Recently, I’ve been looking at all kinds of “address profiling”—a lot of tags, clustering, and those fund-flow arrow charts that look convincing enough. But if I’m being honest, you can only treat it as a reference: the same person can hold dozens of addresses, and the same address could also be a custodian or a consolidation point. Change the parameters of the clustering algorithm, and the conclusions change. I’m not really comfortable using it as a basis for trading.

For me, what’s more practical is to go back to lending: around the collateralization ratio and the liquidation threshold, are there any abnormal adds to collateral or collateral withdrawals? If the interest spread suddenly widens, could it be that someone is moving liquidity around in advance… These are more reliable than “some giant whale is buying.”

Recently, the rate-cut expectations and the whole dollar-related narrative have come back again, and the noise around risk assets rising and falling together is pretty loud. I need to be reminded: it’s fine to listen to macro stories, but risk control shouldn’t follow emotions—first, leave enough buffer around your liquidation line.