"Trust in Allah, but tie your camel": The First Principle of Borrowing

Source: Qianglie Forum public account

A devout believer came to listen to the prophet's sermon. He was fully engaged, remembering every word of the prophet, and was the last to leave after all the other believers had dispersed.

He immediately ran back, exasperated, shouting: "Oh God! I came here riding a camel this morning, but now I can't even find a single hair from the camel! Prophet, I listened devoutly and believe in your absolute divine power, but my camel has been stolen. Is this how God rewards my sincere prayers?"

The prophet was silent for a few seconds and said: "Brother, believe in God, but you also need to tie up your camel."

In the world of cryptocurrency, there is no shortage of faith and sentiment. Whether it's Satoshi Nakamoto or Vitalik, the "crypto prophets" are believed without question, and their followers pursue dreams with fervor, step by step.

However, the prophet is not responsible for watching the camel. When we prepare to implement the prophet's vision in the real world, there are always some basic logics that cannot be ignored. Besides the prophet, we also need "first principles" and must find Aristotle's "most fundamental propositions in every system that cannot be violated or removed" to ride our own camel and safely reach our destination.

From identity to contracts, memes and structures, computation and cognition, the three questions of the soul of money, every discussion in "Wenli" digs deep into the "first principles" in the crypto world. There are certainly more than one "basket," and first principles may appear in every corner— for example, have you ever thought about what the "first principles" of the seemingly mundane concepts of "lending" and "liquidity" should be? When you passionately practice in the crypto world, is your camel tied up?

1. BendDAO Incident: "Liquidity" is like air, often unnoticed, but absolutely essential

BendDAO is an innovative attempt to provide liquidity for NFTs, but it still fell victim to the "lost camel" incident. A liquidity crisis a few weeks ago was even likened to a "subprime crisis" by Fortune ("Subprime crisis: how monkey JPEGs pushed a crypto lender to the brink of insolvency?").

But is this really a "subprime crisis"? Was BendDAO's camel properly tied? As usual, let's start with three soul-searching questions before we assess the "first principles of lending":

First Question: Why do people need to collateralize NFTs for loans? What is the value proposition of NFT lending?

(1) "Selling things that can't be sold" (cash-out premium): If a monkey's market bottom price today is 100 ETH, tomorrow it is 150 ETH, and the day after it becomes 200 ETH—your inner thoughts must be: the more it rises, the more uncomfortable it becomes. Because this means the locked "premium" is increasing, and you can only watch helplessly, unable to take it out.

At this point, if there is a place where you can collateralize the monkey at an X% collateral rate and take out X ETH for flexible use—this satisfies your value orientation.

(2) "Buying things that were originally unaffordable" (mortgage loan): You want to show off but can only sigh at monkeys worth hundreds of ETH.

At this point, if there is a place where you can just pay the down payment to exchange the monkey for an avatar to socialize with big shots, and then slowly pay back later—this satisfies your value orientation.

(3) Similar to financial markets, the most concentrated value orientation remains short-term borrowing needs. For example:

• You need to top up after a margin call at the exchange;

• You want to seize the opportunity for leveraged trades;

• You want to take advantage of GameFi opportunities: for instance, collateralizing a monkey to borrow money to buy shoes and earn money running in stepN;

• You want to donate to Ukraine, but your salary is still a week away;

• You are NFT-rich but FT-poor: using the monkey as collateral, rolling it every three months, you can obtain "long-term" BTC/ETH funds for other trades.

It is evident that "lending" is indeed an objectively existing demand. The question is, in what form should this value proposition be satisfied?

Second Question: Is BendDAO a bank?

Let’s look at BendDAO's main functions:

• Deposits: If you have idle ETH, you can deposit it in BendDAO and earn up to 9-10% annual interest.

• Loans: If you are a holder of one of the seven blue-chip NFTs or want to make a down payment, you can obtain a collateralized loan.

• Interest Spread: BendDAO lends out the ETH deposited by everyone at a higher interest rate (about 25%), and subsidizes borrowers with its governance token BEND to achieve essentially "zero-cost lending" (provided that BEND is regularly converted to ETH)—BendDAO profits from the interest spread.

