One way to test this is with data that has sufficient random walks (say for example daily open and close prices over several decades) and use the additive noise

One way to test this is with data that has sufficient random walks (say for example daily open and close prices over several decades) and use the additive noise method (ANM)[0] to determine if noise from one correlator affects noise in the other (and hence one causes the other). https://arxiv.org/abs/1412.3773 https://arxiv.org/abs/1412.3773