Read the terms and liquidation mechanism of the NFT lending agreement in one article

Author: Nan Zhi

introduction

This article aims to explore the development path and trend of NFT lending by sorting out the lending terms and liquidation mechanisms of each NFT lending agreement and the important proposals that have been passed. Research objects include BendDAO, ParaSpace, JPEG’d, Blend, NFTfi, Arcade, Pine Protocol, Sodium.

All the parameters and proposals in this article come from the protocol white paper and the governance platform. The data comes from Blur, OpenSea, LooksRare, X2Y2. The caliber is the platform floor price data. The deadline is 2023/7/30, and the statistics are summarized through Dune.

List of liquidation mechanisms

This paper divides lending agreements into two categories:

• Point-to-pool type: the borrower borrows money from the collective fund pool;
• Peer-to-peer type: borrowers and lenders are matched through matching.

In peer-to-pool lending agreements, common features include:

• Both use decentralized oracle machines to feed prices, and the basic algorithm is time-weighted average price algorithm (TWAP) to avoid abnormal instantaneous fluctuations. In order to avoid floor price attacks, the specific algorithm is not disclosed;
• The borrowable ratio of BAYC, CryptoPunks, MAYC, and Azuki is significantly higher than other NFTs, and the liquidation threshold is usually better than other series.

In the peer-to-peer pool agreement, because the borrower can borrow immediately, and the overall market has not shown an obvious upward trend since the NFT bull market in April 2022, in order to control risks, each lending agreement only opens the lending function for some blue-chip NFTs, and sets a relatively Stringent control parameters.

For example, in BendDAO, borrowing funds with an LTV of 30% and a liquidation threshold of 65% requires a drop of 62.5% to trigger liquidation, while in ParaSpace, according to an LTV of 30% and a liquidation threshold of 65%, the drop liquidation rate is 53.8%.

Although follow-up market results show that multiple series of blue chips have experienced multiple declines of more than 50%, the following is the Moonbirds price curve starting from 2022/4/18. The floor price on that day was 18.9 ETH, and the highest point was on 2022/4/1 24 to 37.6 ETH, and to 2023/7/30 the floor price is 1.7 ETH, the largest drop of 95.5%.

**The point-to-pool agreement improves the matching efficiency of blue-chip NFTs, without the need to match the consensus of the borrowers and lenders on market conditions, and improves the turnover capacity of blue-chips. However, in order to ensure the safety of the overall fund pool, the economic efficiency is usually low. In addition, the point-to-pool agreement cannot solve the problem of rapid deterioration of the market environment by relying solely on the agreement mechanism. For example, the rapid decline and liquidation brought about by the release of Azuki Elementals will be dealt with by the temporary measures of the project party outside the agreement mechanism. If only relying on the auction liquidation mechanism There may be bad debts. **

Reference report: “ParaSpace: Due to the sharp fluctuations in Azuki’s price, the Azuki mining pool is suspended, including recharge, withdrawal, and strong equality functions”

Among peer-to-peer lending protocols, common features include:

*No floor price is usually involved, and whether the loan can be repaid on time is used as the trigger for liquidation;

• Usually an offer is made for the lender (that is, the parameter setting of the loan conditions), which means that the lender has an advantage in the market;
• Usually can be rolled over to defer liquidation.

The peer-to-peer protocol solves the long-tail asset lending problem through matching. Because the non-blue-chip NFT market is relatively small, the lender has an advantage, so the offer is mainly for the lender. Only in Sodium, because it can lend in many-to-one-multi-tiers, the borrower Make an offer. The liquidation is based on repaying the loan at maturity, which avoids the liquidation problem that the point-to-pool agreement may face after a rapid decline, but it also brings the possibility that the borrower may not consider repayment if the price continues to decline and does not recover. For this reason, Pine Protocol has set up the function that the lender initiates liquidation, but it needs to charge an additional 5% liquidation fee.

In addition, peer-to-peer agreements usually support extensions, but only Blend and NFTfi support rescheduling loan conditions, which can prevent market conditions from being too different from the original loan conditions. Pine Protocol’s targeted setting for this is to limit the number of loan days to only 7 days or 14 days, which reduces the possibility of large fluctuations. **Compared horizontally with the traditional bond market, changes in market interest rates are balanced by reverse changes in bond prices, but the NFT loan agreement does not have sufficient depth of secondary market support. It is also a potential development direction. Pine Protocol V3 has made an initial attempt to NFTize lending positions, which can be transferred between accounts. ** Therefore, for protocols that support lending over a quarterly period, the function of loan reconditioning is indispensable.

Proposal for lending terms and liquidation mechanism

The initial versions of the major lending agreements have relatively strict parameters for lending terms, and are also lacking in flexibility. In the process of development, either to deal with the liquidation crisis caused by the decline, or to promote the development of the agreement to improve economic efficiency, restrictions have been gradually relaxed. This section lists proposals for significant changes to the lending terms and liquidation mechanism of the lending protocol described in the previous section.

There are two main directions in the proposal: improving economic efficiency and improving liquidation.