Deposits, loans, interest spread—at first glance, this seems almost identical to what banks do, except that the collateral (NFT) has more volatile prices and poorer liquidity. Such "lending" has been operating in the real world for hundreds of years (if we count from the first silver coins lent out by the ancient Greeks for maritime adventures, it has been three thousand years), BendDAO is merely moving it to the virtual world.

But it is still not a "bank." No matter how innovative a lending platform is, whether in the real world or the metaverse, it cannot be called a "bank"—no matter how similar they may appear. "Wenli" has already provided a "mind-blowing" conclusion in the "three questions of the soul of money"—any institution that does not "create money" cannot be called a bank. The first principle of a bank is "to create money out of thin air."

"Lending" is a catch-all term that seems to fit any business model. However, upon closer inspection, there are roughly three business models in the world that can accomplish "lending":

(1) Commercial Banks: The banks in the real world need no explanation, but who are the banks in the crypto world?

Strictly speaking, all stablecoin projects are banks because they possess the "first principle of banks"—they issue new currency. Whether this "new currency" is used by anyone or will go to zero is another matter; as long as they engage in the creation of stablecoins, they are "banks."

(2) Pawnshop Model: Commonly seen in the evils of the old society—pawn a ten-thousand-dollar bracelet, and the pawnshop lends you 300 dollars in an emergency; afterward, either you redeem the bracelet with cash or it gets sold by the pawnshop.

(3) Pure Smart Contract Model: This refers to platforms like Compound and AAVE in the DeFi world that operate purely on smart contracts. You deposit assets on AAVE, borrow money, but have no creditor-debtor relationship with AAVE—everything is written into the algorithm and executed by smart contracts. Once asset prices drop, the automatic liquidation mechanism quickly brings in third parties to liquidate your assets for profit.

These three models all seem to be engaged in "lending," but the differences are significant. BendDAO's "peer-to-pool" model clearly employs the third type—attempting to introduce the "fungible token" smart contract lending pool into the world of individual NFTs.

Third Question: Is the lending model of the "fungible" world suitable for the "non-fungible" world?

Recently, NFT prices plummeted, unexpectedly triggering a run on BendDAO, with dozens of monkey collateral facing liquidation.

This timing is terrible. As we all know, everyone in a bear market is already on edge, and at this moment, "forced liquidation" is like a crisp gunshot, causing the market to drop without hesitation—more monkeys are liquidated, leading to a death spiral. This is akin to "the real estate bubble bursting, banks forcibly auctioning houses to cash out, and house prices continuously declining in a spiral." But monkeys are not houses; houses can still be lived in, while in a spiraling market, how many people are willing to hold on to their avatars instead of running away?

A bigger problem lies on the other end: when depositors discover that BendDAO has monkeys that cannot be recovered at a discount, they immediately realize that their principal deposited may not come back—thus a bank run ensues, and deposits in the treasury are quickly withdrawn, causing borrowing interest rates to soar.

At this point, the automatic liquidation mechanism is completely helpless; the liquidation rules are unreasonable, the thresholds are too high, the incentives are insufficient, and the time variables are high (details omitted). Fortunately, the project team promptly modified the liquidation rules, and with the market rebounding, the situation was resolved within days.

Where did the problem lie? It stemmed from the misuse of the logic of "trading" and "credit," applying an automatic lending mechanism meant for "liquid assets" to "illiquid assets."

Aave and Compound are based on the lending logic of liquid assets (fungible tokens). "Liquid assets" refer to those that can always be liquidated when needed—under this premise, we can certainly set a calculation rule: whether it is a 90%, 80%, or 50% liquidation rate, the key is "being able to sell when you want to." This is what can be called "liquid."

In the market for Fungible Tokens (FT), there is the concept of "cumulative depth"—the "price" of each FT is the record of its last transaction, while "liquidity" is the cumulative depth of current market buy orders.

However, for NFTs, which are "individual assets," there is no "cumulative depth," only an "effective buy order," which is the current highest bid. Therefore, there is no relationship between its "price" and "liquidity"; having a "price" does not mean there is "liquidity," nor does it mean it can be sold. As you know, this is referred to as "having a price but no market": a priceless monkey can still have zero liquidity—either due to no demand or no transactions.

