Be wary of discount rate risk: the mechanism and risk of AAVE, Pendle, Ethena's PT leveraged return flywheel

Author: @Web3 _Mario

Abstract: Recently, the work has been a little busy, so the update has been delayed for a period of time, and now the frequency of weekly updates is resumed, and I would like to thank you for your support. This week, we found an interesting strategy in the DeFi space, which has received widespread attention and discussion, that is, using Ethena's staking yield certificate sUSDe in Pendle's fixed income certificate PT-sUSDe as the source of income, and using the AAVE lending protocol as the source of funds to carry out interest rate arbitrage and obtain leveraged income. Some DeFi KOLs on the X platform have made relatively optimistic comments on this strategy, but I think the current market seems to ignore some of the risks behind this strategy. Therefore, I have some experience to share with you. In general, AAVE+Pendle+Ethena's PT leveraged mining strategy is not a risk-free arbitrage strategy, in which the discount rate risk of PT assets still exists, so participating users need to objectively evaluate, control leverage, and avoid liquidation.

Mechanism analysis of PT leveraged returns

First of all, let's briefly introduce the mechanism of this yield strategy, friends who are familiar with DeFi should know that DeFi, as a decentralised financial service, compared with TradFi, the core advantage is the so-called "interoperability" advantage brought by the use of smart contracts to carry core business capabilities, and most DeFi proficient people, or DeFi Degen's work usually has three:

1. Explore the arbitrage opportunities of interest rate differentials between DeFi protocols;

2. Find sources of leveraged funds;

3. Explore high-interest rate and low-risk-return scenarios;

The PT leveraged income strategy comprehensively reflects these three characteristics. The strategy involves three DeFi protocols, Ethena, Pendle, and AAVE. All three of them are popular projects in the current DeFi track, and they are just briefly introduced here. First of all, Ethena is a yield-based stablecoin protocol that captures short interest rates in the perpetual contract market on centralised exchanges with low risk through Delta Neutral's hedging strategy. In a bull market, the strategy has a higher yield due to the extremely strong demand for long positions by retail investors, so they are willing to bear the higher cost of fees, where sUSDe is its income certificate. Pendle is a fixed-rate protocol that decomposes the floating yield income certificate token into Principal Token (PT) and Income Certificate (YT) with a zero-coupon bond similar to a zero-coupon bond by synthesising assets, and if investors are pessimistic about future interest rate changes, they can lock in the interest rate level in advance by selling YT (or buying PT). AAVE, on the other hand, is a decentralised lending protocol where users can use designated cryptocurrencies as collateral and lend other cryptocurrencies from AAVE to increase leverage, hedge, or short.

This strategy is the integration of the three protocols, that is, using Ethena's staking income certificate sUSDe in Pendle's fixed income certificate PT-sUSDe as the source of income, and with the help of the AAVE lending protocol as the source of funds, interest rate arbitrage and leveraged income. The specific process is as follows, firstly, users can obtain sUSDe at Ethena, and fully exchange it for PT-sUSDe locked interest rate through the Pendle protocol, and then deposit PT-sUSDe into AAVE as collateral, and lend USDe or other stablecoins through revolving loans, repeating the above strategy to increase capital leverage. The calculation of the return is mainly determined by three factors, the base rate of return of PT-sUSDe, the leverage multiple, and the spread in AAVE.

The current state of the market and user engagement of the strategy

The popularity of this strategy can be traced back to AAVE's recognition of PT assets as collateral by AAVE, a lending protocol with the largest amount of funds, which released the financing ability of PT assets. In fact, before this, other DeFi protocols have long supported PT assets as collateral, such as Morpho, Fuild, etc., but AAVE can provide lower borrowing interest rates with more abundant loanable funds, which amplifies the yield of this strategy, and AAVE's decision is more symbolic.

Therefore, since AAVE supported PT assets, the pledged funds have risen rapidly, which also shows that the strategy has been recognised by DeFi users, especially some whale users. Currently, AAVE supports two PT assets, PTsUSDe July and PTeUSDe

May, the total supply has reached about $1B.

The maximum leverage currently supported can be calculated based on the Max LTV of its E-Mode, taking PT sUSDe July as an example, the Max LTV of the asset as collateral in E-Mode mode is 88.9%, which means that the leverage ratio can theoretically be about 9x through revolving loans. The specific calculation process is shown in the figure below, that is to say, when the leverage is maximum, without considering the flash loan or capital exchange cost caused by gas and revolving loans, taking the sUSDe strategy as an example, the theoretical return rate of the strategy can reach 60.79%. And this yield does not include Ethena points rewards.

Next, let's take a look at the actual participant distribution, still taking the PT-sUSDe pool on AAVE as an example. The total supply of 450M is provided by a total of 78 investors, which can be said to have a high proportion of whales and a large leverage ratio.

Looking at the top four addresses, the first 0xc693... The leverage of the 9814 account is 9 times, and the principal is about 10M. The 0x5b305 in second place... The leverage of the 8882 account is 6.6 times, and the principal is about 7.25M, and the leverage of analytico.eth in the third place is 6.5 times, and the principal is about 5.75M, and the 0x523b27 in the fourth place... The leverage of the 2b87 account is 8.35x, and the principal amount is about 3.29M.

