APY looks tempting on the surface, but when you click into the contract, it's a three-layer nested structure, and no one knows who's borrowing from the bottom-level fund pool. I've run scripts to dig through a few aggregators, and some of the underlying protocols have TVL lower than my wallet balance—their risk classification is as good as useless.

Regulatory pressure has been tightening recently around here, and on-ramp/off-ramp channels are narrowing. You stare at the on-chain high yields and jump in, only to find the withdrawal channel blocked when you try to exit—that's what real liquidity risk looks like. Anyway, now when I look at an aggregator, I first check the underlying collateral, then calculate the fund pool concentration, and put the APY number on the back burner.

Data doesn't lie, but people use data to lie to you.