OKX's cycle lending product has been criticized by many CEX users, which actually proves that slightly complex DeFi products like this are not suitable for direct integration into CEX at this stage. The issues that arise include things like the APY dropping from over 30% to just above 10% in no time, and excessive wear and tear during mining. The reasons are: First, stablecoin swaps are not entirely 1:1, especially for emerging stablecoins like USDG, which still heavily rely on on-chain liquidity. Factors like AMM slippage, price spreads, and volatility caused by concentrated swaps come into play. Additionally, you have to consider interest rate changes on platforms like Aave. If it’s something like the previous USDe, where redemption required lock-up, then unwinding the cycle and exiting becomes painful. You’d have to go through another round of swap wear and tear. When many people unwind leverage at the same time, on-chain premiums can occur. This kind of strategy is better managed manually or by writing scripts to monitor premium/discount situations. Essentially, this has nothing to do with OKX or Pendle—it’s just a natural phenomenon. It’s clear that if you want to make such products accessible to the masses, directly integrating the underlying without any processing won’t work. If you add a layer of strategy management between the CEX and DeFi ends—incorporating path optimization, wear and tear reduction, and risk control—then package it as a vault, it could work. Plus, if you negotiate with the project team for some subsidies, it would be much better.