Therefore, it can be seen that most investors are willing to allocate higher capital leverage for this strategy, but the author believes that perhaps the market is a little too aggressive and optimistic, and this deviation of sentiment and risk perception will easily cause large-scale stampede liquidation, so let's analyse the risk of this strategy.

The discount rate risk cannot be ignored

The author sees that most DeFi analysis accounts will emphasise the low-risk nature of the strategy, and even advertise it as a risk-free arbitrage strategy. However, this is not the case, and we know that there are two main risks associated with leveraged mining strategies:

1. Exchange rate risk: When the exchange rate between the collateral and the borrowing target becomes smaller, there will be liquidation risk, which is easier to understand, because the mortgage rate will become lower in this process.

2. Interest rate risk: When the borrowing interest rate increases, it may lead to a negative overall return on the strategy.

Most analysts will believe that the exchange rate risk of this strategy is extremely low, because as a more mature stablecoin protocol, USDe has experienced market tests, and the risk of its price de-anchoring is low, so as long as the borrowing target is a stablecoin type, the exchange rate risk is low, and even if de-anchoring occurs, as long as the borrowing target is USDe, the relative exchange rate will not drop significantly.

However, this judgement ignores the particularity of PT assets, and we know that the most critical function of the lending protocol is that it must be liquidated in a timely manner to avoid bad debts. However, there is a concept of duration for PT assets, and during the duration period, if you want to redeem the principal assets in advance, you can only trade through the AMM secondary market provided by Pendle. Therefore, the transaction will affect the price of the PT asset, or the yield of the PT, so the price of the PT asset is constantly changing with the transaction, but the general direction will gradually approach 1.

After clarifying this characteristic, let's take a look at AAVE's oracle design scheme for PT asset price. In fact, before AAVE supported PT, the strategy primarily leveraged Morpho as a source of leveraged funds, where the price oracle for PT assets used a design called PendleSparkLinearDiscountOracle. To put it simply, Morpho believes that during the duration of the bond, PT assets will receive income at a fixed interest rate relative to the primary assets, ignoring the impact of market transactions on interest rates, which means that the conversion rate of PT assets relative to the primary assets is constantly increasing linearly. Therefore, it is natural to ignore the exchange rate risk.

However, in the process of studying the oracle scheme of PT assets, AAVE believes that this is not a good choice, because the scheme locks in the yield and is not adjustable during the duration of PT assets, which means that the model cannot actually reflect the impact of market transactions or changes in the underlying yield of PT assets on the price of PT, and if the market sentiment is bullish on the change of interest rates in the short term, or the underlying yield has a structural upward trend (such as a sharp rise in the price of incentive tokens, new revenue distribution scheme, etc.), which may cause the oracle price of PT assets in Morpho to be much higher than the real price, which can easily lead to bad debts. In order to reduce this risk, Morpho usually sets a benchmark interest rate that is much higher than the market interest rate, which means that Morpho will actively reduce the value of PT assets and set a more ample room for volatility, which in turn will lead to the problem of low capital utilisation.

In order to optimise this problem, AAVE adopts an off-chain pricing solution, which can enable the oracle price to follow the pace of structural changes in PT interest rates as much as possible, and avoid the risk of market manipulation in the short term. We will not discuss the technical details here, there is a discussion dedicated to this issue in the AAVE forum, and interested partners can also discuss with the author in X. Here we will only present the possible price following effect of PT Oracle in AAVE. It can be seen that in AAVE, Oracle's price performance will be similar to the piecewise function, which follows the market interest rate, which is higher capital efficiency than Morpho's linear pricing model, and also better mitigates the risk of bad debts.

Therefore, this means that if there is a structural adjustment in the interest rate of PT assets, or when the market has a consistent direction for interest rate changes in the short term, AAVE Oracle will follow this change, so this introduces discount rate risk to the strategy, that is, assuming that the PT interest rate rises for some reason, the price of PT assets will fall accordingly, and the excessive leverage of the strategy may have liquidation risk. Therefore, we need to clarify the pricing mechanism of AAVE Oracle for PT assets, so that we can rationally adjust the leverage and effectively balance the risk and return. Here are some of the key features for you to think about:

1. Since in the mechanism design of Pendle AMM, liquidity will be concentrated towards the current interest rate over time, which means that the price changes brought about by market transactions will become less and less obvious, and it can be said that the slippage is getting smaller and smaller. Therefore, the expiration date is approaching, the price change caused by market behaviour will be smaller and smaller, and for this feature, AAVE Oracle has set up the concept of heartbeat, which is used to indicate the frequency of price updates, the closer to the expiration date, the larger the hearbeat, the lower the update frequency, that is, the lower the discount rate risk.

2. AAVE Oracle will follow a 1% interest rate change as another adjustment factor for the price update, which will trigger a price update when the market rate deviates from the Oracle interest rate by 1% and deviates for more than hearbeat. Therefore, this mechanism also provides a time window to adjust the leverage ratio in time to avoid liquidation. Therefore, for users of this strategy, it is necessary to monitor interest rate changes as much as possible and adjust the leverage ratio mechanism.