Liquidity Crisis in Crypto Lending. What Should We Learn? (Part 2)

2023-02-17, 04:25

[//]:content-type-MARKDOWN-DONOT-DELETE
![](https://gimg2.gateimg.com/image/article/167660652811.png)
## Types of Crypto Lending
Crypto lending can be classified by lending venues, credit rating, collateral types and counterparties.

### By Lending Venues
Crypto lending could happen in both centralized (CeFi) or decentralized (DeFi) venues. Currently, the share of centralized crypto lending exceeds 90% of the crypto lending market size.

CeFi venues are mainly exchanges and lending platforms. Exchanges (such as Binance, OKX, [Gate.io](https://www.gate.io/ "Gate.io"), etc.) are the most accessible lending places for individual lenders. Lending platforms in this article refer to those lending platforms or financial service platforms. Loans from exchanges are often required to be used only in the exchange, while loans from lending platforms can be moved among different venues, so they are more flexible and often used for institutional players. Currently, there are multiple leading CeFi lending platforms in the market, such as BlockFi, Celsius, NE<area>XO, etc..

There is an active inter-institutional lending network which is used when some lending institutions have a temporary liquidity gap, or the treasury departments want to make arbitrage on interest rates. However, under the recent liquidity crisis triggered by LUNA / UST, stETH and GBTC, the redemption by panic lenders forced institutions to withdraw their allocations, causing a liquidity crisis for the whole crypto lending market.

DeFi venues are protocols built on blockchain-based smart contracts. So far, most [DeFi protocols](https://www.blockchain-council.org/defi/defi-protocols/ "DeFi protocols") (such as Co<area>mpound, Aa<area>ve, BenQi…) implement the cash pooling model to achieve the rapid matching of loans and real-time adjustment of loan interest rates, while very few protocols are in the broker model like Maple Finance.

One key difference between CeFi lending and DeFi lending is the service. DeFi lending protocols can be lower in interest rate than CeFi, but the terms are too rigid for borrowers, such as initial and maintenance LTV, and there is no buffer time before forced liquidation.

### By Credit Rating
As bank loans can be divided into credit loans and secured loans (including credit enhancement measures like mortgage, pledge, etc.), crypto lending can be similarly categorized as unsecured and secured loans.

Lenders normally lend to borrowers based on collateral, but some borrowers with a strong track record and good reputation can borrow under or even without collateral within a credit limit, which can either be a formal loan agreement, or an earning program for investment.

For borrowers with relatively weaker credit (individuals, small institutions, etc.), the most common way is secured loans by posting collateral over the loan.

Besides the collateral loan (or bilateral loan), there’s another way of loan by DMA prime brokerage service (PB), where PB provides trading accounts of centralized exchanges funded with loans to the borrowers (usually qu<area>ant traders). The borrowers should provide and keep certain collateral in the trading accounts, and pay trading fees and interests to PB. It is a common way in the market for asset managers to leverage up and get lower transaction fee rates in exchanges. We will publish articles to discuss Crypto Prime Brokerage business in detail in the future.

### By Collateral Types
There are generally two types of collateral in the market for crypto lending: Standardized and Non-standardized collateral.

Cryptocurrency itself, with the characteristics of strong liquidity, is always accepted by lenders as standardized collateral. In most cases, stablecoins are usually borrowed using large-cap cryptocurrencies such as BTC and ETH as collateral. And stablecoins such as USDT and USDC might be pledged to borrow non-stablecoins in rare cases.

Non-standardized collateral refers to the collateral that is not cryptocurrency. As long as the borrowers and lenders reach an agreement, those crypto assets like blockchain and DeFi assets (DeFi AMM LP tokens, blockchain addresses, etc.), mining machines, NFT or even SAFT, can be used as collateral for loans. For example, NE<area>XO recently issued a loan of 1,200 ETH by receiving two [CryptoPunks Zombies NFT](https://mobile.twitter.com/zcryptopunks "CryptoPunks Zombies NFT") as collateral. The interest rate of this 60-day loan was 21% P.A.. The interest rate is much higher than the market average to cover the high risk in collateral value (NFT).

### By Counterparties
Depending on who they lend to, crypto lending institutions can be divided into to-B (Business) and to-C (individual).

To-B lending institutions means their sources (lenders) and allocations (borrowers) of capital are both strictly from institutions. Leading players are Anchorage and Genesis.