Methods to improve economic efficiency include: increasing the borrowing ratio of top blue-chip NFTs, or increasing the borrowing ratio of rare funds in a certain blue-chip series. For example, in BendDAO, BAYC’s LTV in the initial version is only 10% greater than other NFTs, and the gap widens to 30% in BIP#15. In the ParaSpace V1.4 version, the LTV of golden skin BAYC and Koda increased by 6 times.

Ways to improve liquidation include: reduce liquidation, such as JPEG’d PIP-61 allows partial repayment, and delay the time of liquidation. Another category is to ensure successful liquidation and prevent unsold auctions. BendDAO’s BIP#9 removes the reserve price limit for auctions, which greatly increases liquidation profit margins. And the major agreements are shortening the auction time. According to JPEG’d statistics, most auctions take place in the last few hours or minutes. It is only necessary to ensure that users have enough reaction time to participate in the auction.

In order to study the relationship with the market situation, the time involved in the following is based on the time when the proposal starts to vote.

BendDAO

**Improve liquidation: **BIP#9 (2022/8/22)

Remove the restriction that the liquidation Bid price must be > 95% of the floor price;

The liquidation threshold is adjusted to 70%, and the adjustment is completed in three times;

Adjusted auction period to 4 hours: "The 48 hour window was designed to protect NFT holders from liquidation without waking up and losing their PFP. But now we have TWAP-protected on-chain oracles, which means floor Price attacks will be very difficult. So we want to reduce the time from 48 hours to 4 hours to improve the liquidity of the auction.”

**Improve economic efficiency and improve liquidation: **BIP#10 (2022/8/23)

The liquidation threshold is adjusted to 80%;

Adjust the auction period to 24 hours.

**Improve economic efficiency: **BIP#15 、BIP#15-1 (2022/11/11)

ParaSpace

Improving Economic Efficiency: ParaSpace v1.4 (2023/1/28)

Increase the mortgage value of rare NFT, such as Koda mortgage is six times that of regular Otherdeed;

Otherdeed LTV increased from 30% to 35% -40%.

JPEG’d

**Improve liquidation: **PIP-51 (2023/4/29)

After liquidation, the repayment time is reduced from 72 hours to 48 hours.

**Improve economic efficiency: **PIP-58 (2023/6/11)

Increase the LTV of each collateral, with a minimum of 5% and a maximum of 25% (BAYC);

Increase the pledged tokens to enhance the LTV function.

**Improve liquidation: **PIP-61 (2023/6/15)

Partial repayments are allowed to avoid liquidation.

**Improve liquidation: **PIP-71 (2023/7/17)

Auction time reduced from 24 hours to 12 hours.

NFTfi

Improve liquidation: Loan Renegotiations (2022/8/30)

NFTfi borrowers and lenders can renegotiate loan terms at any time before the loan is foreclosed (i.e. NFTfi’s liquidation).

Pine Protocol

**Improve liquidation: **PIP-8 (2023/4/28)

A liquidation fee is levied, payable by the lender when the lender initiates liquidation, in the amount of 5% of the loan amount.

**Improve economic efficiency: **PIP-12 (2023/7/12)

Limiting loan terms to only 7 days, 14 days increases market predictability and eases decision making.

Improve economic efficiency: Pine V3 (2023/1/13)

In V1 and V2 of the Pine protocol, the addresses of lenders and borrowers are written directly into the loan smart contract. This cannot be changed as long as the loan position exists. To make it more decentralized and flexible, lending positions are tokenized as NFTs in the Pine V3 smart contract design. Any blockchain address holding these NFTs will be the lender and debit of the corresponding loan position.

NFT blue chip index and time series of major changes

Based on the 2022/4/18 BAYC, MAYC, Azuki, Pudgy Penguins, CloneX, Doodles, Moonbirds and other 7 blue-chip prices, the NFT blue-chip index is obtained by averaging the relative percentages of daily prices as follows. CryptoPunks are not included because the series has not been traded for many days, which is easy to cause distortion, so it is excluded.

The following conclusions can be drawn from the relationship between the time when each proposal was launched and the index:

Improvement proposals for liquidation are usually made when the price is low. For example, in 22 years, BendDAO had four times when the loan interest rate exceeded 100%. When this happened for the second time in August, BendDAO changed significantly The liquidation mechanism is removed, and Bid’s reserve price requirement is removed.

Measures to improve economic efficiency have no obvious relationship with NFT prices, measures to improve LTV such as BendDAO BIP#15 and JPEG’d PIP-58 were proposed when prices were low, or related to the development needs of the protocol itself. Another potential factor is the health of the platform’s borrowed NFTs. Since the fourth loan interest rate of BendDAO exceeded 100% in September 2022, the loan interest rate for the remaining three months has been at a low level for the whole year, that is, the loan health coefficient is relatively high. , which becomes the window period for it to increase LTV.

With the crisis brought about by Azuki Elementals being resolved, lending agreements have also had the opportunity to re-improve the protocol mechanism and parameters. It remains to be seen whether the lending agreement can become a booster for the recovery of NFT, or there will be another liquidity crisis.