: 50m - enabled once migrated
DC-IAM ttl: 8h - enabled once migrated
Cut: 0.99
Step: 90 seconds
Buf: 1.20
Cusp: 0.4
Tail: 140 minutes
Chip: 0.1%
Tip: 300 DAI
Ilk.chop: 20%
Tolerance: 0.5
Ilk.hole: 30m DAI
Dust: 10.000 DAI

Sources:

11 Likes

50% of what? The total outstanding DAI? Or total outstanding being paid under variable rates?

USDC in the protocol does not pay a fee and due to this one fact alone can’t be considered in these calculations of variable rate borrowers. At this time Variable rate borrowers compromise perhaps 1/2 of this system? We already are at 50% as I see this.

A few other points.

I agree we very much want to accommodate institutional vaults.

My issue here is with the DC and rates potentially going below 1%. See below.

I would also prefer the origination fee be basically converted to a refundable vault performance bond.

I do not believe Maker can agree to this 1% rate change lock every 6 months. Or put a better way the SF rate discrepancy on any fixed term vault should not exceed more than 1/2 the average variable length rate for greater than 6-12 months. Maker is the only protocol that needs to use rates to manage liquidity demands (changing DCs can limit minting) but messing with LRs to manage critical situations is completely NOT the way to deal with liquidity issues and IF ever implemented will damage Makers reputation more than any single factor:

also penalty fees are already high.

Exactly how is one going to do this on the PSM without implication on the PEG? Now granted in a PEG downside scenario the PSM will probably be empty.

Then wouldn’t prudence suggest we just limit the available DC now to be that 50% mark of the available variable rate debt?

Again this is precisely where partial vault liquidations could help. Isn’t this again a good reason to have vaults with staggered CRs but an average CR for total funds they need to maintain to reduce this default liquidation clumping risk.

This is exactly the reason I prefer a slash-able vault performance bond that was based on the total DC* 5-10bps over the period that had terms that needed to be met. In this case for each 1B for 6 month term this would be 2.5M (in DAI) slashable and refundable (rolloverable) bond. What the slashing terms would be to be discussed. The fact that it can roll over means it isn’t a fee but a bond conditional on some minimum performance requirements is important. It forces IVs to commit to Maker as much as Maker is committing to them and the bond IS the Maker recourse for failure to satisfy agreements. Now should/could Maker put up a counter bond to the IV party for similar performance (maybe).

Not addressed is the fact that with no capital controls, and a hard PEG, Maker has only one other lever (rates) to control the DAI supply. I have seen little consideration in this report of this primary concern and hazard other than a speculation of how to apply capital controls, while somehow magically the PEG remains ‘managed’ and a 50% of something hard cap which is not addressed in the DC vault limits.

Of note here is a complete absence of discussion of any of the ‘fixed rate’ implementations. Deco, or other which makes one wonder if given the above analysis risk even feels the need to have any sort of system fixed rate implementation other than governance promising not to fiddle with rate knobs for fear of IV and/or PR backlash?

I am for the following:

A provisional 12month trial period for any IV looking for vaults.

In Nexo case.

1.5% with 10bps on each 100M with the vault rate never going below 1% and a DC of 900M (since according to risk they won’t need 1.5B) on both the ETH and wBTC vaults. This gives about 1.8B of borrow at as low as 1%.

BUT in exchange I want to have a performance deposit of 5bps on each 900M in DAI as a performance guarantee for the 12 months against the following:

  1. Removal of more than 1/2 the collateral or borrow dropping below the 400M mark
  2. At each point of 100M increase this level must be maintained or face losing the entire bonus gained for the period. Terms on this can be adjustable to just losing the 10bps on the 100M back down. As a penalty if the borrow goes below 400M the rate goes up beyond 1.5% to 1.6 for each 100M and the bond potentially forfeit in the period
  3. CR must be maintained above the agreed levels (I was thinking they should target 200 and we should only penalize/slash if they go below 180) otherwise the 10bps per 100M performance can be revoked or reduced and/or the bond slashed.