For to-C lending institutions, there might be both individual and institutional as their lenders and borrowers. Leading players are BlockFi, Celsius and NE<area>XO. To-C crypto lending institutions usually face higher requirements in regulations. In February 2022, BlockFi was fined 100 million USD by the SEC for ‘failing to register the offers and sales of its retail crypto lending product’.

## How to Build A Crypto Lending Business
### Crypto Lending Procedure
Crypto lending institutions use similar procedures to traditional lending:
![](https://gimg2.gateimg.com/image/article/167660702777.png)
### Key Factors of Success
Risk Control (pre- or post-loan issuance) is definitely the core of the lending business.

In the crypto loan business, institutions tend to focus on post-loan issuance risk control, i.e. collateral management.

### Optimized Lending Solution
For lenders to attract borrowers, cost of capital (interest rate) is definitely the most important factor. While borrowers also care about other service features, like liquidation mechanism, cross-order collateral management…

Better service to borrowers might mean higher risk for lenders, so lenders have to balance the risk and service level, and seek support from various financial tools to protect their downside.

### Risk Control Overview in Pre- and Post- Loan Issuance
The lenders are exposed to two major risks. The credit risk of borrowers not fulfilling the obligations to repay the principal and interest, and the collateral depreciation risk, which means the value of collateral drops lower than the total principal and interest.

Below table shows the main measures to address those risks.
![](https://gimg2.gateimg.com/image/article/167660716688.png)
### Pre-loan
Counterparty Rating and Credit Limit Setting

Lenders know their borrowers by due diligence on the borrowers’ business and financial status.

However, credit rating is more subjective than objective in the crypto industry, simply because financial reporting and auditing of crypto institutions is not as regulated and solid as traditional companies. Those large and well-known institutions might be rated higher than they should be and get over-leveraged. (LUNA / UST as a recent example)

Credit line can be set either in notional (loan value) or exposure (collateral value) terms.

### Regulatory Approval
Crypto lending institutions can be regarded as banks or security issuers in some jurisdictions, especially those to-C institutions, thus requiring a license to conduct business for local clients.

There were cases leading to-C crypto lending institutions like BlockFi and Celsius being fined for to-C lending business, but no such case from to-B institutions like Anchorage and Genesis.

KYC and AML are also important to ensure the counterparties are using capital for purposes complying with the regulations. Nowadays travel rules are rolled out in different jurisdictions.

### Post-loan
Counterparty Operating Status Tracking

In the post-loan stage, lenders need to track the borrower’s credit status and regularly evaluate the rating and credit line through the latest operating status, financial conditions, etc., to make sure their counterparty has the ability to repay the agreed principal and interest.

### Collateral Management
The target of [collateral management](https://en.wikipedia.org/wiki/Collateral_management "collateral management") is to ensure the value from disposal of collateral can cover the principal and interest. Most institutions track LTV (Loan-to-Value, has to be <100% to keep collateral more valuable than loan, and usually starts from 60-65% in secured loans), some others prefer to use collateral coverage ratio which is the inverse of LTV (so this has to be >100% and usually starts with 150-160% in secured loans)

The collateral in traditional lending includes ‘liquid assets’ (stocks, bills, etc.) and ‘illiquid assets’ (real estate, land use rights, etc.). For ‘liquid assets’, licensed financial institutions usually use public market price as the valuation of collateral. For ‘illiquid assets’, financial institutions usually hire professional 3rd party agencies to evaluate fair market value. In the post-loan stage, it’s easy to monitor value and liquidate liquid assets, and difficult to handle illiquid assets.

In the crypto lending market, most popular collaterals like BTC and ETH are extremely liquid. Those trading markets (exchange and OTC) operate 7*24 non-stop. Therefore, compared with traditional lending, post-loan collateral disposal in crypto lending is easy.

However, since the volatility in crypto price is much higher than in the traditional securities market, lenders have to monitor the real-time value of collateral and take actions when certain thresholds are reached. Only when the loan business scale is very small (e.g. limited loan orders, very simple collateral), it is possible for lenders to manually manage the collateral (setting alerts on crypto prices, manually sending margin calls or performing forced liquidations). After reaching a scale threshold, software can manage the collateral in an efficient and scalable way. From 1Token’s experience with clients, this threshold is at 10-20 loan orders.

