A relevant question in the econometrics literature is whether a jump in one stochastic process Z triggers a jump in a related process Y. Starting with the work by Li, Todorov and Tauchen (2017), several papers have discussed this issue, typically in the situation where a jump in Z forces Y to have a jump as well, with the size of the jump in Y given as a function of the simultaneous jump in Z. Asymptotics are then derived in a high-frequency setting, often with the functional relation being linear and based on finite activity jumps in Y and Z. In this talk, we will discuss more realistic scenarios, including infinite activity jumps and a more classical regression assumption, namely that the jump sizes in Y are not given exactly by a function of the corresponding jump size in Z, but involving additional i.i.d. errors. We will sketch how asymptotical results can be obtained in two different situations.