We study the estimation and prediction of the risk measure Value at Risk for Cryptocurrencies. Using Generalized Random Forests (GRF) (Athey et al., 2019) that can be adapted to specifically fit the framework of quantile prediction, we show their superior performance over other established...