Nowadays, there are new non-standard collaterals in the crypto industry, which could fall into ‘liquid assets’ (AMM LP token, blockchain address) and ‘heavy assets’ (miners, VC investment ‘SAFT’, NFT).

### Cashflow Projection
Cash pooling model exposes lending institutions to liquidity risk.

When all liquid assets have been consumed to redemptions, the lending institution will have to get immediate liquidity either from short term borrowing (from other retail users, or from other institutions) or quit their position (stETH’s role in Celsius incident), with a cost that incurs negative effects to their profitability or even business continuity.

Treasury departments of to-C crypto lending institutions should take Cashflow projection and asset liquidity ratio as a high priority.

### [Stress Test](https://corporatefinanceinstitute.com/resources/knowledge/credit/loan-stress-test/ "Stress Test")
Crypto lenders need to know when they need to request further collateral for each order / borrower, or which price ranges are most ‘dangerous’ that trigger margin calls or forced liquidations for a large chunk of loan orders.

On the other hand, when collateral value increases with coin price, borrowers will apply to increase loan or withdraw collateral. Lenders should be able to simulate those scenarios and judge how to fulfill borrower’s request while still keeping a healthy LTV / collateral coverage ratio.

### Maker-checker Approval and Audit Trail
Key operations through order lifetime require manual confirmation, e.g., when lender receives collateral from borrower’s exchange account, lender cannot automatically match the deposit to borrower, so manual confirmation is necessary.

Functions like different user’s roles (maker/ check), approval process and operation log are required for large institutions to prevent mistakes, and provide necessary documentation to meet the standard from internal control and external auditors.

### Analysis Reports
For financial reporting purposes, lending business data should be fed into accounting software like NetSuite and QuickBooks.

For the middle-back office, there are various analyses and reports for internal and external functions, across the company (Treasury view) or for each book / desk. Such as asset and liabilities, income and cost, asset liquidity, counterparty balance clearing and settlement, invoice and statements…

To generate such analysis with a spreadsheet is possible for a small lending scale. After reaching a scale threshold, a software is much more efficient and scalable, and an API connection or an export file that can be loaded into accounting software directly would save a lot of effort.

### Optimized Lending Solution
Lenders need to fully understand how borrowers operate with the loan and potential risks.

Within acceptable risk, lenders need to design the right terms (price and service) to attract borrowers with the following considerations.

### Interest Rate

Price (interest rate) is decided by cost (sources of capital) and profit. Normally, whoever has lower costs can offer a lower price to the market.

Main sources of capital can be (cost ranked from high to low) retained earnings, shareholder’s funding, external funds raised from earning programs, or borrowing from other institutions.

Generally speaking, crypto lenders expect higher yields than those in traditional finance, so nowadays crypto lending institutions are trying to seek funding from traditional finance by bridging their assets into the crypto industry, to reduce capital cost.

### [LTV](https://www.investopedia.com/terms/l/loantovalue.asp "LTV") (Initial LTV and Maintenance LTV)
There was a comparison on 2021 August between NE<area>XO and Celsius [4], based on a Bi<area>tcoin collateral of $5 million worth, how much they can max. lend.

The max. loan is 4.5 million USDC on NE<area>XO while 2.5 million USDC on Celsius.

If other terms are close, NE<area>XO definitely has higher attractiveness to borrowers.

### Cross-collateral
Cross-collateral means sharing collateral between orders.

It’s common between two counterparties to have multiple lending orders, with multiple coins in collateral and loan, and different entry prices. Sometimes there are other financial services between two institutions, such as OTC trading, structured products, and margin trading by DMA prime brokerage… To fit those scenarios, in crypto lending institutions there can be two-levels of cross-collateral mode.

–       1st level is to share collateral between multiple collateralized lending orders

–       2nd level is to share borrower’s equity / collateral among multiple business, including collateralized lending

Cross-collateral will be helpful to maximize borrower’s capital efficiency, but requires the lenders to precisely monitor risk across multiple orders and business.

### Delayed Liquidation
Delayed liquidation means there’s a buffer time from 1st margin call till forced liquidation.
Normally this is only applicable in institutional lending, where there are two LTV levels of margin call. Let’s call them a) liquidation alert and b) liquidation execution.

a) Liquidation alert is usually at LTV 85-95% (depending on the collateral coin’s liquidity and loan size) where the lenders usually offer a buffer time up to 24 hours from margin call.

