Blockchain records are an incredibly rich source to extract intelligence related to the behavior of relevant parties in the crypto ecosystem such as exchanges, miners, large holders (whales), long-term holders and many others. From manual labeling to machine learning classification models, there are several strategies that help identify the entities associated with blockchain addresses. That information can enable strategies that detect trading signals based on the flow of capital in and out of the relevant addresses such as an exchange’s hot wallets that are accessible online. From that perspective, blockchain represents a new source of alpha that can spark new types of quant strategies specifically tailored to crypto assets.
3 Factors That Make Quant Trading in Crypto Unique — CoinDesk
