The Ultimate Form of DeFi Lending: How PFund Lending Mechanism Turns "Leverage" into a Rational Tool

"Leverage" has always been a high-risk financial instrument in the crypto market.

While it amplifies funding effects, it also frequently triggers systemic risks. In May 2022, during the UST de-pegging chain liquidation storm, major lending protocols (Aave, Compound, MakerDAO, etc.) triggered billions of dollars in forced liquidations within 72 hours. Users' collateral was liquidated at low prices by the system, leading to widespread position wipeouts. This extreme liquidation is an inherent systemic consequence of all lending protocols that use liquidation mechanisms—when market volatility is intense, the liquidation process exacerbates panic, triggering a death spiral of "price decline → liquidation → sell-off of collateral → further price drops."

PFund fundamentally cuts off the occurrence mechanism of this liquidation spiral at the protocol architecture level.

**Underlying Logic and Implementation Path of No-Liquidation**

The core reason traditional DeFi lending protocols rely on liquidation mechanisms is the risk of price decoupling between collateral and borrowed assets. When the market price of collateral assets (like ETH) plummets and their value can no longer cover the borrowed assets (like USDC), the protocol faces bad debt and must enforce liquidation to cut losses.

PFund’s lending design fundamentally avoids this risk exposure. Its implementation is based on two core rules:

● First: The borrowing limit is strictly anchored to the borrower’s own PFund fund pool shares.
- The maximum borrowing amount (quoteBorrowMax) for a user is calculated as:
- B0 = stakedToken × savings / (liquidBalance + stakedToken)
- Borrowing capacity = B0 - existing borrowings (borrowed)
- The essence of this model is that users withdraw liquidity equivalent to their PFund minimum guarantee shares. Users use these guarantee shares as collateral, and the value of these shares is insulated from AMM market price fluctuations, only increasing unilaterally with the protocol’s overall yield accumulation.

● Second: The borrowing limit is constrained by a dual rigid cap—whichever is smaller: the individual contribution cap and the pool’s available liquidity.
- Audit reports confirm: totalGrossBorrowed ≤ savingsQuote (the total borrowed amount never exceeds the pool’s available liquidity), and execBorrow enforces a pre-check that grossPreview > savingsQuote before execution, ensuring no single borrow causes the pool’s liquidity to run out.

Under this architecture, sharp fluctuations in AMM token prices do not threaten the safety of the borrowing system because the anchor for borrowing limits is "absolute shares in the fund pool," not "token market prices." This logic fundamentally eliminates the need for a liquidation mechanism.

**Code-Level Verification of Lending Safety**

The audit report verifies PFund’s lending safety based on five core invariants:

● I-1: Individual borrowing limits will not exceed their own PFund minimum contribution.
- Derivations show that the user’s maximum borrow limit is strictly calculated based on their savings (guarantee shares) and existing holdings, making it impossible for borrowing to encroach on others’ guarantee principal.

● I-2: Total borrowed amount never exceeds savingsQuote.
- savingsQuote represents the available liquidity of the PFund pool. Borrowing reduces savingsQuote in real-time; repayment restores it equivalently. The pool’s available liquidity forms a hard cap on borrowing scale.

● I-3: Borrowing does not affect others’ PFund guarantee principal and yield index.
- Borrow operations only modify savingsQuote (liquidity accounting), without touching totalSavingsShares (principal shares) or globalNVP (global yield index), thus safeguarding other holders’ rights.

● I-4: Users’ savings and nvp0 remain unchanged during borrowing.
- The function mFinalizeBorrow does not modify savings or nvp0. Only when users actively sell shares does the guarantee share count decrease; borrowing itself incurs no loss.

● I-5: Post-repayment, the released collateral ≤ original collateral, and remaining collateral can cover remaining debt.
- The algorithm logic in _repayUnstakeTokens ensures that whether partial or full repayment, the released collateral stays within safety thresholds, fully covering the debt exposure.

**Safety Boundaries of Circular Lending**

To address concerns that circular borrowing could infinitely amplify leverage and collapse the protocol, PFund’s mathematical structure provides a definitive defense mechanism.

After the first loan, the user’s borrowed data increases, and the next borrowing limit resets to B0 - borrowed_prev, forming a strictly decreasing trend. Simultaneously, the pool’s global available liquidity savingsQuote decreases due to borrowing. Each cycle reduces the user’s available leverage and the system’s overall leverage space. The borrowing system has an inherent self-regulating property, eliminating the need for external governance warnings. Because the leverage ceiling is determined by the protocol’s internal mathematical closed loop, it avoids external oracle manipulation risks, establishing the reliability of its no-liquidation architecture.

**Application Scenarios of Lending Strategies**

Taking a typical asset allocation as an example: a user invests 2000 USDT in a project, with 1000 USDT forming an AMM token position and 1000 USDT recognized as PFund fund pool shares. When the AMM position’s price doubles (up 100%), the user can adopt the following strategy:

● Stake tokens to trigger lending.
● Borrow up to an amount equivalent to their PFund guarantee shares in liquidity assets.
● Reinvest the borrowed funds to buy more of the project, increasing their PFund fund share proportion, thereby boosting future dividend rights.
● The entire process is interest-free, with no liquidation risk, only a one-time 3% borrowing fee.

This low-cost leverage avoids the risk of amplifying price gaming, instead focusing on increasing the foundation for stable returns (the dividend rights of the PFund pool). The lending mechanism becomes a foundational tool for enabling rational asset allocation.