Other point: IMO rates for fixed vaults need to move at least 50% of the spread difference between similar vaults in the system and the current IV vault (based on CR, and collateral type). Honestly if I was negotiating this I would suggest Maker will put up a similar performance bond of 2.5bps on the 900M (2.25M or 4.5M on both vaults) that over say the next period we won’t raise rates more than 1% during the first 6month rollover. The bond only claimable at the end of the second period provided Nexo stays with us at this 400M minimum level through the higher than normal rate rise.

I personally think we should focus on a term agreement no more than the next 12 months, go through one renegotiation period with the intention if this works to roll it another year. So basically the term for Nexo would be to put up 4.5M on each 900M vault as a refundable performance bond for the year (9M). Maker will put up a 2.25M performance bond per vault 900M that we won’t increase rates beyond +1% at the 6 month mark claimable should Nexo stick around to the end of the 12 months.

The point here is I really want to do a deal. But I want both parties to put up a bond against the required/demanded performance measures (or failure to perform). I want this to be of a term that is exploratory, and with terms that don’t put us over limits. 900M on each DC with a bottom rate of 1% should be more than sufficient to encourage Nexo to take advantage, and for us to move forward on this. 12 months provisional is good enough for Maker to come up with a better set of longer term agreements for all Institutional investors.

Realize if this works reasonably well and everyone can be on board about the details we can easily extend these to longer terms and heap in additional details and add other vaults.

BTW: The whole performance bond is meant to be an absolute replacement for the origination fee since the origination fee is meant to stop IVs from shopping their debt needs around. I am of mixed mind on origination fee provided we have the performance bonds. I think these are better for Nexo and for Maker generally. It would allow us to compete better too I think.

Thank you for the report @primoz and I hope some real thought to this whole impossible trinity issue with respect to rate management will be forthcoming in the future.

1 Like

“A severe DAI demand shock, potentially combined with a DAI supply increase would create an even more extreme scenario. If this happens, MakerDAO may have to consider increasing rates on fixed-rate vaults to protect the integrity of the DAI peg. The consequences of deploying such a solution would be highly undesirable. Increasing rates on a fixed-rate product negates the purpose and intent of the product. As a result, MakerDAO’s reputation may be damaged and lead to negative publicity.”

At least this is being admitted to beforehand and not hidden away.

I see it as a classic case of underestimating the risk of black swans and thereby making incorrect inference from historical data. The end result is fragility.

1 Like

@MakerMan regarding MakerDAO getting a better deal or hedge against potentially offering too low rates in the future, this was already discussed and polled with the community a month ago, but community decided to go with terms as they are.

And this has to do mostly with the revenue aspect of Maker, not the risk aspect which we evaluate here. The risk aspect of fixed rates was covered in this report, see asset liability duration risks. We did mention a concern here:

It may not be clear: 50% ratio is meant versus stablecoin backing which sits at 3.8bn. In other words, we’d need to increase rates for vaults once 3.8bn gets drained if we wanted to protect the peg, but of course we probably wanted to react before it gets completely drained - and therefore proposal was at 50% (limit the exposure), which is quite conservative. Alternatively you could also increase DSR, but there is a limit to that.

I do share your concern what happens if we do need to react and limit/lower this exposure in say 12 months or later? We’d need to abandon the terms and this wouldn’t look ok. This is why we recommended some kind of formula to say how much Maker wants to lend fixed rate on specific duration, relative to stablecoin backing. But then this then needs to be part of the terms offered, which is not right now, all we have in terms is this and that may not be enough:

So in short, yes there is a concern from peg management risk aspect versus client relationship and we brought it forward. How big it is? - depends how you see DAI demand and supply moving next year and going forward. The other concern is more about revenue aspect which community decided on and we don’t evaluate.

2 Likes

Well then we both have expressed our concerns and the DAO has decided. Problem here will be when the second or third IV wants to be added with similar terms when it comes to this 50% mark. It is basically something which will limit IV DAI (or should at least according to risk).

Thanks for the reply.

One thing I was researching a bit and wondered if you had an answer to. What do rates to borrow look like if one factors in farming returns. Looked to me like rates across the board were either 0 or negative on Compound/Aave. Does your team have a list of actual borrow rates once farming returns are factored into the deposit and borrow sides?