It is well known that liquidation is based on "liquidity," not on "price"—and for something with liquidity equal to "0," how do we liquidate it?

The "disposal" of illiquid assets cannot be called "liquidation"—this is also a basic logic in the real world. Banks have "illiquid asset mortgage loans," such as home loans; they also have "liquid asset mortgage loans," such as margin financing (on-market stock pledge financing). The former's defaults can only be "disposed of and auctioned," while the latter's margin calls can be timely "liquidated" using warning lines and closing lines—assets differ, and thus the methods of tying up camels must also differ.

2. How should NFT lending be done reasonably?

The misuse of financial trading market logic does not mean that NFTs cannot be used for lending; it just means that using a "liquidity pool plus liquidation" model is indeed not very suitable. For NFTs, which are non-fungible and illiquid assets, a more reasonable approach might be:

(1) First: Do not emulate the liquidity pool model of commercial banks—you really can't learn it; they print their own money. Banks do not take depositors' money from the deposit pool to lend to you; they lend you money they printed themselves, so there is no issue of borrowers not repaying and thus not being able to pay depositors. Banks inherently have high leverage capabilities and possess infinite risk resistance in "lending."

Under the same risk control system, a lending platform can never compete with a bank. The Catholic doctrine divides sins into "seven deadly sins," and the financial industry has seven major risks—banks can touch all seven risks, at most suffering some injuries; while a lending platform, if it violates any one of them, may face dire consequences—possibly never to rise again—before it even gets to commit the other six sins.

(2) Pawnshop Model: Strictly speaking, using depositors' money to lend falls under the "pawnshop model," and among the three models, the pawnshop carries the highest risk and the most uncontrollable factors. Since lending platforms have zero risk resistance in the face of "non-repayment" and lack the "fungible" market depth to assist, the most reasonable method is to learn from the evils of the old society and set extremely low collateral rates—such as lending around a thousand dollars against a ten-thousand-dollar monkey might be more reasonable.

(3) Peer-to-Peer Model: This refers to the NFTFi model. If a loan can be fully tied to a specific monkey NFT, it indicates that the person wanting to lend money has an awareness of this monkey and is willing to acquire it. The borrower has the right to know, and the collateralized lending is specific to each asset, with each transaction being unique, and there is no automatic liquidation process. However, the downside is also evident: both parties may have to wait a long time to connect.

(4) Shareholder Rights (Liquidity Provider DAO): Shareholders lock tokens and, if necessary, "convert equity to debt," using the DAO's capital to buy back collateralized NFTs. It's simply transferring the risk of depositors to shareholders and the DAO.

There are two golden rules in the lending industry: first, never lend without collateral; second, never lend only on collateral (Never lend only on collateral, and never lend without collateral). If you cannot understand your clients better than they understand themselves, then firmly adhere to this basic rule: believe in God, but also tie up your camel.

3. The Magic of DeFi "Transparency"

To be frank. A lending platform that just experienced a bank run managed to get back on track within a week. After the community voted to modify the rules, depositors eagerly deposited their ETH back—this is unimaginable in traditional finance.

This is a completely transparent system where one can see at any time how much money is in the platform's treasury, whether it exceeds the loans given out, whether people have started depositing again, seeing borrowers gradually repaying, and seeing the platform successfully auctioning NFTs for ETH—there's no need to wait for BendDAO to issue an announcement; depositors automatically return—this level of "transparency" is almost impossible to achieve in the real world. In your bank branch, how much liquidity is there daily? Besides the treasury department, no one knows—and this is precisely the key factor that keeps banks safe: opacity.

In the financial system, complete "transparency" is a "revolution," using a new machine to destroy the old one. The financial system in the real world is like the apps in the Apple App Store—relying on the iOS operating system; if iOS has major flaws, no matter how transparent or well-designed the apps are, it cannot guarantee that the system will not fail.

Regardless, if we can't even find the "bugs," how can we treat them? "Opacity" will hinder our ability to accurately identify "bugs." This is why "DeFi" seems to always offer a glimmer of "revolutionary" hope—the models and logic of DeFi will at least improve the transparency of the financial system, providing a "diagnostic" tool—seeing what is happening allows us to decide which surgical knife to use.