If borrowers wish not to be liquidated, they have to 1) respond to lenders’ margin calls and 2) show proof of capital abundance and deposit attempt. In a drastic market like March 12 2020, blockchain is jammed, and deposit could take several hours. So a transfer record snapshot can be acknowledged.
  
If the borrower fails to respond to margin calls or provide proof within agreed time, lenders will trigger liquidation to recover the loan.

b) Liquidation execution is usually above LTV 95% (sometimes can be above 100%) where the lenders will immediately liquidate collateral to cover the loan.

Different lenders might set different parameters and procedures about those two margin calls. And overall, this delayed liquidation mechanism is favored by borrowers due to flexibility.

### Potential Advancements
In the current crypto lending ecosystem, we see potential for crypto lending institutions to enhance their competitiveness.

#### Real-time monitoring of market rates

If the optimal loan interest rate across different CeFi and DeFi venues can be monitored, sources of capital and assets in the market can be effectively connected. Moreover, financial institutions can carry out lending and arbitrage in the best place to improve their profitability. Currently, Paradigm has established a lending marketplace to match the RFQ of borrowers and lenders.

#### Non-standard collateral loans
As mentioned above, non-standardized collateral like NFTs and mining machines are already recognized as collateral in some cases. Lending institutions should establish a reasonable valuation mechanism for non-standardized collaterals based on market’s demand, so as to increase attractiveness to borrowers.

Here are some worthy considerations for setting up the standards: Instead of taking the floor price, is it possible to use the market price in Opensea (or other mainstream marketplaces) for NFT valuation? Should mining machines be valued based on market quotations, or total future cash f<area>low, or even both?

### Defensive measures
Non-performing loans are inevitable for lenders, and the bad debt provisions will have huge impacts on lenders’ profitability.

The major risk to lenders is the market fluctuation, when the market crashes, the drop in collateral value dramatically increases the risk on the lenders’ side. So it is always good to hold certain positions of put options to remedy the potential loss caused by market plummet.

It will not be a surprise if in the future credit insurance instruments emerge in the crypto industry. For example, CDS (Credit Default Swap) which helps transfer the risk of credit default from lenders to a 3rd party. There are three parties in CDS transactions: the borrower, the lender (CDS buyer) and CDS seller (banks or insurance companies). Similarly, in the crypto industry, lenders can also pack the non-performing crypto loans and purchase CDS from banks or insurance companies, the risks therefore are mitigated and CDS sellers will get premium instead.

***Disclosure: This article is from 1Token, crypto native technology provider since 2015.
1Token is a software provider for crypto-financial institutions, offering a one-stop tech solution. For more information on 1Token, please visit [https://1token.tech/](https://1token.tech/ "https://1token.tech/")***

**Reference**
1.Lending Global Market Report 2022 – By Type (Corporate Lending, Household Lending, Government Lending), By Interest Rate (Fixed Rate, Floating Rate), By Lending Channel (Offline, Online) – Market Size, Trends, And Global Forecast 2022-2026
https://www.thebusinessresearchcompany.com/report/lending-global-market-report

2.Genesis 2021 Q4 Market Observations
https://genesistrading.com/wp-content/uploads/2022/01/Genesis21Q4QuarterlyReport-final3.pdf

3.Arcane Research: Banking on Bi<area>tcoin The State of Bi<area>tcoin as Collateral
https://www.ceicdata.com/en/indicator/united-states/non-performing-loans-ratio

4.NE<area>XO vs Celsius: A Comparison of Crypto Lending Platforms
 https://p2pmarketdata.com/blog/<a href="/price/nexo-nexo" target="_blank" class="blog_inner_link">NEXO</a>-vs-celsius/
 
5.The Crypto Credit Report, Issue 7, Q4 2020
https://credmark.com/pdf/CryptoCreditReport-Q4-2020.pdf

6.VanEck research：Crypto Lending and the Search for Yield
https://www.vaneck.com/us/en/blogs/digital-assets/matthew-sigel-crypto-lending-and-the-search-for-yield/

<div class="blog-details-info">
<div>Author:** Yixuan Huang**, BD Director of 1Token; ** Phil Yang**, Head of BD of 1Token
<div class="info-tips">\*This article represents only the views of the researcher and does not constitute any investment suggestions.
<div>\*Gate.io reserves all rights to this article. Reposting of the article will be permitted provided Gate.io is referenced. In all cases, legal action will be taken due to copyright infringement.
</div>